Convention Program

AAEESS 112244tthh CCOONNVVEENNTTIIOONN PPRROOGGRRAAMM May 17 – 20, 2008

RAI International Exhibition & Congress Centre, Amsterdam

The AES has launched a new opportunity to recognize 1University of Salford, Salford, Greater student members who author technical papers. The Stu- Manchester, UK dent Paper Award Competition is based on the preprint 2ICW, Ltd., Wrexham, Wales, UK manuscripts accepted for the AES convention. Forty-two student-authored papers were nominated. This paper gives an account of work carried out The excellent quality of the submissions has made the to assess the effects of metallized film selection process both challenging and exhilarating. polypropylene crossover capacitors on key sonic The award-winning student paper will be honored dur- attributes of reproduced sound. The capacitors ing the convention, and the student-authored manuscript under investigation were found to be mechani- will be published in a timely manner in the Journal of the cally resonant within the audio frequency band, Audio Engineering Society. and results obtained from subjective listening Nominees for the Student Paper Award were required tests have shown this to have a measurable to meet the following qualifications: effect on audio delivery. The listening test methodology employed in this study evolved (a) The paper was accepted for presentation at the from initial ABX type tests with set program AES 124th Convention. material to the final A/B tests where trained test (b) The first author was a student when the work was subjects used program material that they were conducted and the manuscript prepared. familiar with. The main findings were that (c) The student author’s affiliation listed in the manu- capacitors used in crossover circuitry can exhibit script is an accredited educational institution. mechanical resonance, and that maximizing the (d) The student will deliver the lecture or poster pre- listener’s control over the listening situation and sentation at the Convention. minimizing stress to the listener were necessary ***** to obtain meaningful subjective test results. Convention Paper 7314 The winner of the 124th AES Convention Student Paper Award is: Modeling Frequency-Dependent Boundaries as Digital Impedance Filters in FDTD 09:30 and K-DWM Room Acoustic Simulations —Konrad Kowalczyk (Presenting Author) P1-2 Perceptual Study and Auditory Analysis on and Maarten van Walstijn, Digital Crossover Filters—Henri Korhola, Matti Convention Paper 7430 Karjalainen, Helsinki University of Technology, Espoo, Finland To be presented on Monday, May 19 in Session P18 Digital crossover filters offer interesting possibili- —Room and Architectural Acoustics and Sound ties for sound reproduction, but there does not Reinforcement exist many publications on how they behave perceptually. In this paper phase and magnitude * * * ** errors in digital implementations of linear phase FIR as well as Linkwitz-Riley crossover filters are Friday, May 16 13:30 Room H studied perceptually and by auditory analysis. In Standards Committee Meeting on SC-02-02 Digital a headphone simulation listening experiment we Input/Output Interfacing explored the just noticeable level of degradation due to crossover filter artifacts. In a real loud- Session P1 Saturday, May 17 speaker experiment we explored rough guidelines 09:00 – 12:00 Room C-D for “safe” filter orders of linear-phase FIR crossover filters, which would not produce audible LOUDSPEAKERS, PART 1 errors. Possibilities to predict the perceived errors were then explored using auditory analysis, Chair: Stanley Lipshitz, University of Waterloo, including also third-octave magnitude spectrum Waterloo, Ontario, Canada and group delay as simple auditory correlates. Linear-phase FIR crossovers were found to pro- 09:00 duce different kinds of phase errors than Linkwitz- Riley crossovers. The auditory analysis can quali- P1-1 Audio Capacitors. Myth or Reality?—Philip tatively explain the perceptibility degradation. Duncan,1 Paul Dodds,2 Nigel Williams2 Convention Paper 7315 ¯

Audio Engineering Society 124th Convention Program, 2008 Spring 1 Technical Program 10:00 11:30 P1-3 The Air Spring Effect of Flat Panel P1-6 Audibility of Phase Response Differences in Loudspeakers—Daniel Beer,1 Michaela a Stereo Playback System. Part I: Headphone Schuster,2 Michael Jahr,2 Alexander Reich1 Reproduction of Wide-Band Stimuli—Geoff 1Fraunhofer Institute for Digital Media Martin, Sylvain Choisel, Bang & Olufsen A/S, Technology, Ilmenau, Germany Struer, Denmark 2Technical University Ilmenau, Ilmenau, Germany The audibility of phase distortion in sound repro- Flat panel loudspeakers are characterized by duction systems has been the subject of many their low manufactured depth. Compared with studies. However, it remains a topic of contro- conventional loudspeakers the space saving versy, in particular in the field of loudspeaker or integration in existing surroundings is an advan- headphone equalization. Most studies lead to tage. From the acoustics point of view disadvan- the conclusion that, although phase distortion tages come along with the low manufactured may be audible for specific stimuli, in realistic lis- depth that influence the reproduction in the lower tening situations in a room, they will go largely and middle frequency range. Based on mea- unnoticed. These studies, however, have surements and FEM-simulations the reasons for focused on monophonic phase distortion; a this behavior were analyzed. Supplementary severe limitation, since ignoring phase response methods for solving this problem have been con- in equalization can result in different phase dis- sidered that are derived from conventional loud- tortion in different channels. It is the purpose of speaker technologies. the present study to investigate the audibility of Convention Paper 7316 stereophonic phase mismatch in the specific case of headphone reproduction. In addition, the 10:30 implications on microphone design and production are discussed. P1-4 The Inertial Air Load of a Loudspeaker Convention Paper 7319 Diaphragm—John Vanderkooy, University of Waterloo, Waterloo, Ontario, Canada, and B&W Group Ltd., Steyning, West Sussex, UK Workshop 1 Saturday, May 17 A typical bass loudspeaker driver has an inertial 09:00 – 10:30 Room L air load, which is about 30% of its actual cone mass. This air mass is often poorly understood, PRODUCTION OF SURROUND SOUND WITH HDTV but it is significant in defining the resonance frequency; and the purpose of this paper is to Chairs: Kimio Hamasaki, NHK Science and understand the concept, clarify important as- Technical Research Laboratories, pects, and present corroborative measurements. Tokyo, Japan The immediate surroundings of the diaphragm Christian Hugonnet, Audio Engineering determine the low-frequency air load, and mea- Consultant, Paris, France surements on a test driver with different mount- ings arrangements are made and assessed, including measurements in vacuum. A loud- Panelists: Florian Camerer, ORF - Austrian TV speaker box presents its own complications. Toru Kamekawa, Tokyo University of Fine Simulations are used to show how the air load Arts and Music, Tokyo, Japan depends on baffle size. In general, the air load may not be accurately represented by the usual HDTV is currently broadcasted in many countries over approximations that apply to a piston in an infi- the world. 5.1 channel surround sound has made great nite baffle or to a freely oscillating disk, but they progress in digital broadcasting. Many workshops have do give a rough estimate. discussed the 5.1 channel surround sound at previous Convention Paper 7317 AES conventions; however, 5.1 channel sound production and broadcasting accompanying HDTV has not 11:00 been discussed as often at previous conventions. This P1-5 Horn Loudspeaker Nonlinearty Comparison workshop will focus on the 5.1 surround sound produc- and Linearization Using Volterra Series— tion and broadcasting for HDTV and discuss various Delphine Bard, University of Lund, Lund, Sweden issues such as sound recording, sound design, postproduction, and aesthetic approaches in terms of 5.1 chan- The characterization of a weakly nonlinear elec- nel sound with large and high-resolution pictures. troacoustic device with usual methods of measurement (THD, intermodulation) does not Tutorial 1 Saturday, May 17 illustrate the nonlinearities themselves, but only 09:00 – 12:00 Room N101 some of their effects. Device linearization can be achieved by applying the inverse nonlinearity PERCEPTUAL AUDIO EVALUATION upstream of the device, under the condition that the nonlinearity law is known in detail. This paper presents nonlinearities behavior compari- Presenters: Søren Bech son of horn loudspeakers of different frequency Nick Zacharov ranges using an experimental method of weak The aim of this tutorial is to provide an overview of per- nonlinearity characterization and compensation, ceptual evaluation of audio through listening tests, based based on a representation of the nonlinearity by on good practices in the audio and affiliated industries. Volterra series using multitone excitations. The tutorial is geared to anyone interested in the evalua- Convention Paper 7318 tion of audio quality and will provide a highly condensed

2 Audio Engineering Society 124th Convention Program, 2008 Spring overview of all aspects of performing listening tests in a used are usually well managed corporate net- robust manner. works with good Quality of Service (QoS) and usu- Topics will include: ally high bandwidth. Due to its availability, the Internet is also increasingly used for various cases (1) Definition of a suitable research question and asso- of radio and television contribution, especially over ciated hypothesis; longer distances. However, the use of high bit (2) Definition of the question to be answered by the rates and reliable contribution transmissions over subject; the Internet cannot be guaranteed. Correspon- (3) Scaling of the subjective response; dents have the choice in their equipment to use (4) Control of experimental variables such as choice of either ISDN or the Internet to deliver their reports. signal, reproduction system, listening room, and selec- More than 20 manufacturers now provide equip- tion of test subjects; ment for audio over IP applications. The EBU has (5) Statistical planning of the experiments; and issued and verified a standard, EBU TECH 3326- (6) Statistical analysis of the subjective responses. 2007, which allows for interoperability between The tutorial will include both theory and practical previously not compatible Audio over IP codecs. A examples including discussion of the recommendations plug-test between nine manufacturers held in Feb- of relevant international standards (IEC, ITU, ISO). The ruary 2008 proved that earlier incompatible units presentation will be made available to attendees and an now can connect according to the new standard. extended version will be available in the form of the text Convention Paper 7322 Perceptual Audio Evaluation, authored by Søren Bech [This paper was presented by Mathias Coinchon] and Nick Zacharov. 10:00 Tutorial 2 Saturday, May 17 P2-2 Audio Fingerprint and its Applications to 09:00 – 11:00 Room B Peer-to-Peer Systems—Antonello D’Aguanno, Goffredo Haus, Università degli Studi di Milano, DIGITAL AUDIO SIGNALS, FILTERS, Milano, Italy AND EQUALIZERS In this paper we want to analyze the applicability Presenter: Jamie Angus, University of Salford, Salford, of audio-fingerprint technology to peer-to-peer Greater Manchester, UK systems. Audio-fingerprint is a technology commonly applied to scopes like audio identification In this tutorial we will first present the principles of or digital rights management. Peer-to-peer is a sampling and the basic form and function of digital filters, common Internet paradigm to share various digi- explain the link between impulse response and frequen- talized contents. We propose an improvement cy response, and how it leads to different forms of for typical peer-to-peer architectures (query filtering. Next we will relate that to the similarities and dif- flooding, centralized directory, hybrid architec- ferences between the designs of analog filters and digital ture) that permits the application of audio finger- filters. Finally, we will discuss some of the filter structures print technology to these systems. that might be used for audio equalizers. Convention Paper 7321

Saturday, May 17 09:00 Room H 10:30 Standards Committee Meeting on SC-02-08 Audio- File Transfer and Exchange P2-3 Audible ICMP Echo Responses for Monitoring Ultra Low Delayed Audio Streams—Alexander Carôt,1 Alain B. Renaud,2 Christian Werner1 Saturday, May 17 09:00 Room K 1University of Lübeck, Lübeck, Germany Technical Committee Meeting on Audio for Games 2Queen’s University Belfast, Belfast, Northern Ireland, UK Session P2 Saturday, May 17 Playing live music on the Internet is very 09:30 – 12:00 Room E-F demanding in terms of delay, loss, or jitter and hence requires extremely reliable network condi- AUDIO NETWORKING tions. Jitter is the most problematic factor because it has a direct influence on the required Chair: Sascha Spors, Technische Universität Berlin, network buffer sizes for receiving low delay Berlin, Germany audio streams. Therefore, measuring the amount of jitter is a very complex task due to the multi-hop architecture of the Internet. So far it 09:30 has been impossible to know at which hop these P2-1 EBU Tech.doc. 3326 for Interoperability delay variances appear. The authors propose a between Audio over IP Units—Lars Jonsson,1 solution that is able to generate an audible Mathias Coinchon2 impression of the jitter problem for each hop. 1Swedish Radio, Stockholm, Sweden Convention Paper 7320 2European Broadcasting Union, Geneva, Switzerland 11:00 Audio over IP end units are now common in P2-4 A Grid-Based Approach to the Remote radio and TV operations for streaming programs Control and Recall of the Properties of over IP networks. The units are used to create IEEE1394 Audio Devices—Philip Foulkes, contribution circuits from remote sites or local Richard Foss, Rhodes University, Grahamstown, offices into main studio centers. The IP networks South Africa ¯

Audio Engineering Society 124th Convention Program, 2008 Spring 3 Technical Program Typically, the configuration of audio hardware In this workshop experiences and utilization by manufac- and software is not integrated. This paper turers and users of the EBU RF 64 standard file format discusses a software system that has been for surround and stereo in TV and radio production will developed to remotely control and recall the be reviewed. AES Standard Working Group SC-02-08 properties of IEEE1394 (FireWire) audio devices have begun working on the adoption of this format as an via a series of graphical routing matrices. The AES standard. The format began to be used with radio software presents sound engineers with a graph- automation systems in Germany and in editors. It has ical routing matrix that shows, along its axes, the now found applications also for file-based TV 5.1 audio available FireWire audio devices on a FireWire operations in Sweden and elsewhere. network. Inter device connection management may be performed by selecting the cross points Workshop 3 Saturday, May 17 on the grid, and intra device control may be per- 10:30 – 12:00 Room O formed via device editors that are displayed via the axes of the matrix. The software application RECORDING STUDIO DESIGN may be hosted by a compatible Digital Audio Workstation (DAW) application to allow for the storing and recalling of the various properties Chair: Fritz Fey associated with the devices. Editor of Studio Magazin, Fritz Fey has successfully Convention Paper 7323 developed its activities to studio design (Studioplan). He will present various types of studio concepts, from musi- 11:30 cian personal rooms to more sophisticated installations, with testimonials of users. P2-5 Can the Public Internet Be Used for Being presented will be some VIPs (very important Broadcast?— Simon Daniels, Audio Processing points) on studio design from a practical point of view Technology, Belfast, Northern Ireland, UK that are REALLY helpful for participants. There are differ- This paper will look at a number of examples of ent aspects that come across besides the usually well- remote broadcasts over contended IP links and bundled package of formula and theoretical know how, examine the key points in their success. We will like basic geometry, ergonomics, economics, stereo + talk about issues such as jitter and latency and surround, working atmosphere, multi-use, loudspeaker considerations regarding essential features on IP environment, and many others. This workshop will show codec equipment. The experiences of major Euro- that studio design is not just a matter of absorbers and pean broadcasters trialing audio over the Public reverb time issues. Internet will form the basis of a discussion of the pitfalls and possibilities associated with using the Live Sound Seminar 1 Saturday, May 17 public Internet for essential broadcast links. 10:30 – 12:00 Forum Convention Paper 7324 DEMYSTIFYING AUDIO

Historical Event Presenter: Ralf Zuleeg, d & b HISTORICAL CAFÉ Saturday/Sunday/Monday 10:00 – 17:00 This session describes the principles of acoustics from Tuesday 10:00 – 13:00 the wavelength to the properties of a good loudspeaker. Room N The session is spiced with demonstrations and experiments to make it easy to digest the information. In the historical Café there will be on display a selection of vintage CD and compact cassette equipment from the early days; a book selection of vintage acoustic matters; Saturday, May 17 11:00 Room H a Slide Rule used for electro acoustical calculations; Standards Committee Meeting on SC-05-05 samples of triode applications (100 years triode use) Grounding and EMC Practices

Saturday, May 17 10:00 Room K Saturday, May 17 11:00 Room K Technical Committee Meeting on Coding of Audio Technical Committee Meeting on Hearing Signals and Hearing Loss Prevention

Workshop 2 Saturday, May 17 Special Event Saturday, May 17 Room B 10:30 – 12:00 Room L 12:00 – 13:30

USE OF RF 64: A FILE FORMAT FOR SURROUND AWARDS PRESENTATION AND KEYNOTE ADDRESS SOUND Opening Remarks: • Executive Director Roger Furness • President Bob Moses Chair: Lars Jonsson, Swedish Radio • Convention Chair Peter Swarte Program: • AES Awards Presentation Panelists: Axel Holzinger, D.A.V.I.D. GmbH, Munich, • Introduction of Keynote Speaker by Germany Convention Chair Peter Swarte Erik Lundbeck, Swedish Television Stockholm, • Keynote Address by Tammo Stockholm, Sweden Houtgast Heinz-Peter Reykers, WDR Cologne, Cologne, Awards Presentation Germany Please join us as the AES presents special awards to Mark Yonge, AES Standards, UK those who have made outstanding contributions to the

4 Audio Engineering Society 124th Convention Program, 2008 Spring Society in such areas of research, scholarship, and pub- This workshop investigates the impact of interactivity in lications, as well as other accomplishments that have audio environments on improving the feeling of presence contributed to the enhancement of our industry. within those environments. This workshop extends a Keynote Speaker series of related workshops on perception and interactivi- This year’s Keynote Speaker is Tammo Houtgast. Hout- ty presented at recent AES Conventions. The workshop gast was trained as a physical engineer and worked at considers the technical, creative, and psychoacoustic the Human Factors Laboratory of the Dutch Organization constraints when building interactive environments. A for Applied Physical Research. In 1974 he received his selection of panelists, with varying perspectives on the PhD with a thesis on “Lateral Suppression in Hearing.” In meaning of interactive audio, will discuss the importance the field of speech communication, he was involved in of presence within their own application areas. These the development of the STI (Speech Transmission include gaming, training, and education. The workshop Index), i.e. modeling and predicting the effect of any has an open format and encourages discussion between speech transmission channel on speech intelligibility. panelists and the audience relating to the meaning of Since 1994 he has been affiliated with the ENT depart- interactivity and presence. The workshop is therefore ment of the VU University Medical Center in Amsterdam, suitable to experts and newcomers alike and offers a as professor of experimental audiology, working on chance to share different perspectives of this topic of speech reception by hearing impaired persons. His main growing importance. interests are the relation between types of hearing impairment and the reduced intelligibility of speech in Workshop 5 Saturday, May 17 noise; signal-processing strategies for hearing aids to 13:30 – 15:30 Room L (partly) compensate for that impairment. He retired in 2006, but continues the scientific coordination of an SPATIAL AUDIO AND SOUND DESIGN IN SPORTS extensive EU-funded project HearCom, aimed at improving the participation of hearing impaired persons in our Chair: Gerhard Stoll, IRT, Munich, Germany modern communication society. Houtgast’s keynote address is entitled: “Speech Panelists: Dennis Baxter, Sound for the Olympics Reception in Noise: a Review from a Personal Perspective.” Andrea Borgato, Dolby It will be described how the concept of the Articulation Hans Huber, IRT Index, as developed over fifty years ago by French and Ales Koman, Slovenia TV Steinberg and others, led to the present models for Sports productions are a great platform for the combina- quantifying the effect of noise on the intelligibility of tion of HDTV and multichannel audio, i.e. the combina- speech. Among these models, the Speech Transmission tion of “best picture and best sound.” Since audio has a Index (STI) is of some importance since, besides the ef- very high impact on the emotions of the viewer, in partic- fect of interfering noise, it also accounts for the effect of ular in the case of very good pictures, more and more room acoustics (reverberation). In addition to its use for attention should be paid to a stimulating sound design, design purposes, the STI has proven a valuable tool for which conveys a major part of the tension in a sport assessing the quality of speech transmission in actual event. The forthcoming Euro 2008, the Beijing 2008 situations. Besides the properties of the sound transmis- Summer Olympics, and the Vancouver 2010 Winter sion path from the source to the listener, also the hearing Olympics will be important examples for bringing exciting characteristics of the listener are to be considered. Many sound together with crisp pictures to the home. The hearing impaired and older persons experience great dif- applied techniques range from special microphones and ficulties in understanding speech in noisy situations. The according placement to the use of additional sound present insight in this field will be briefly reviewed. effects with samplers, etc., as well as the specific issues of crowd pickup in surround sound. The panelists will Live Sound Seminar 2 Saturday, May 17 explain and discuss their ideas to provide new tools for 12:15 – 14:30 Forum the production and design of exciting sound in sports.

LINE ARRAYS Tutorial 3 Saturday, May 17 13:30 – 15:00 Room N101 Presenter: Ralf Zuleeg, d & b This is a session about the principles of line arrays and ABSORBERS AND DIFFUSERS describes in a non-mathematical way their behavior. This unique style of explaining the physics makes it easy to Presenter: Trevor Cox, University of Salford, Salford, adapt the theory to the real world. Demonstrations will Greater Manchester, UK round this session up. Absorbers and diffusers are two of the main design tools for altering the acoustic conditions of rooms such Saturday, May 17 13:00 Room H as studios and auditoria, semi-enclosed spaces such Standards Committee Meeting on SC-05-02 Audio as stadia and the outdoor environment. Absorbers Connectors also have a crucial role within loudspeaker enclosures and in the reduction of machinery noise. This tutorial will describe state-of-the-art designs for diffusers and Workshop 4 Saturday, May 17 absorbers, as well as touching on techniques for 13:30 – 15:30 Room O measuring and modeling these treatments. Surface diffusion is a relatively young subject area, and case MULTISENSORY INTERACTION: DOES studies will be drawn upon to show when are where INTERACTION IMPROVE PRESENCE? diffusers can be used to improve conditions. Absorp- tion is a more established technology, and yet new materials and techniques continue to be developed Chair: Renato S. Pellegrini, sonic emotion ag, today, for instance driven by the need for more sus- Oberglatt, Switzerland tainable buildings. ¯

Audio Engineering Society 124th Convention Program, 2008 Spring 5 Technical Program Session P3 Saturday, May 17 model, and the assumption of small listener 14:00 – 18:00 Room C-D movements, the source image becomes unstable at a relatively low frequency. Given that for SPATIAL AUDIO PERCEPTION AND PROCESSING, such low frequencies the spherical head model PART 1 is a reasonable approximation of measured HRTFs, this work suggests that individualized Chair: Francis Rumsey, University of Surrey, HRTF and pinnae functions are of little benefit Guildford, Surrey, UK when designing a virtualizer system that allows for some listener mobility. 14:00 Convention Paper 7327 P3-1 Objective and Subjective Evaluation of Urban 15:30 Acoustic Modeling and Auralization—Yuliya P3-4 On the Use of Directional Loudspeakers to Smyrnova, Yan Meng, Jian Kang, University of Create a Sound Source Close to the Listener Sheffield, Western Bank, Sheffield, UK —Aki Härmä, Steven van de Par, Werner This paper presents the results of objective and de Bruijn, Philips Research Laboratories, subjective evaluation of a simulation and auraliza- Eindhoven, The Netherlands tion system based on model CRR—combined It is sometimes desired to create an illusion that a ray-tracing and radiosity. Auralization of an sound source appears closer to the listener than urban square has been carried out with various the nearest loudspeaker location. By using highly boundary reflection patterns (purely specular, directional loudspeakers one may manipulate the purely diffuse, and a mix of specular and diffuse) relation between direct and reverberant energy using two audio stimuli. The subjective evaluation and therefore change the distance cues to make results reveal a strong impact of sound sources the sound source appear very close to the listener. and reflection pattern. Despite similarities in In this paper we present a method combining high- objective measures, there are noticeable differ- ly directional sound with surround audio reproduc- ences in subjective attributes between signals tion to produce controllable distance effects based on simulated and measured impulse between the listener location and the nearest responses, but current auralization algorithms loudspeakers. are still adequate in simulating real urban Convention Paper 7328 environments. Convention Paper 7325 16:00

14:30 P3-5 Directional Analysis of Sound Field with Linear Microphone Array and Applications P3-2 Virtual vs. Actual Multichannel Acoustical in Sound Reproduction—Jukka Ahonen,1 Recording—Gavin Kearney,1 Jeff Levison2 Ville Pulkki,1 Fabian Küch,2 Markus Kallinger,2 1Trinity College, Dublin, Ireland Richard Schultz-Amling2 2Euphonix, Inc., Palo Alto, CA, USA 1Helsinki University of Technology, Espoo, Finland 2Fraunhofer Institute for Integrated Circuits IIS, We present a comparison of live recordings of a Erlangen, Germany choral ensemble versus dry recordings of the same players, with the acoustic environment The use of a linear microphone array composed reconstructed from impulse responses of the of two closely spaced omnidirectional micro- original reverberant performance space. Binaur- phones as input to teleconference application of al measurements are used to objectively classify Directional Audio Coding (DirAC) is presented. the recordings, and the perceptual attributes are DirAC is a method for spatial sound processing, investigated through a series of subjective listen- where the direction of the arrival of sound and dif- ing tests. It is shown that the differences fuseness are analyzed and used for different pur- between dry recordings convolved with linear poses in reproduction. Two-dimensional plane ar- time-invariant (LTI) impulse responses and actu- rays have been used so far to generate input al acoustical recordings can be perceived by a signals for DirAC, in which case it is possible to panel of expert listeners. measure directly a two-dimensional sound field. In Convention Paper 7326 this paper a one-dimensional linear array is used to provide input signals for one-dimensional direc- 15:00 tion and diffuseness analysis in DirAC. Listening tests are conducted to evaluate the intelligibility of P3-3 Virtual Sources and Moving Targets—Glenn speech with simultaneous talkers when the linear Dickins, David Cooper, David McGrath, Dolby array is used in teleconference applications. Laboratories, Sydney, NSW, Australia Convention Paper 7329

This paper presents an analysis of the effects of 16:30 listener mobility on the stability of virtual source images created by a pair of loudspeakers. A P3-6 The SoundScape Renderer: A Unified Spatial spherical head is used to generate analytic head Audio Reproduction Framework for Arbitrary related transfer functions from which we create a Rendering Methods—Matthias Geier, Jens simple perceptual localization model for the for- Ahrens, Sascha Spors, Technische Universität ward half of the horizontal plane. This model is Berlin, Berlin, Germany then used to investigate changes in perceived source localization as the listener moves. The The SoundScape Renderer is a versatile soft- analysis demonstrates that even with this simple ware framework for real-time spatial audio ren-

6 Audio Engineering Society 124th Convention Program, 2008 Spring dering. The modular system architecture allows Session P4 Saturday, May 17 the use of arbitrary rendering methods. Three 14:00 – 18:00 Room E-F rendering modules are currently implemented: Wave Field Synthesis, Vector Base Amplitude LOW BIT-RATE AUDIO CODING Panning, and Binaural Rendering. After a description of the software architecture, the Chair: Karlheinz Brandenburg, Fraunhofer Institute implementation of the available rendering meth- for Digital Media Technology, Ilmenau, Germany ods is explained and the graphical user interface is shown as well as the network interface for the remote control of the virtual audio scene. Finally, 14:00 the Audio Scene Description Format, a system- P4-1 Time-Varying Transform for High Quality independent storage file format, is briefly Audio Communication Codecs—Pierrick presented. Philippe,1 David Virette,2 Balázs Kövesi2 Convention Paper 7330 1France Télécom R&D, Cesson Sevigne, France 2France Télécom R&D, Lannion, France 17:00 High quality audio communication is a current P3-7 Initial Investigation of Signal Capture challenge addressed by the standardization Techniques for Objective Measurement committees. In this context, ITU and MPEG of Spatial Impression Considering Head recently issued standards for high quality coding Movement—Chungeun Kim, Russell Mason, of both speech and music contents. Transform Tim Brookes, University of Surrey, Guildford, coding is used and allows quality commensurate Surrey, UK with bit rates regardless of the audio content. Up In a previous study it was discovered that listen- to now, only constant transform sizes were used ers normally make head movements attempting in these coding schemes since time varying to evaluate source width and envelopment as transform needed look-ahead for perfect recon- well as source location. To accommodate this struction, hence adding further delay. In this finding in the development of an objective mea- paper we demonstrate how variable transform surement model for spatial impression, two cap- sizes can be used without affecting the coding turing models were introduced and designed in delay. Based on filter bank theory, a framework this research, based on binaural technique: 1) avoiding look-ahead is presented. The quality rotating Head And Torso Simulator (HATS), and improvement offered by the proposed solution is 2) a sphere with multiple microphones. As an ini- illustrated in the context of MPEG4 enhanced tial study, measurements of interaural time dif- low delay AAC. ference, level difference and cross-correlation Convention Paper 7333 made with the HATS were compared with those made with a sphere containing two micro- 14:30 phones. The magnitude of the differences was P4-2 Differential Graph-Based Coding of Spikes in judged in a perceptually relevant manner by a Biologically-Inspired Universal Audio Coder comparing them with the just-noticeable differ- — Ramin Pichevar, Hossein Najaf-Zadeh, Louis ences of these parameters. Thibault, Hassan Lahdili, Communications Convention Paper 7331 Research Centre, Ottawa, Ontario, Canada In a previous work we showed that it is possible to 17:30 code audio materials using a biologically- inspired universal audio coder based on matching P3-8 A Second Order Differential Microphone pursuit. The best atoms/kernels chosen by match- Technique for Spatially Encoding Virtual ing pursuit are represented by spikes to reflect the Room Acoustics— Alexander Southern, biologically-inspired nature of the algorithm. In Damian Murphy, University of York, Heslington, that work, each spike or atom was defined by York, UK parameters such as timing, channel frequency, Room acoustics modeling using a numerical amplitude, chirp factor, etc., that were encoded simulation technique known as the Digital Wave- independently. However, encoding each guide Mesh (DWM) has previously been pre- atom/spike as a separate entity is very bit con- sented as a suitable method for measuring suming. In the present paper, we propose algo- spatial Room Impulse Responses (RIR) of virtual rithms to encode only the difference between enclosed spaces. In this paper a new method for parameters associated with spikes. Hence, we capturing the DWM modeled soundfield using an assume that each spike/atom is a node in a graph array of spatially distributed pressure-sensitive and choose the sequence of spikes that will mini- receivers is presented. The polar response of mize the differential encoding costs. Methods the formed 2nd order virtual microphone is mea- based on minimum spanning tree and traveling sured and compared to the theoretical polar salesman are proposed and compared for the response. This approach is proven to be capable graph-based optimization of the code. of decomposing the modeled soundfield into Convention Paper 7334 second order spherical harmonic components that are typically associated with 2nd order Ambisonics. Convention Paper 7332 ¯

Audio Engineering Society 124th Convention Program, 2008 Spring 7 Technical Program 15:00 Tele- and video conferencing systems for modern business communication are managed by P4-3 Unraveling the Relationship between Basic central hubs, so-called multipoint control units Audio Quality and Fidelity Attributes in Low (MCU). One major task of these units is the mix- Bit-Rate Multichannel Audio Codecs—Paulo ing of audio streams from the participating sites. Marins, Francis Rumsey, Slawomir Zielinski, This is traditionally done by decoding the University of Surrey, Guildford, Surrey, UK streams, mixing in time domain, and then Prior to this study the evaluation of multichannel re-encoding of the mixed signals. This requires audio codecs has been done mainly according to additional processing power, leads to increased the ITU-R standards BS.1116 and BS.1534. Basic delay, and degraded audio quality. The paper audio quality is the only perceptual attribute demonstrates how the recently standardized assessed in the majority of these tests. This MPEG-4 Enhanced Low Delay AAC (AAC-ELD) approach, although efficient for measuring the codec offers a solution to these problems by effi- overall quality of several codecs at once, does not cient and delayless mixing in the transform provide reasons why a particular codec is rated as domain of the codec. better as or worse than another. In this paper Convention Paper 7337 fidelity attributes were included; these were based on the attributes suggested in the ITU-R standards 16:30 but have not been used explicitly in codec evaluation up to now. In this experiment the perceptual P4-6 Low-Power MPEG-4 HE-AAC Version-2 importance of these attributes and their contribu- Encoder— Chi-Min Liu, Han-Wen Hsu, Chung- tion to the basic audio quality of low bit-rate Han Yang, Wen-Chieh Lee, National Chiao Tung surround sound codecs were investigated. University, Hsinchu, Taiwan Convention Paper 7335 In MPEG-4 HE-AAC version-2 encoder, the 15:30 analysis/synthesis complex-exponential modula- P4-4 A New Perceptual Model for Audio Coding tion filter banks are used in spectral band repli- Based on Spectro-Temporal Masking— cation (SBR) and parametric stereo (PS) coding. Steven van de Par,1 Jeroen Koppens,2 Due to the aliasing interference, the complex Armin Kohlrausch,1,3 Werner Oomen2 banks instead of real banks are adopted in the 1Philips Research Europe, Eindhoven, SBR and PS coding. However, the additional The Netherlands; overhead from the complex values in the 2Philips Applied Technologies, Eindhoven, CEMFB and the subsequent processing have The Netherlands; led to high operational overhead. Our previous 3Eindhoven University of Technology, work has designed the SBR encoders based on Eindhoven, The Netherlands; the real-domain cosine modulation filter banks; we proposed a complexification-based approach In psychoacoustics, considerable advances for the SBR coding. This paper extends the work have been made recently in developing into PS coding. An approximate method for pa- computational models that can predict the dis- rameters estimation is proposed to save opera- criminability of two sounds taking into account tional overhead with only one CEMFB-analysis spectro-temporal masking effects. These models channel. Also, a phase-adjustment down-mixing operate as artificial observers by making predic- method is proposed to reduce energy vanish tions about the discriminability of arbitrary sig- effects. nals [e.g., Dau et al., J. Acoust. Soc. Am. 99, Convention Paper 7338 Vol. 36(15), 1996]. Therefore, such models can be applied in the context of a perceptual audio coder. A drawback, however, is the computation- 17:00 al complexity of such advanced models, espe- P4-7 Low Complexity Bit Allocation Algorithms cially because the model needs to evaluate each for MP3/AAC Encoding—S Nithin,1 quantization option separately. In this paper a Kumaraswamy Suresh,2 T. V. Sreenivas2 model is introduced and evaluated that is a com- 1National Institute of Technology, Surathkal, India putationally lighter version of the Dau model but 2Indian Institute of Science, Bangalore, India maintains its essential spectro-temporal masking predictions. Listening test results in a transform We have developed two reduced complexity bit- coder setting show that the proposed model out- allocation algorithms for MP3/AAC based audio performs a conventional purely spectral masking encoding, which can be useful at low bit-rates. model and the original model proposed by Dau. One algorithm derives optimum bit-allocation Convention Paper 7336 using constrained optimization of weighted noise-to-mask ratio and the second algorithm 16:00 uses decoupled iterations for distortion control P4-5 Delayless Mixing—On the Benefits of MPEG- and rate control, with convergence criteria. 4 AAC-ELD in High Quality Communication MUSHRA based evaluation indicated that the Systems—Markus Schnell,1 Markus Schmidt,1 new algorithm would be comparable to AAC but Per Ekstrand,2 Tobias Albert,1 Daniel Przioda,1 requiring only about 1/10th the complexity. Manfred Lutzky,1 Ralf Geiger,1 Vesa Ruoppila,3 Convention Paper 7339 Fredrik Henn,2 Erlend Tårnes4 [Paper presented by TV Sreenivas] 1Fraunhofer IIS, Erlangen, Germany 2Dolby Sweden, Stockholm, Sweden 3Dolby Germany, Nuremberg, Germany 4Tandberg, Oslo, Norway

8 Audio Engineering Society 124th Convention Program, 2008 Spring 17:30 More and more functions are installed to a mobile phone, but the size of the handset has P4-8 Linear Filtering in MDCT Domain— become smaller. Devices installed in the mobile Kumaraswamy Suresh, T. V. Sreenivas, Indian phone have been required to be downsized or Institute of Science, Bangalore, India thinner. Micro-loudspeakers installed to the In this paper expressions for convolution multipli- mobile phone are required to be thinner. They cation properties of MDCT are derived starting are required to become both thinner and repro- from equivalent DFT representations. Using duce high quality sound. However, it has been these expressions, methods for implementing very difficult to make thinner micro-loudspeakers linear filtering through block convolution in the without deteriorating the acoustic performance MDCT domain are presented. The implementa- because the structure of conventional dynamic tion is exact for symmetric filters and approxi- micro-loudspeaker is not suitable to make it thin- mate for non-symmetric filters in the case of rec- ner. We have succeeded in developing an ultra- tangular window-based MDCT. For a general thin micro-loudspeaker using Oblique Magnetic MDCT window function, the filtering is done on Circuit, which is 1.5 mm thick (45% thinner than the windowed segments and hence the convolu- conventional dynamic micro-loudspeakers) with- tion is approximate for symmetric as well as non- out deteriorating the acoustic performance. symmetric filters. This approximation error is Convention Paper 7342 shown to be perceptually insignificant for symmetric impulse response filters. Moreover, the in- 14:00 herent 50% overlap between adjacent frames P5-3 A Novel Glass Laminated Structure for Flat used in MDCT computation does reduce this ap- 1 proximation error similar to smoothing of other Panel Loudspeakers—Olivier Mal, Marek Novotny,1 Bart Verbeeren,1 Neil Harris2 block processing errors. The presented tech- 1 niques are useful for compressed domain pro- AGC Research & Development Centre, Jumet, Belgium cessing of audio signals. 2 Convention Paper 7340 New Transducers Ltd. (NXT), Cambourne, UK [Paper presented by TV Sreenivas] A new, patented “sandwich structure” has been developed for various audio applications, in which thin glass sheets are laminated with a special PVB (Polyvinyl Butyral) film to eliminate Session P5 Saturday, May 17 typical acoustical weaknesses of monolithic 14:00 – 15:30 Topaz Lounge glass and standard laminated solutions. The glass improvements include suppression of ring- POSTERS: MICROPHONES AND LOUDSPEAKERS ing of the audio signal and a much more flexible and lightweight glass structure. It results in flatter 14:00 frequency response (both on-axis and 180° power response) and better transport of vibrations in P5-1 A Study of Electrostatic Forces in Single- the glass surface. In addition, better acoustical Acting Condenser Digital Transducer—Libor sensitivity and mechanical resistance are Husník, Czech Technical University in Prague, achieved. In this paper, after defining the struc- Prague, Czech Republic ture of the developed laminated glass solution, we compare its performances to previously tried One of the possibilities to design a transducer monolithic and laminated glass solutions. We with the direct digital-to-analog conversion, also emphasize the key factors influencing the sometimes called a digital loudspeaker, is the final acoustical properties. Finally, we introduce miniature condenser transducer manufactured potential application fields for the developed on a silicon chip. Only recently has this micro structure. technology been made available commercially, Convention Paper 7343 which can open further application possibilities. This paper is aimed at the study in which the back electrode of the electrostatic transducer is 14:00 partitioned into sections having total areas pro- P5-4 A Digitally Direct Driven Dynamic-Type portional to powers of 2. Since electrostatic force Loudspeaker—Ryota Saito, Akira Yasuda, acting on the membrane is affected by the distri- Kazushige Kuroki, Tomohiro Tsuchiya, Naoto bution of bit groups, which cannot be even, said Shinkawa, Hosei University, Koganei, Tokyo, electrostatic force will not be a linear function of Japan the signal voltage. Correcting coefficients for some arrangements are searched for. If a loudspeaker can be driven digitally, all Convention Paper 7341 processes from the input to the output can be performed digitally without the use of analog components such as power amplifiers; and a 14:00 small, light, and high-quality speaker system can be realized. In this paper we propose the P5-2 Ultra-Thin Micro-Loudspeaker Using Oblique basic principle behind Digital-Speaker, and Magnetic Circuit—Toshiyuki Matsumura,1 a digitally driven dynamic-type loudspeaker Shuji Saiki,1 Sawako Usuki,1 Koji Sano2 provided with multiple voice coils employing 1Matsushita Electric Industrial Co., Ltd., Kadoma multibit delta-sigma modulation. The piezoelec- City, Osaka, Japan tric-speaker used in our previous study is 2Panasonic Electronic Devices Co., Ltd., replaced by the voice coil. The prototype is Matsusaka City, Japan implemented along with a FPGA, CMOS dri- ¯

Audio Engineering Society 124th Convention Program, 2008 Spring 9 Technical Program vers, and a dynamic-type loudspeaker. The Kunimaro Tanaka, Teikyo Heisei University, THD and SPL are approximately 0.1% and 104 Tokyo, Japan dB, respectively, and the output power is 1 W Junichi Yoshio, Pioneer Corporation, Tokyo, even when the power supply voltage is 1 V. Japan Convention Paper 7344 Various technologies in terms of storage are currently used for audio recording, mastering, and archiving. Necessary 14:00 data rate is also increasing due to the complexity of audio P5-5 Accelerated Power Test Analysis Based on production —such as multichannel sound production. This Loudspeaker Life Distribution—Xu Wang, workshop reviews the current situation of technologies re- Yong Shen, Zhicheng Wu, Nanjing University, garding the storage for audio recording, mastering, and Nanjing, China archiving and will discuss the future of storage and archives. Panels from both media industries and production will dis- For the loudspeaker manufacturers, the long cuss the requirements for the next-generation storage and time spent on power tests made by relative stan- next-generation audio recording and archiving systems. dards or buyers has deeply influenced the period of product design and development. The authors Historical Event apply the theory of reliability to cut the duration SHANNON, BEETHOVEN, AND THE COMPACT DISC of loudspeaker power tests. On the basis of Saturday, May 17, 14:00 – 15:00 experiment data, a model of loudspeaker life dis- Room N tribution is propounded, from which an accelerated factor of the loudspeaker power test is derived, and then the characteristics of the loud- Presenter: Kees Schouhamer Immink speaker under normal working conditions can be estimated. The method can be conveniently used An audio compact disc (CD) holds up to 74 minutes, 33 on relative power tests and shorten the duration seconds of sound, just enough for a complete mono of the tests effectively. recording of Ludwig van Beethoven’s Ninth Symphony Convention Paper 7345 (“Alle Menschen werden Brüder”) at probably the slowest pace it has ever been played, during the Bayreuther 14:00 Festspiele in 1951 and conducted by Wilhelm Furtwän- gler. Each second of music requires about 1.5 million P5-6 Perception and Physical Behavior of bits, which are represented as tiny pits and lands ranging Loudspeaker Nonlinearities at Bass from 0.9 to 3.3 micrometers in length. More than 19 bil- Frequencies in Closed vs. Reflex Enclosures lion channel bits are recorded as a spiral track of alter- —Jukka Rauhala, Jukka Ahonen, Miikka nating pits and lands over a distance of 5.38 kilometers Tikander, Matti Karjalainen, Helsinki University (3.34 miles), which are scanned at walking speed, 4.27 of Technology, Espoo, Finland km per hour. This year it is 25 years ago that Philips and Sony introduced the CD. We will discuss the various cru- This paper examines loudspeaker nonlinearities cial technical decisions made that would determine the at bass frequencies in closed and reflex enclo- technical success or failure of the new medium. sures using signal analysis and perceptual evaluation methods. The nonlinearities are investigated by driving the loudspeakers to be Special Event compared with sinusoidal and musical test WAVE FIELD SYNTHESIS DEMONSTRATION tones. The produced responses are evaluated in Saturday, May 17, 14:00 – 17:00 terms of diaphragm displacement, harmonic dis- Sunday, May 18, 10:00 – 17:00 tortion, and bandwise distortion. In addition, a Monday, May 19, 10:00 – 17:00 listening experiment is conducted in order to Tuesday, May 20, 10:00 – 15:00 determine how the nonlinearities are perceived Room G in both reflex and closed enclosures. The results In 1988, Guus Berkhout published his paper “A Holo- show that with signals that have energy close to graphic Approach to Acoustic Control” [J. Audio Eng. the tuning frequency of the reflex port produce Soc. 36, pp. 977 - 995] where he introduced Wave Field more distortion with the closed enclosure. On Synthesis as a new concept for sound reproduction with- the other hand, acoustic bass test tone behaved out “sweet spot” limitations. Now, 20 years later, WFS is in an opposite way causing more distortion with recognized as a favorite technique for spatial sound the reflex enclosure. These phenomena were reproduction with a high potential of applications. verified with the listening tests. Research is done at many institutes to match the percep- Convention Paper 7346 tion of critical listeners with the technical state-of-the-art. The production of content dedicated to WFS perfor- Workshop 6 Saturday, May 17 mance is slowly but steadily increasing. 14:00 – 16:00 Room B In this special event, “20 Years WFS,” a series of examples of such content will be demonstrated. Old TRENDS OF STORAGE TECHNOLOGIES FOR AUDIO friends that took part in the development of WFS from RECORDING AND MASTERING the beginning as well as later converts will cooperate in a program where traditional and electronic music, acoustic Chair: Kimio Hamasaki, NHK Science and landscapes, and movie fragments illustrate the possibili- Technical Research Laboratories, Tokyo, Japan ties of Wave Field Synthesis. The demos are realized with the portable WFS system of the organization for Panelists: Noboru Harada, NTT Communication spatial rendering of electronic music “The Game of Life.” Science Laboratories, Tokyo, Japan Moreover, this special event will be a platform where Koichi Kitamura, Recording Industry people can discuss past, present, and—most impor- Association of Japan, Tokyo, Japan tant—the future of Wave Field Synthesis.

10 Audio Engineering Society 124th Convention Program, 2008 Spring Saturday, May 17 14:00 Room K wideband voice communication. Since SBC was Technical Committee Meeting on High Resolution first designed with audio compression, it does Audio not incorporate the features that speech coders commonly use. The use of voice activity detec- Live Sound Seminar 3 Saturday, May 17 tion and comfort noise generation to reduce 15:00 – 18:00 Forum bandwidth usage and power consumption is an example. In this paper we investigated exten- MIXING CONSOLES sions for SBC that would make it better suited for voice compression in the Bluetooth frame- Presenter: Ruben van der Goor work. The proposed enhancements were evaluated on the basis of their impact on voice quality, The digital mixing console as the center of a PA audio their implementation requirements, and their system. Under this headline theoretical and practical bandwidth power savings. aspects of system EQs, room alignment, sound creation, Convention Paper 7347 signal processing, and applications of VCM technology will be explained as well as analog and digital pros and 16:00 cons with live demonstrations. P6-2 Efficiently Shuffling Large Sets of Clips— Saturday, May 17 15:00 Room H Ulrich Herrmann, austriamicrosystems, Graz, Standards Committee Meeting on SC-04-03 Austria Loudspeker Modeling and Measurement A method for randomly shuffling through large sets of video or audio clips is presented in this Saturday, May 17 15:00 Room K paper. Many up-to-date devices have only a Technical Committee Meeting on Human Factors in rather limited capability of shuffling only up to Audio Systems 200 or 256 songs. This algorithm presents a way of shuffling even large sets with an almost unlim- Tutorial 4 Saturday, May 17 ited number of items. It also provides the ability 15:30 – 18:00 Room N101 to traverse back and forth with little processing power on today’s micro controllers. All this is INTELLIGENT AUDIO SYSTEMS: A REVIEW OF THE done with few bytes of code and almost no FOUNDATIONS AND APPLICATIONS OF SEMANTIC RAM. AUDIO ANALYSIS AND MUSIC INFORMATION Convention Paper 7348 RETRIEVAL 16:00 Presenters: Perfecto Herrera-Boyer, Universitat P6-3 Hardware/Software Co-Design of Multi- Pompeu Fabra Format Audio Decoder—ChangYong Son, Jay LeBoeuf, Imagine Research KangEun Lee, DoHyung Kim, Soojung Ryu, This tutorial will target students, researchers, and indus- Shihwa Lee, Samsung Advanced Institute of try audio engineers who are unfamiliar with the field of Technology, Suwon, Korea Music Information Retrieval (MIR). We will demonstrate This paper presents a hardware/software the myriad of exciting technologies enabled by the fusion co-design method for the implementation of a of basic signal processing techniques with machine multi-format audio decoder with ultra low power, learning. The presentation will be a high-level, applied, small chip size, and high flexibility, which are the multimedia-rich, overview of the building blocks of MIR most critical factors in embedded devices. This systems. Our goal is to make highly-interdisciplinary approach can provide both flexibility and low technologies and dauntingly-complex algorithms power with high performance in such a way that approachable. In the spirit of modern cooking shows, we hardware implementation has been focused on will perform numerous demonstrations, including on-the- the commonly used critical blocks of multiple fly coding of basic "intelligent audio" systems, prepared audio decoders having intensive computations. system demonstrations, and prepared audio examples Hardware blocks are well modularized to allow demonstrating more complex systems. easy and rapid architecture exploration of several digital audio standards. The proposed system Session P6 Saturday, May 17 can decode an MP3 bitstream using only about 16:00 – 17:30 Topaz Lounge 4 MHz clock frequency and AAC bitstream using only about 7 MHz clock frequency on average at POSTERS: MOBILE PHONE AUDIO the sampling rate of 48 kHz and the target bit rate of 128 kbps/stereo. Convention Paper 7349 16:00 P6-1 Enhancements to the SBC CODEC for Voice 16:00 Communication in Bluetooth Devices— P6-4 Audio Enhancement for Portable Device- 1 2 Laurent Pilati, Mohammad Zad-Issa Based Speech Applications—Rory Turnbull, 1 Broadcom Corp., Sophia Antipolis, France Peter Hughes, Steve Hoare, BT Group CTO, 2 Broadcom Corp., Irvine, CA, USA Ipswich, Suffolk, UK The Bluetooth Audio Distribution profile has uses Portable devices with audio capabilities necessi- low complexity sub-band coder (SBC) as its tate the use of small transducers, often with poor mandatory audio compression codec. More frequency responses. This can be a limiting factor recently, SBC has been selected for Bluetooth in the perception of the speech quality of VoIP ¯

Audio Engineering Society 124th Convention Program, 2008 Spring 11 Technical Program services hosted on such a device. This paper Workshop 7 Saturday, May 17 seeks to investigate the problem and provide prac- 16:00 – 18:00 Room B tical solutions through the use of appropriate enhancement technologies. The paper covers the LOUDNESS—A CHANGE OF PARADIGMS? use of equalization, dynamic range compression, and psychoacoustic bass enhancement as possi- Chair: Thomas Lund, TC Electronic A/S ble methods for improving intelligibility. Subjective tests are used to evaluate the enhancements prior Panelists: Andrew Mason, BBC R&D to making practical recommendations. Jean-Paul Moerman, VRT – Flemish Convention Paper 7350 Broadcasting [Paper was presented by Peter Hughes] Gerhard Spikofski, IRT Munich Alessandro Travaglini, Fox Italia 16:00 Canada and Fox TV in Italy have proven that the switch P6-5 An Efficient, Low-Noise Filter Architecture for to loudness measurement instead of peak measurement Bass Processing on a Processor Core—Peter is a possible route—and one which serves to solve the Eastty, Nathan Bentall, Duncan Stott, Oxford most frequent complaint by consumers. Are we at the Digital Limited, Stonesfield, Oxfordshire, UK beginning of a new practice in audio levels in Broadcast- ing? Standards are also in place (ITU-R BS.1770), Bass Enhancement is becoming popular in metering tools incorporating it appear, etc. The panel will many forms of consumer devices. Whatever discuss the state of things, the transition workflow, the technique is used on whatever processor, the tools available, different possible solutions, and practical low frequency filtering involved is frequently the experiences, also with different methods of measurement. major determinant of system signal-to-noise ratio. The architecture described combines an Historical Event efficient, cascaded, low-pass FIR filter and a ELECTRO ACOUSTIC SLIDE RULE poly-phase adaptation of standard low frequency Saturday, May 17, 16:00 – 17:00 IIR filtering. The resulting circuit achieves a 20 to Room N 30 dB improvement in signal to noise ratio at the cost of only 12 instructions per sample. The Presenter: Peter Swarte technique may be applied to any bass process- In the transition of the seventies to the eighties in the ing using fixed or floating point processors. previous century, the programmable calculators could be Complete design tables for the cascaded FIR fil- applied for the design of a sound reinforcement system. ters are given as are noise spectrum plots of the The disadvantage was that you were not in the position results. to make quick changes in the parameters, e.g., the Convention Paper 7351 reverberation time of the room, the directivity factor of the sources, and the speech intelligibility based upon the development work of Peutz as he had published in a 16:00 paper during the AES convention in 1971, New York. A P6-6 Implementation of Dynamic Voltage and slide rule reduces these disadvantages in a dramatic Frequency Scaling on Portable Audio Players way. By sliding you were able to make an estimation of —Dahyanto Harliono, Woon-Seng Gan, the results (costs) and the margins you worked in. Peter Nanyang Technological University, Singapore Swarte will demonstrate some examples of real situations with the help of this slide rule. The results are still Current portable computing devices demand not valid and based upon the statistical approach of sound in only higher performance but also lower power (semi) enclosed rooms. Furthermore, the relationship consumption. For the same reason, this between several parameters like acoustic power, direc- research aims to build a framework that enables tivity, diffuse sound, direct sound, absorption, and rever- a rapid design of energy-efficient embedded sys- beration are part of the demonstration. tems. Specifically, this research is focused on a dynamic voltage scaling algorithm, which has Saturday, May 17 16:00 Room K been found effective in saving power consump- Technical Committee Meeting on Studio Practices tion. We developed a method of scaling voltage and Production and frequency dynamically on the latest embedded processor, jointly designed by Analog Devices and Intel. The rationale behind this Student Event/Career Development method is to avoid the processor being idle in OPENING AND STUDENT DELEGATE ASSEMBLY high operating frequency and voltage. Instead, MEETING – PART 1 the processor can save power by running its Saturday, May 17, 16:30 – 18:00 task at a lower frequency and voltage, and com- Room O pleting it just before the real-time deadline. Fur- Chair: Suzana Jakovic thermore, our method can also be implemented in other embedded processors with voltage- Vice Chair: Misato Yamada frequency scaling features. The first Student Delegate Assembly (SDA) meeting is Convention Paper 7352 the official opening of the convention’s student program and a great opportunity to meet with fellow students from all corners of the world. This opening meeting of the Stu- dent Delegate Assembly will introduce new events and election proceedings, announce candidates for the coming year’s election for the Europe/International Regions, announce the finalists in the recording competition cate-

12 Audio Engineering Society 124th Convention Program, 2008 Spring gories, hand out the judges’ sheets to the nonfinalists, 09:00 and announce any upcoming events of the convention. Students and student sections will be given the opportu- P7-1 Low-Frequency Extension of Gated nity to introduce themselves and their activities, in order Loudspeaker Measurements—Juha to stimulate international contacts. Backman, Nokia Corporation, Espoo, Finland All students and educators are invited to participate in The free-field response of a loudspeaker system this meeting. Election results and Recording Competition can be approximated through a gated measure- and Poster Awards will be given at the Student Delegate ment, made in a sufficiently large space. The fre- Assembly Meeting–2 on Tuesday, May 20, at 14:30. quency resolution is nominally determined by the time gap between the direct sound and the first Saturday, May 17 17:00 Room H reflection, but the actual low-frequency accuracy Standards Committee Meeting on SC-04-01 Acousics of gated measurements is reduced also by the and Sound Source Modeling group delay of the loudspeaker itself. The group delay at low frequencies may cause a large fraction of the energy sound radiation to be cut off, Saturday, May 17 17:00 Room K underestimating the low-frequency response. A Technical Committee Meeting on Signal Processing method is presented to estimate the approximate low-frequency response from the imped- Special Event ance measurement of the loudspeaker and to MIXER PARTY use the response to pre-process the acoustical Saturday, May 17, 18:00 – 19:30 measurement to improve the accuracy of the See Flyer for Details gated measurement. Convention Paper 7353 A mixer party will be held on Saturday evening to enable convention attendees to meet in a social atmosphere 09:30 after the opening day’s activities to catch up with friends and colleagues from the industry. There will be a cash P7-2 Measurement and Fourier-Bessel Analysis bar and snacks. of Loudspeaker Radiation Patterns Using a Spherical Array of Microphones—Filippo M. Fazi,1 Vincent Brunel,1 Philip A. Nelson,1 Lars Student Event/Career Development Hörchens,2 Jeongil Seo3 STUDENT PARTY 1University of Southampton, Highfield, Saturday, May 17, 20:00 –22:00 Southampton, UK De Valk 2Delft University of Technology, Delft, To welcome you to Amsterdam the Dutch student sec- The Netherlands tion has organized a party in the evening on the first day 3Electronics and Telecommunications Research of the convention (Saturday, May 17). The party will take Institute (ETRI), Daejeon, Korea place in the venue De Valk at the IJplein 3 (at the back of Loudspeakers are widely used in three-dimen- Amsterdam central station, across the water). At the par- sional sound field reconstruction systems, but ty there will be audio colleagues from all over the world, their spatial directivity features are relatively lit- drinks, and music. Music is provided by the band Dear tle-known. In this paper a hemispherical array of Watson and later in the evening DJ’s will heat up the 40 microphones was designed and built in order dance floor. Of course everybody is welcome to this par- to measure the pressure field radiated by differ- ty—you don’t have to be a student to party like a student. ent commercially available loudspeakers. The Contribution toward expenses: 5 EUROS. Tickets are spatial samples of the acoustic pressure were sold at SDA-1. processed in order to estimate the truncated The Student party is sponsored by SAE Institute. Fourier-Bessel expansion of the sound field, which allows the reconstruction of the 3-D radia- Special Event tion pattern. An analysis of the errors involved in BOAT TRIP the estimation was also performed with a numer- Saturday, May 17, 20:30 – 22:00 ical model of the array. Convention Paper 7354 Join us for the Grachten Tour boat trip through the Amster- dam canals. The tour lshows parts of Amsterdam that 10:00 cannot be seen “on land.” What better way to explore Amsterdam’s ancient city center than by going on a tour P7-3 Turbulent and Viscous Air Friction in the through the city’s canals? This is definitely an experience Mid-High Frequency Loudspeaker—Ivan not to be missed during a visit to Amsterdam. Tickets will be Djurek,1 Antonio Petosic,1 Danijel Djurek2 available at the Special Events desk. 1University of Zagreb, Zagreb, Croatia 2Alessandro Volta Applied Ceramics (AVAC), Zagreb, Croatia Session P7 Sunday, May 18 09:00 – 11:30 Room C-D Mid-high frequency loudspeaker with resonant frequency f = 982 Hz has been studied in atmos- 4 pheres of air, He , D2 and H2 at pressures rang- LOUDSPEAKERS, PART 2 ing 0-1 bar. The measurements of viscous and turbulent contributions to the friction entering Q- Chair: Ronald Aarts, Philips Research, Eindhoven, factor showed significant difference as compared The Netherlands to a low frequency loudspeaker. The resonant frequency in air is considerably lower in an evacu- ¯

Audio Engineering Society 124th Convention Program, 2008 Spring 13 Technical Program ated space than at 1 bar, and this differs from the Session P8 Sunday, May 18 low frequency loudspeaker, when the opposite is 09:00 – 12:30 Room E-F true. Measurements showed that imaginary part of viscous friction in Navier-Stokes equation is domi- WAVE FIELD SYNTHESIS nant, while contribution of the real part to the fric- Chair: Diemer de Vries, Delft University of tion term is less significant, and Navier-Stokes Technology, Delft, The Netherlands equation reduces to the Stokes form [. . .], when imaginary part of the viscous force reduces effective vibration mass, which in turn enables opera- 09:00 tion of the loudspeaker at high frequency. The P8-1 The Theory of Wave Field Synthesis Revisited data were interpreted in terms of Greenspan theo- —Sascha Spors,1 Rudolph Rabenstein,2 ry of the piston radiator. Jens Ahrens1 Convention Paper 7355 1Technische Universität Berlin, Berlin, Germany 2University of Erlangen-Nuremberg, Erlangen, 10:30 Germany P7-4 Modeling of an Electrodynamic Loudspeaker Wave field synthesis is a spatial sound field including Membrane Viscoelasticy—Antonio reproduction technique aiming at authentic Petosic,1 Ivan Djurek,1 Danijel Djurek2 reproduction of auditory scenes. Its theoretical 1University of Zagreb, Zagreb, Croatia foundation was developed almost 20 years ago 2Alessandro Volta Applied Ceramics (AVAC), and has been improved considerably since then. Zagreb, Croatia Most of the original work on wave field synthesis is restricted to the reproduction in a planar listening The model is proposed based upon viscoelastic area using linear loudspeaker arrays. Extensions properties of the loudspeaker membrane, and like arbitrarily shaped distributions of secondary properties considered include stress-strain hys- sources and three-dimensional reproduction in a teresis, creeping effect, initial stress effect, and listening volume have not been discussed in a uni- appearance of the temperature fluctuations on fied framework so far. This paper revisits the theo- the membrane surface. The creeping displace- ry of wave field synthesis and presents a unified ment response dependent on the step-like exci- theoretical framework covering arbitrarily shaped tation current has been measured on different loudspeaker arrays for two- and three-dimensional loudspeaker configurations, and listed effects reproduction. The paper additionally gives an were analyzed in terms of the N-order Ben- overview on the artifacts resulting in practical newitz-Rötger differential equation, commonly setups and briefly discusses some extensions to used for description of the system of vibrating the traditional concepts of WFS. viscoelastic body. The main parameter in this Convention Paper 7358 equation is inverse stress parameter which con- nects friction and restoring term in the loudspeaker vibrating system. 09:30 Convention Paper 7356 P8-2 A Finite Difference Time-Domain Approach to Analyzing Room Effects on Wave Field 11:00 Synthesis Reproduction—Robert Oldfield, Ian Drumm, Jos Hirst, University of Salford, Salford, P7-5 On a Novel Concept of Membrane Suspension Greater Manchester, UK in an Electrodynamic Loudspeaker—Danijel Djurek,1 Ivan Djurek,2 Antonio Petosic2 Probably the largest pit-fall to accurate audio 1Alessandro Volta Applied Ceramics (AVAC), reproduction using wave field synthesis (WFS) is Zagreb, Croatia the listening space. The WFS theory assumes 2University of Zagreb, Zagreb, Croatia free field, source free conditions that are seldom the case for practical sound reproduction. There A laboratory model of an electrodynamic loud- is consequently a need to determine what effect speaker has been realized with the membrane the reproduction room has upon the synthesized suspended on a hollow elastic torus positioned in sound field. This paper presents a finite differ- the bottom of the membrane, close to the voice ence time-domain (FDTD) approach to predicting coil. This geometry removes the torque in the the sound field in a room with arbitrary geometry membrane coming from the maximum possible and frequency dependent absorbing boundaries. distance of the suspension on the outer rim from A significant benefit to using FDTD is that the the voice coil. The suppressed torque results in WFS system can be modeled both as part of the the suppression of the Bessel vibration modes, room and also in free-field conditions; therefore which generate stochastic deformation tilts on the distortion of the sound field from the acoustics of membrane surface. Such tilts contribute to the the reproduction room can be quantified. intrinsic friction of the membrane, and their Convention Paper 7359 absence results in minor viscoelastic losses. Lat- eral rigidity of the torus is sufficient for operation of the loudspeaker without centric fixation. 10:00 Convention Paper 7357 P8-3 Wave Field Synthesis Evaluation Using the Minimum Audible Angle in a Concert Hall— Georgios Marentakis,1 Etienne Corteel,2 Stephen McAdams1 1McGill University, Montreal, Quebec, Canada 2sonic emotion ag, Oberglatt, Switzerland

14 Audio Engineering Society 124th Convention Program, 2008 Spring Localization accuracy with Wave Field Synthesis In this paper we outline a basic framework for the (WFS) was estimated in a variable-acoustics reproduction of the wave field of moving virtual concert hall. Contrary to previous studies, we sound sources. Conventional implementations employed a Minimum Audible Angle (MAA) par- usually reproduce moving virtual sources as a adigm as a measure of localization performance. sequence of stationary positions. This process The MAA was estimated for three different leads to various artifacts as reported in the litera- listening positions, three orientations of the lis- ture. On the example of wave field synthesis, we teners (0, 60, 90 degrees) and two acoustical show that the explicit consideration of the physi- conditions. WFS was found to produce satisfying cal properties of the wave field of moving sources localization cues that depend little on the rever- avoids these artifacts and allows for the accurate beration time of the room and only weakly on the reproduction of the Doppler Effect. However, position of the listener. numerical simulations suggest that the artifacts Convention Paper 7360 inherent to the reproduction system can lead to a heavy degradation of the reproduction quality. 10:30 Convention Paper 7363 P8-4 Objective and Subjective Analysis of Localization Accuracy in Wave Field 12:00 Synthesis—Joseph Sanson,1 Etienne Corteel,2 Olivier Warusfel1 P8-7 A Graphical Tool Set for Analyzing Wave 1IRCAM, Paris, France Field Synthesis Algorithms—Thomas Korn, 2sonic emotion ag, Oberglatt, Switzerland Fraunhofer Institute for Digital Media Technology, Ilmenau, Germany This paper analyses localization inaccuracies in the synthesis of virtual sound sources using Current Wave Field Synthesis (WFS) rendering Wave Field Synthesis (WFS), particularly at high realizations consist of large structures of audio frequencies. Objective and perceptual analyses signal processing components (filters, delays, are conducted through a binaural simulation of amplitude weighting) that are controlled by com- the actual sound field reproduced at the listen- plex algorithms based on the virtual source's er’s ears. The simulation consists in summing properties. This paper proposes a set of tools the respective contribution of each array trans- that is used to analyze the underlying WFS coef- ducer after filtering it with the appropriate HRTF ficient calculation algorithms visually by mapping according to the considered listener’s position. characteristic measures dependent on the High-pass filtered white noises are used as a source's and listener's position. These measures critical signal to investigate the impact of aliasing are derived from the reproduction system's ideal- on localization accuracy. Objective and percep- ized transfer function and parametric impulse tual observations show that localization accuracy response description. They reveal functional may degrade for off-centered listening positions, aspects of the algorithm’s behavior. The mea- which can be mainly attributed to a mismatch in sures aim at supporting an intuitive understand- the elicited Interaural Level Differences (ILD) ing of the perception of virtual sound events in a above the aliasing frequency. Wave Field Synthesis system, but also they Convention Paper 7361 facilitate the basic algorithm development process. Convention Paper 7364 11:00 P8-5 Wave Field Synthesis with Increased Aliasing Frequency—Etienne Corteel, Renato Pellegrini, Workshop 8 Sunday, May 18 Clemens Kuhn-Rahloff, sonic emotion ag, 09:00 – 11:00 Room L Obergltt (Zurich), Switzerland SURROUND SOUND IN RADIO Wave Field Synthesis (WFS) is a sound reproduction technique that enables the synthesis of Chair: Bosse Ternstrom, Swedish Radio target sound fields without any assumption on the listening position. Spatial aliasing is one of the remaining artifacts of WFS that limits the Panelists: Karl Petermichl, ORF Radio exact synthesis below a corner frequency Heinz-Peter Reykers, WDR Köln referred to as spatial aliasing frequency. This Gerhard Stoll, IRT – Munich, Germany paper presents a new technique that enables an Multichannel audio in radio is steadily catching up and increase of the spatial aliasing frequency of sometimes surpassing the TV realm with regards to WFS assuming a preferred listening area. The number of productions, diversity of genres, and dedica- presented technique is fully scalable and may be tion of the people involved. Although there shouldn't be a adapted to any listening zone shape or location. direct competition, the inherent greater freedom of radio Applications in the domain of simulation environ- concerning the dramaturgical creativity in genres like ments and home entertainment are discussed. radio drama or experimental and electro-acoustic music Convention Paper 7362 has led to a promising output of surround sound programs in those stations that pioneer this new format. The 11:30 panel consists of what you already can call “veterans” of surround sound in radio, and they will discuss the state P8-6 Reproduction of Moving Virtual Sound of the art, current issues, and solutions like file formats, Sources with Special Attention on the LFE, signal distribution, etc., as well as showcase some Doppler Effect—Jens Ahrens, Sascha Spors, of their respective work. Technische Universität Berlin, Berlin, Germany ¯

Audio Engineering Society 124th Convention Program, 2008 Spring 15 Technical Program Workshop 9 Sunday, May 18 We present an integrated set of audio-visual 09:00 – 11:30 Room O tracking and synthesis tools to aid matching of the audio to the video position in both horizontal UNEVEN BASS REPRODUCTION IN AUTOMOBILES and periphonic sound reinforcement systems. Compensation for screen size and loudspeaker Chair: Tom Nousaine, Listening Technology layout for high definition formats is incorporated and the spatial localization of the source is rendered using advanced spatialization techniques. Panelists: David Calristrom, Chrysler Automotive A subjective comparison of several original and David Clark, Alpine, USA enhanced film sequences using the Vector Base Robert Klacza, Chrylser Automotive Amplitude Panning (VBAP) method is presented. Alfred Svobodnik, Harman International The results show that the encoding of non-con- Mark Ziemba, Panasonic-US tradictory audio-visual spatial information, for In the evaluation of 800 OEM Autosound systems Tom presentation on different loudspeaker layouts Nousaine, Listening Technology, has found that a majori- significantly improves the naturalness of the lis- ty of systems suffer from spectral uniformity problems at tening/viewing experience. low frequencies. This simply means that bass sounds are Convention Paper 7365 uneven. For example, on a program with an acoustic bass solo, some notes almost disappear while others may 09:30 seem unusually loud. It is commonly felt that this is a sim- P9-2 Encoding Higher Order Ambisonics with ple equalization issue. Thus, it is puzzling that this prob- 1 2 lem continues to exist when modern car electronics have AAC—Erik Hellerud, Ian Burnett, Audun Solvang,1 U. Peter Svensson1 significant active sound control capabilities. The panel will 1 discuss the issue, causes, and possible solutions. Norwegian University of Science and Technology, Trondheim, Norway 2University of Wollongong, Wollongong, New Tutorial 5 Sunday, May 18 South Wales, Australia; 09:00 – 11:00 Room N101 In this paper we explore a simple method for ELECTROACOUSTIC MEASUREMENTS, PART 1 reducing the bit rate needed for transmitting and storing Higher Order Ambisonics (HOA). The Presenter: Christopher J. Struck, CJS Labs, San HOA B-format signals are simply encoded using Francisco, CA, USA Advanced Audio Coding (AAC) as if they were individual mono signals. Wave field simulations This tutorial focuses on the fundamentals of electroa- show that by allocating more bits to the lower coustic measurements, including principles of acoustics, order signals than the higher the resulting instrumentation, and data interpretation as well as practi- error is very low in the sweet spot but increases cal information on how to perform appropriate tests. The as function of distance from the center. Encod- basic components of a measuring system are examined ing the higher order signals with a low bit rate followed by a review of basic acoustics, signals, sound does not lead to a reduced audio quality. The sources, and sound fields. Psychoacoustics and hearing spatial information is improved when higher- are examined with respect to objective measurements. order channels are included, even if these are Measurement transducers are explained in terms of encoded with a low bit rate. selection and application. Finally, an overview of fre- Convention Paper 7366 quency analysis and the FFT is presented. This tutorial is intended to enable the participants to perform accurate 09:30 audio and electroacoustic tests and provide them with the necessary tools to understand and correctly interpret P9-3 Virtualized Listening Tests for Loudspeakers the results. —Timo Hiekkanen,1 Aki Mäkivirta,2 Matti Karjalainen1 1Helsinki University of Technology, Espoo, Finland Sunday, May 18 09:00 Room H 2 Standards Committee Meeting on SC-02-01 Digital Genelec Oy, Iisalmi, Finland Audio Measurement Techniques The precise location of a loudspeaker in a listening room is known to affect loudspeaker prefer- Sunday, May 18 09:00 Room K ence ratings. When multiple loudspeakers are Technical Committee Meeting on Network Audio compared the evaluation is limited by the poor Systems human auditory memory. To overcome these problems, a method to evaluate and compare Session P9 Sunday, May 18 loudspeakers using headphones is proposed. 09:30 – 11:00 Topaz Lounge The method utilizes personal head-related transfer functions in rendering the sound field recorded in a standard listening room with an artificial POSTERS: SPATIAL AUDIO PERCEPTION head. Equalization of circumaural headphones AND PROCESSING and the artificial head are investigated. Formal listening tests are conducted to examine differ- 09:30 ences between the proposed binaural method and real loudspeakers in a standard listening P9-1 Audio-Visual Processing Tools for Auditory room. Listening tests show that the virtualized Scene Synthesis—Gavin Kearney, Rozenn loudspeakers can be nearly imperceptible from Dahyot, Frank Boland, Trinity College Dublin, reality in many but not in all cases. Dublin, Ireland Convention Paper 7367

16 Audio Engineering Society 124th Convention Program, 2008 Spring 09:30 employing as little as three bits per sample. This quantization strategy seems very promising for P9-4 Binaural Rendering in MDCT Domain for Multi- reducing the rate of HOA. Object Audio Coding—Shinya Iizuka, Kei Convention Paper 7370 Kikuiri, Nobuhiko Naka, NTT DoCoMo, Inc., Yokosuka, Kanagawa, Japan 09:30 We propose a binaural rendering method in Modified Discrete Cosine Transform (MDCT) P9-7 A Binaural Auditory Model for the Evaluation of Reproduced Stereophonic Sound— domain. It has good compatibility with audio 1 2 codecs because a number of audio codecs uti- Marko Takanen, Gaëtan Lorho 1Helsinki University of Technology, Espoo, Finland lize an MDCT filter bank for time-frequency 2 transform. The proposal maps MDCT coeffi- Nokia Corporation, Helsinki, Finland cients to the real part of the Modulated Complex Binaural cues describing the differences in phase Lapped Transform (MCLT) coefficients and and power between signals at the two ears processes the amplitudes and phases according enable our auditory system to localize sound to the binaural information. The inverse MCLT is sources and segregate spatially multiple auditory applied to the coefficients with a synthesis win- events. Recent publications on binaural auditory dow function, which is derived from the perfect models have shown how the interaural coher- reconstruction condition for the phase shifted ence can be utilized to estimate these cues and signal under the assumption of linear phase therefore model the localization ability of our property. The proposed method is applicable to auditory system. This approach is exploited in the Binaural Cue Coding Type I and offers this paper to estimate the binaural cues at differ- equivalent subjective quality to the original bin- ent frequency bands and identify the spatial loca- aural signal. tion of sound sources from recorded broadband Convention Paper 7368 signals. We illustrate the application of a binaural auditory model to evaluate sound reproduced by 09:30 a stereophonic loudspeaker setup in terms of source localization and specific loudness. P9-5 Room-Dependent Preference of Virtual Convention Paper 7371 Surround Sound—Frederick Scott, Agnieszka Roginska, New York University, 09:30 New York, NY, USA P9-8 An Augmented Reality Audio Mixer and A common method for simulating surround Equalizer—Ville Riikonen, Miikka Tikander, sound over headphones, so-called virtual Matti Karjalainen, Helsinki University of surround sound, is the convolution of content Technology, Espoo, Finland information with binaural cues. Often, room information is included. This paper examines if using In Augmented Reality Audio (ARA) applications HRTFs with room impulse responses cus- the real sound environment of the user is tomized to the room the listener is in enhances extended with virtual objects. The real environ- the listening experience. Perceptual experiments ment is reproduced as a pseudo-acoustic world were conducted to evaluate whether or not lis- via a special ARA headset that consists of bin- teners prefer a room accurate rendering versus aural microphones and headphones. However, a room that is dissimilar to the one a listener is the headset causes coloration to the pseudo- seated in. A preference test was conducted acoustic representation. In order to make the using music as the test material. headset acoustically transparent, equalization is Convention Paper 7369 needed. Digital equalization easily causes unac- ceptable delays. This paper presents a novel ARA mixer with real-time analog equalization to 09:30 correct the coloration caused by the leakage P9-6 Quantization of 2-D Higher Order Ambisonics through the headset and changed resonances in Wave Fields—Audun Solvang, U. Peter the closed ear canal. Svensson, Erik Hellerud, Norwegian University Convention Paper 7372 of Science and Technology, Trondheim, Norway 09:30 The spatial distribution of the quantization noise for a 2-D Higher Order Ambisonics (HOA) signal P9-4 Sub-Band Adaptive Crosstalk Cancellation: is investigated analytically. Uniformly distributed A Novel Approach for Immersive Audio— loudspeakers radiating plane waves in a nonre- Stefania Cecchi,1 Lorenzo Palestini,1 Paolo verberant environment and frequency domain Peretti,1 Francesco Piazza,1 Ferruccio Bettarelli2 quantization are presumed. It is found that 1Universita Politecnica delle Marche, Ancona, employing the same quantization interval for all Italy orders leads to uniformly distributed quantization 2Leaff Engineering, Porto Potenza Picena (MC), noise in space. Assigning a larger quantization Italy interval (i.e., fewer bits) to higher orders leads to a radially increasing quantization noise. Match- In the field of immersive audio, crosstalk ing the quantization error to the reproduction canceller is required when a virtual sound is ren- error at the near perfect reconstruction boundary dered over two loudspeakers. In the last decade suggests that as little as four bits per sample can several adaptive algorithms have been pro- be used for quantization. Furthermore, high-pass posed: nowadays the least square (LMS) algo- filtering the HOA components opens up for rithm seems to be the best compromise ¯

Audio Engineering Society 124th Convention Program, 2008 Spring 17 Technical Program between simplicity and robustness although its Special Event convergence is weakened for colored inputs. In GEOFF EMERICK/SGT. PEPPER this paper a new approach for crosstalk cancel- Sunday, May 18, 11:00 – 12:30 lation based on a sub-band adaptive algorithm Room L will be derived. The effectiveness of this algorithm, considering colored input, will be present- Marking the 40th Anniversary of the release of Sgt. Pep- ed in terms of matrix inversion quality and fast per’s Lonely Hearts Club Band, Geoff Emerick, the Beat- convergence rate comparing it with the conven- les engineer on the original recording was commissioned tional LMS algorithm. by the BBC to re-record the entire album on the original Convention Paper 7373 vintage equipment using contemporary musicians for a unique TV program. Celebrating its own anniversary, the APRS is proud to Workshop 10 Sunday, May 18 present for a select AES audience, this unique project 09:30 – 11:00 Room B featuring recorded performances by young UK and US artists including the Kaiser Chiefs, The Fray, Travis, PROJECT MANAGEMENT IN PRODUCT Razorlight, the Stereophonics, the Magic Numbers, and DEVELOPMENT a few more—and one older Canadian, Bryan Adams. These vibrant, fresh talents recorded the original Chair: Jeroen Langevoort, NXP Semiconductors, arrangements and orchestrations of the Sgt. Pepper Nijmegen, The Netherlands album using the original microphones, desks, and hard- learned techniques directed and mixed in mono by the As car radio systems become more and more complex Beatles own engineering maestro, Geoff Emerick. proper project management is key to deliver the products Hear how it was done, how it should be done, and how on time and in full to the market. many of the new artists want to do it in the future. Geoff The development project of a car radio system with will be available to answer a few questions about the ROM-codes and IC’s, like the LeafDice and Dirana 2, will recording of each track and, of course, more general be taken as reference. In the workshop attention will be questions regarding the recording processes and the paid to the requirements definition, risk management, innovative contribution he and other Abbey Road wizards and design processes within NXP. Special requirements made to the best ever album. on quality that are common in the automotive industry APRS, The Association of Professional Recording will be briefly touched on, as these are very different Services, promotes the highest standards of profession- from those of the consumer industry. Among the deliver- alism and quality within the audio industry. Its members ables of such projects is an extensive communication are recording studios, postproduction houses, master- package for the OEM to enable them to use the system ing, replication, pressing, and duplicating facilities, and in a proper way. This includes application board and providers of education and training, as well as record application notes. producers, audio engineers, manufacturers, suppliers, and consultants. Its primary aim is to develop and Live Sound Seminar 4 Sunday, May 18 maintain excellence at all levels within the UK’s audio 09:30 – 13:00 Forum industry.

CABLES AND CONNECTORS Exhibitor Seminar 1 Sunday, May 18 Presenters: Wolfgang Feß, Klotz-a.i.s 11:00 – 12:00 Room P Norbert Nachbauer, Neutrik, Schaan, Liechtenstein SENNHEISER ELECTRONIC CORPORATION This session describes in a wide range the structures, materials, and application ranges of cables used in PA Presenters: Sebastian Schmitz systems (copper - balanced, unbalanced; fiber optic - Gregor Zielinsky single mode, multi mode, a.o.). Parameters that have MKH 800 Twin with Remote Controllable determining influence on the characteristic impedance of Directivity copper cables will be pointed out as well as an overview being shown about typical audio, video, and data appli- MKH 800 Twin The MKH 800 Twin is a Doublecapsule cations and the required cables (analog, digital, data, Microphone with remote-controllable pattern for stereo antenna, etc.). Audible effects will be demonstrated. and surround recording. Each of the capsules can be recorded separately as they have separate outputs. This Student Event/Career Development enables you to seamlessly adjust the pickup pattern at CAREER/JOB FAIR post-production stage. The presentation shows several Sunday, May 18, 09:30 – 11:30 examples of how to use this technique in stereo and sur- Topaz Lounge round examples. The Career Fair will feature several companies from the exhibit floor. All attendees of the convention, students Sunday, May 18 11:00 Room H and professionals alike, are welcome to come talk with Standards Committee Meeting on SC-04-04 representatives from the companies and find out more Microphone Measurement and Characterization about job and internship opportunities in the audio industry. Bring your resume! Sunday, May 18 11:00 Room K Sunday, May 18 10:00 Room K Technical Committee Meeting on Archiving, Technical Committee Meeting on Audio Restoration, and Digital Libraries for Telecommunications

18 Audio Engineering Society 122nd Convention Program, 2007 Spring Workshop 11 Sunday, May 18 Workshop 12 Sunday, May 18 11:30 – 13:30 Room B 12:00 – 13:00 Room O

CREATING A MUSIC DOWNLOAD WEBSITE AUTOMATED DIRECTIONAL LIVE SOUND REINFORCEMENT Chair: Vicki Melchior, Consultant, San Anselmo, CA, USA Chair: Dave Haydon, Out Board Electronics, UK

Panelists: Brian Armour, CC Technology, Glasgow, Panelists: Tom Strebel, Audiopool, Switzerland Scotland Cees Wagenaar, Peutz, The Netherlads Philip Hobbs, Linn Audio Robin Whittaker, Out Board Electronics, UK Bosse Ternstrom, Swedish Radio Karsten Wolstadt, Danish Royal Theatre, Denmark For labels, bands, and artists facing the decline of physical media, the potential now exists to release new and Directional amplification, also referred to as Source Ori- prior catalogs directly through digital downloads. Down- ented Reinforcement (SOR), describes a practical tech- loads can readily be targeted to specialized markets as nique to deliver amplified sound to a large listening area well as to the broader online community, and at the with even coverage while providing directional informa- same time may offer selective digital formats—for exam- tion to reinforce visual cues and create a realistic and ple, high resolution or lossless standard resolution, multi- noncontradictory auditory panorama. Audio demonstra- channel, surround algorithms such as ambisonics, tions of the fundamental psychoacoustic techniques surround with designated height channels, etc. In this employed in an SOR design will be presented and limits workshop a number of individuals with recent web devel- discussed. The panel of presenters will outline the histo- opment experience will describe the challenges and ry of SOR from Steinke & Fels pioneering work with their methodology of creating a music download website. The Delta Stereophony System in the mid 1970s (later discussion should provide insight for those getting start- licensed to AKG) to Out Boards current-day TiMax Audio ed in this area and will include: what is involved with set- Imaging Delay Matrix including the very latest ground ting up for web delivery from a producer’s viewpoint, breaking technology employed to enable control of overview of server hardware and network requirements, precedence by radar tracking the actors on the stage. programming and software development, website issues Descriptions of venues and productions that have and licensing, business models and expected volume, employed SOR include the new Copenhagen Royal The- choosing download formats, reaching target audiences, atre Drama House, Einsiedeln Welttheater 2007, the and economies of scale including when it’s better to join opera Aïda in Rotterdam Ahoy and Basel Tattoo among a much larger website. others. TiMax, a DSP-based audio control system, was launched in 1997 and has since received a number of industry awards for innovation. TiMax has become synony- mous with the techniques for localization of sound in the- Tutorial 6 Sunday, May 18 aters and auditoria, and was one of the first commercially 11:30 – 13:00 Room N101 available products offering dynamic control of time delay and level for recreating realism in sound reinforcement. YESTERDAY’S FX TODAY Human ability to determine the direction from which sound arrives is due to binaural hearing, the brain being Presenter: Alex Case, University of Massachusetts able to detect differences between sounds received by Lowell, Lowell, MA, USA the two ears from the same source and thus to deter- With affordable digital audio tools being continuously mine angular direction. Alan Blumlein 1931 invented, refined, improved, extended, and upgraded, we If two successive sounds are heard as fused, the loca- are lucky to be a part of the audio industry at this tion of the total sound is determined largely by the loca- moment. We have no excuse not to create original, tion of the first arriving sound. This, among other things, beautiful art. What we do with today’s ability to do any- stops you getting confused when a sound comes at you thing can be informed by the creative and technical from two speakers or audio sources at once. Helmut achievements expressed in touchstone recordings Haas 1946 decades ago. This tutorial takes a close look at some iconic moments of signal processing innovation in recorded music history, undoing, isolating, and analyzing the effects for our edification. Tutorial 7 Sunday, May 18 12:30 – 14:30 Room L

Student Event/Career Development SURROUND SOUND CONTRIBUTION METHODS EDUCATION FAIR Sunday, May 18, 11:30 – 13:30 Presenters: John McClintock, APT UK (Chair) Topaz Lounge Karl Petermichl, ORF Radio Heinz-Peter Reykers, WDR Institutions offering studies in audio (from short cours- Geir Skaden, Neural Audio es to graduate degrees) will be represented in a “table top” session. Information on each school’s respective Surround contribution over networks are coming into use programs will be made available through displays and in, e.g., Austria, Germany, by the EBU world wide and in the U.S. The intention is to describe these practical academic guidance. There is no charge for schools to ¯ participate. Admission is free and open to all conven- implementations as case studies. tion attendees.

Audio Engineering Society 124th Convention Program, 2008 Spring 19 Technical Program Session P10 Sunday, May 18 new approach that allows for an approximation of 13:00 – 17:30 Room C-D the required B-format signals but is based on a planar microphone configuration only. Compar- SPATIAL AUDIO PERCEPTION AND PROCESSING, isons with the standard DirAC approach confirm PART 2 that the proposed method is able to correctly estimate the desired parameters within a wide range of frequency and the spatial resolution matches Chair: Renato Pelligrini, sonic emotion ag, Obergltt the human perception. (Zurich), Switzerland Convention Paper 7375

13:00 14:00 P10-1 Analysis and Adjustment of Planar Microphone Arrays for Application in Directional Audio Coding—Markus Kallinger,1 P10-3 User-Dependent Optimization of Wave Field 1 1 Synthesis Reproduction for Directive Sound Fabian Kuech, Richard Schultz-Amling, 1,2 1 2 Fields—Frank Melchior, Christoph Giovanni Del Galdo, Jukka Ahonen, Ville 1,3 3 2 2 Sladeczek, Diemer de Vries, Bernd Fröhlich Pulkki 1 1Fraunhofer Institute for Integrated Circuits IIS, Fraunhofer Institute for Digital Media Technology, Ilmenau, Germany Erlangen, Germany 2 2Helsinki University of Technology, Espoo, Finland Delft University of Technology, Delft, The Netherlands; Directional Audio Coding (DirAC) is a well-estab- 3Bauhaus Universität Weimar, Germany lished and efficient way to capture and reproduce a spatial sound event. In a recording room, DirAC The use of wave field synthesis (WFS) enables requires four spatially coincident microphones to the correct localization of sources over a large estimate the desired parameters, i.e., direction-of- listening area. This works well for simulated arrival and diffuseness of sound: one omnidirec- point sources outside the listening area. The tional and three figure-of-eight microphones point- perception of focused sources is only correct for ing along the axes of a three-dimensional a subspace of the listening area. The subspace Cartesian coordinate system. In most consumer depends on the selected set of loudspeakers applications only two dimensional scenes need to used for the reproduction of the focused source. be reproduced, implying that only two figure-of- If the position of the listener is known the selec- eight microphones are required. Furthermore, tion of loudspeakers as well as the signal instead of directional microphones, arrays of processing can be optimized. By the use of con- omnidirectional microphones are considered for tinuous tracking of the listener this adaptation economic reasons. Therefore, we investigate vari- can be done in real time. The same data can be ous two-dimensional microphone configurations used to simulate a specific directivity of a source with respect to their usability for DirAC. We derive and optimize a corresponding room simulation theoretical limits for the correct estimation of both for the tracked listener. We present a wave field direction-of-arrival and diffuseness for the most synthesis system for the simulation of directive suitable planar arrays. Furthermore, we suggest a focused sources including room simulation, way to equalize the systematic bias for the direc- which is continuously optimized for the position tion-of-arrival estimation, introduced by the of a tracked listener. Our observations confirm discrete planar arrays. that this approach significantly improves the Convention Paper 7374 localization and sound quality of focused sources located inside the listening area. Convention Paper 7376 13:30 P10-2 Planar Microphone Array Processing for the 14:30 Analysis and Reproduction of Spatial Audio P10-4 Spatial Audio Object Coding (SAOC)—The Using Directional Audio Coding—Richard Upcoming MPEG Standard on Parametric 1 1 Schultz-Amling, Fabian Kuech, Markus Object Based Audio Coding—Jonas 1 1 2 Kallinger, Giovanni Del Galdo, Jukka Ahonen, Engdegård,1 Barbara Resch,1 Cornelia Falch,2 2 Ville Pulkki Oliver Hellmuth,2 Johannes Hilpert,2 Andreas 1 Fraunhofer Institute for Integrated Circuits IIS, Hoelzer,2 Leonid Terentiev,2 Jeroen Breebaart,3 Erlangen, Germany Jeroen Koppens,4 Erik Schuijers,4 Werner 2 Helsinki University of Technology, Espoo, Finland Oomen4 1 Recording and reproduction of spatial audio Dolby Sweden, Stockholm, Sweden 2 becomes more and more important, as multi- Fraunhofer Institute for Integrated Circuits IIS, channel audio applications gain increasing atten- Erlangen, Germany 3 tion. Directional Audio Coding (DirAC) represents Philips Research Laboratories, Eindhoven, a well proven approach for the analysis and The Netherlands; 4 reproduction of spatial sound. In the analysis Philips Applied Technologies, Eindhoven, part, the direction-of-arrival and the diffuseness The Netherlands of the sound field is estimated in subbands using Following the recent trend of employing para- B-format signals, which can be created with 3-D metric enhancement tools for increasing coding omnidirectional microphone arrays. However, or spatial rendering efficiency, Spatial Audio 3-D microphone configurations are not practical Object Coding (SAOC) is one of the recent addi- in consumer applications, e.g., due to physical tions to the standardization activities in the design constraints. In this paper we propose a MPEG audio group. SAOC is a technique for

20 Audio Engineering Society 124th Convention Program, 2008 Spring efficient coding and flexible, user-controllable directional response pointing straight forward. rendering of multiple audio objects based on Assuming an ideal coincident stereo micro- transmission of a mono or stereo downmix of the phone, the directional response of this center object signals. The SAOC system extends the channel is effectively time and frequency invari- MPEG Surround standard by re-using its spatial ant. Further, the look direction can be steered to rendering capabilities. This paper will describe left and right front directions. The technique is the chosen reference model architecture, the based on the insight that the signal that predicts association between the different operational left from right, modified by limiting the magnitude modes and applications, and the current status of the frequency domain prediction gains, has a of the standardization process. center forward directional response. The center Convention Paper 7377 channel is generated using both, a left-right and a right-left magnitude-limited-predictor signal. Applications of the proposed scheme are use of 15:00 stereo microphones as “digital steerable shot P10-5 Focusing of Virtual Sound Sources in Higher gun microphones” and center channel genera- Order Ambisonics—Jens Ahrens, Sascha tion for music recording. Spors, Technische Universität Berlin, Berlin, Convention Paper 7380 Germany Higher order Ambisonics (HOA) is an approach 16:30 to the physical (re-)synthesis of a given wave P10-8 Spatial Sound in the Use of Multimodal field. It is based on the orthogonal expansion of Interfaces for the Acquisition of Motor Skills the involved wave fields formulated for interior —Pablo F. Hoffmann, Aalborg University, problems. This implies that HOA is per se only Aalborg, Denmark capable of recreating the wave field generated by events outside the listening area. When a vir- This paper discusses the potential effectiveness tual source is intended to be reproduced inside of spatial sound in the use of multimodal inter- the listening area, strong artifacts arise in certain faces and virtual environment technologies for listening positions. These artifacts can be signifi- the acquisition of motor skills. Because skills are cantly reduced when a wave field with a focus generally of multimodal nature, spatial sound is point is reproduced instead of a virtual source. discussed in terms of the role that may play in However, the reproduced wave field only coin- facilitating skill acquisition by complementing, or cides with that of the virtual source in one substituting, other sensory modalities. An half-space defined by the location and nominal overview of related research areas on audiovisu- orientation of the focus point. The wave field in al and audiotactile interaction is given in connec- the other half-space converges toward the focus tion to the potential benefits of spatial sound as point. a means to improve the perceptual quality of the Convention Paper 7378 interfaces as well as to convey information that may prove critical for the transfer of motor skills. Convention Paper 7381 15:30

P10-6 Listener Envelopment—What Has Been Done 17:00 and What Future Research Is Needed?—Dan Nyberg, Jan Berg, Luleå University of Technology, P10-9 Evaluating the Sensation of Envelopment Piteå, Sweden Arising from 5-Channel Surround Sound Recordings—Sunish George,1 Slawomir In concert hall acoustics, the perceived spatial im- Zielinski,1 Francis Rumsey,1 Søren Bech2 pression and/or spaciousness are characterized 1University of Surrey, Guildford, Surrey, UK by the two attributes apparent source width (ASW) 2Bang & Olufsen a/s, Struer, Denmark and listener envelopment (LEV). For LEV there are no clear consensus across the results of previ- This paper discusses a series of listening tests ous work. This paper aims to discuss the research conducted in the UK and Denmark to evaluate performed on LEV and how these research results the perceived envelopment of surround audio are confirming or contradicting each other. There recordings. The listening tests were designed is a consensus on the arrival angle of the later to overcome some drawbacks (such as range sound energy and its influence on LEV, whereas equalization bias) present in the scores of a there is no clear agreement on the delay time and listening test based on ITU-R. BS. 1534-1 Rec- frequency content of the late reflections. ommendation (MUSHRA). In this method the Convention Paper 7379 listeners were asked to evaluate the envelopment of 5-channel surround sound recordings using a 100-point continuous scale. In order to 16:00 calibrate the scale, two anchor recordings were P10-7 Obtaining a Highly Directive Center Channel used to define points 15 and 85 on the scale. from Coincident Stereo Microphone Signals The anchor recordings were selected by means —Christof Faller, Illusonic LLC, Lausanne, of a formal listening test and interviews with the Switzerland listeners. According to the obtained results, the proposed method provides repeatable results. Time-frequency based postprocessing applied to Convention Paper 7382 a coincident stereo recording is proposed to generate an audio signal with a highly directive ¯

Audio Engineering Society 124th Convention Program, 2008 Spring 21 Technical Program Session P11 Sunday, May 18 This paper presents work carried out with the 13:00 – 17:30 Room E-F objective of designing a novel percussion synthesizer based on a physical model of a plate- ANALYSIS AND SYNTHESIS OF SOUND based percussion instrument. The algorithm has been implemented in real-time for the first time, Chair: Olivier Warusfel, IRCAM, Paris, France on a Field Programmable Gate Array (FPGA) chip allowing a number of parameters such as 13:00 excitation value, stroke location, and plate stiff- ness, to be changed in real-time. This presents P11-1 An Improved Pattern-Matching Method for the player with a number of new modes of playa- Piano Multipitch Detection—Luis Ortiz- bility but requires the definition and design of a Berenguer, Francisco J. Casajus-Quiros, Elena flexible interface that gives the extensive access Blanco-Martin, Technical University of Madrid, to the sound world of the synthesis model. Madrid, Spain Details of the hardware implementation architec- A previous method presented by the authors car- ture are put forward as well as fixed point/float- ried out multipitch piano sound identification by ing point computation aspects that impact the using a pattern-matching process. In that method, instrument’s playability. the identification required, besides the matching- Convention Paper 7385 metric calculation, both a spectral predetection process and a validation step. Predetection 14:30 allowed selection of a subset out of the eighty- eight patterns, whereas the validation verified P11-4 Dual Noise Suppression in Hearing Aids— whether the detected note were actually in the Anton Schlesinger, Marinus M. Boone, Delft analyzed spectrum. Both highly increased the University of Technology, Delft, The Netherlands true-positive detections ratio, but they imposed A combined processing scheme for the restrictions to the identification of complex real enhancement of speech intelligibility in hearing sounds (e.g., two-hands playing). This paper pre- aids is presented. The approach utilizes an opti- sents an improvement in the method that allows mized beam-forming method in connection with getting rid of both, predetection and validation, by a biologically inspired processing model of mod- using a modified matching-metric algorithm. This ulation perception and binaural interaction. work has been supported by the Spanish National Convention Paper 7386 Project TEC2006-13067-C03-01/TCM. Convention Paper 7383 15:00 13:30 P11-5 Automatic Sound Recognition for Security Purposes—Pawel Zwan, Gdansk University P11-2 Polyphonic Piano Transcription Based on of Technology, Gdansk, Poland Spectral Separation—Julio Jose Carabias-Orti, Pedro Vera-Candeas, Nicolas Ruiz-Reyes, Raul In the paper an automatic sound recognition Mata-Campos, Francisco Jesus Cañadas- system is presented. It forms a part of a larger Quesada, University of Jaén, Linares, Jaén, security system developed in order to monitor Spain outdoor conditions for non-typical audio-visual events. The analyzed audio signal is being We propose a discriminative model for polyphonic recorded from a microphone mounted outdoor, piano transcription. Spectral features are obtained thus non-stationary noise of a significant energy individually for each note. To solve the overlap- may be present in it. In the paper an especially ping partial problem, we apply spectral separation designed algorithm for an outdoor noise reduc- by estimating the spectral envelope for each note. tion is presented, non-typical events in audio For classifying purposes, support vector stream are automatically detected and parame- machines (SVM) are trained on the spectral ener- terized. Parameter values of various audio gy inferred from these spectral features. We apply events are analyzed and sounds are automati- a scheme of one-versus-all (OVA) SVM classifiers cally recognized. The automatic recognition to discriminate frame-level note instances. To accuracy obtained for various feature vectors decrease the high frequency notes residual ener- and some chosen recognition systems is com- gy due to the downward notes shared partials, a pared. The conclusions are derived and a future method to cancel the interferences from the plan of experiments is proposed. downward notes to the upward notes has been Convention Paper 7387 developed. The classifier output is filtered with a hidden Markov model. Our approach has been 15:30 tested with synthesized and real piano recordings obtaining very promising results. P11-6 Multipitch Estimation of Harmonically- Convention Paper 7384 Related Event-Notes by Improving Harmonic Matching Pursuit Decomposition—Francisco 14:00 Jesus Canadas-Quesada, Pedro Vera-Candeas, Nicolas Ruiz-Reyes, Raul Mata-Campos, Julio P11-3 Toward a Real-Time Implementation of a Jose Carabias-Orti, University of Jaén, Linares, Physical Modeling Based Percussion Jaén, Spain Synthesizer—Katarzyna Chuchacz, Roger Woods, Sile O’Modhrain, Queen’s University In this paper we propose a note detection Belfast, Belfast, Northern Ireland, UK approach based on harmonic matching pursuit (HMP) and specifically designed to detect simul-

22 Audio Engineering Society 124th Convention Program, 2008 Spring taneous notes. However, HMP is not able to 17:00 decompose harmonic sounds in different harmonic atoms when their fundamental fre- P11-9 Circular Pitch Space Based Musical Tonality Analysis—Gabriel Gatzsche,1 Markus quencies are harmonically-related. To solve this 2 2 problem, we propose an algorithm, called atomic Mehnert, Daniel Arndt, Karlheinz Brandenburg1 spectral smoothness (SS), which works over the 1 harmonic atoms obtained by HMP. This algo- Fraunhofer Institute for Digital Media Technology, Ilmenau, Germany rithm is based on the spectral smoothness prin- 2 ciple that supposes that the spectral envelope of Technische Universität Ilmenau, Ilmenau, a harmonic sound usually forms smooth con- Germany tours. Our proposal shows promising results for The focus of this paper is to give an overview of polyphonic musical signals with two harmonical- existing circular pitch spaces, its special properly-related note-events. ties and application for semantic audio analysis. Convention Paper 7388 Beside this the symmetry model is proposed as a framework to describe the inter-model relation- 16:00 ships between different circular pitch spaces. Similar to color spaces in vision musical pitch P11-7 Amplitude Modification Algorithms within spaces organize pitches in a way that seman- the Framework of Physical Modeling and of tic/cognitive/theoretical/physical relationships Haptic Gestural Interaction—Alexandros between tones become geometrically apparent. Kontogeorgakopoulos, Claude Cadoz, Institute Within the last years pitch spaces were mainly National Polytechnique de Grenoble, Grenoble, the subject of music theory. But they become France more and more interesting for semantic analysis Every underlying technique that has been used of musical audio signals. Pitch spaces can be for the realization of audio effects since the applied to key and chord recognition, similarity beginning of electronic and computer music, calculation of musical pieces, genre estimation, introduced different types of sound modifications tension analysis, or harmonic change detection. and proposed new ways of control. The advent Convention Paper 7391 of digital signal processing has stimulated the audio processing researchers to a great extent; Exhibitor Seminar 2 Sunday, May 18 thus a variety of algorithms were designed to 13:00 – 14:00 Room P provide novel sound modifications. On the other hand, physical modeling and digital simulation formalisms have been principally used for the FERROTEC merely imitation and emulation of older sound processing systems. The aim of this paper is to Presenters: M. Klasco propose three physical models conceived to B. Moskowitz offer sound modifications that mainly alter the The Past, Present and Future of Ferrofluids amplitude of audio signals. The originality of this in the Loudspeaker Industry case is not the resulted audio modifications but their transposition in the framework of physical Ferrofluids were introduced to the loudspeaker market- modeling and digital simulation, which outlines place by Ferrotec (formerly Ferrofluidics) in the early an alternative control procedure. 1970s. Millions of loudspeakers have been produced Convention Paper 7389 with ferrofluid in the voice coil assembly to improve performance and speaker life. New advancements in 16:30 research have expanded the range of ferrofluid products available to the industry. P11-8 Circular Pitch Space Based Harmonic 1 Change Detection—Markus Mehnert, Sunday, May 18 13:00 Room H Gabriel Gatzsche,2 Daniel Arndt,1 Karlheinz 2 Standards Committee Meeting on SC-03-12 Brandenburg Forensic Audio 1Technische Universität Ilmenau, Ilmenau, Germany 2Fraunhofer Institute for Digital Media Sunday, May 18 13:00 Room K Technology, Ilmenau, Germany Technical Committee Meeting on Audio Recording and Mastering Systems This paper introduces a novel method for detecting harmonic boundaries in musical audio sig- Workshop 13 Sunday, May 18 nals. These boundaries are important for chord 13:30 – 16:30 Room N101 analysis and between two boundaries is just one particular chord. This event-driven analysis of AUDIO AND AUDIO-TACTILE WARNINGS musical audio signals is a better basis for a fol- AND ALERTS: FORMING TODAY’S SOUNDSCAPE lowing chord analysis than the traditionally frame-based-only concept. The method itself works with circular pitch spaces (CPS). The idea Chairs: Durand Begault, NASA Ames Research behind CPS is the calculation of parameters that Center, Moffett Field, CA, USA summarize high level aspects of the audio signal Ellen Haas, Army Research Lab such as semantic and music theoretical relationships. Using CPSs entails good results in detect- Panelists: Caryl Baldwin, George Mason University ing harmonic changes. Judy Edworthy, University of Plymouth ¯ Convention Paper 7390 Anne Guillaume, ISMA

Audio Engineering Society 124th Convention Program, 2008 Spring 23 Technical Program Elif Ozcan, Technical University of Delft movie frames can be corrected. The paper pre- Rene VanEgmond, Technical University of sents the description of the image based drift, Delft wow, and flutter determination method and the experiments confirming the theoretical findings. Audio warning and alert signals are omnipresent in the Convention Paper 7392 soundscape of modern society, ranging from warnings (sirens), to sounds from human interfaces (pushbutton sounds on devices), to cell phone ring tones. Attention 14:00 has only been given fairly recently in the human factors research community toward the best means of making P12-2 The Norwegian Institute of Recorded Sound: caution and warning signals discriminable, audible, and From Collection to Archive to Public Private tolerable; some high-stress human interfaces are also Partnership—Mark Drews,1 Jacqueline von Arb2 concerned with increasing the richness of the semantic 1University of Stavanger, Stavanger, Norway content of these alarms. Ironically, while it is well docu- 2Norwegian Institute of Recorded Sound, mented that people have ignored critical audio alarms in Stavanger, Norway flight and railroad operations, personalized cell telephone In 2006, the Norwegian Institute of Recorded ring tones have an apparently high “hit” rate, while simul- Sound (NIRS) entered into a partnership with taneously alerting (and annoying) others. Discussion Memnon Audio Archiving Services to form Mem- topics for this workshop include assessment of current Nor, a commercial audio archiving service based standards and best practices for design of auditory in Stavanger, Norway. This paper traces the alerts; new approaches to forming the sound design of evolution of the Norwegian Institute of Recorded alarms, including multimodal approaches such as includ- Sound from a private collection of music recording tactile displays; new approaches to forming the ings to a municipally funded audio archive to sound design of alarms, and defining what best (or a public private partnership and discusses worst!) practices might be. We may obtain as a result an the past, the current, and the future challenges idea of how the soundscape of future decades may involved. Details of ongoing activities is included. sound like (or what we would like it to sound like). Convention Paper 7393

Tutorial 8 Sunday, May 18 14:00 13:30 – 15:00 Room O P12-3 Cable-free Audio Delivery for Home Theater Entertainment Systems—Andreas Floros,1 CURRENT STANDARDS Nicolas-Alexander Tatlas,2 John Mourjopoulos,2 OF FILE-INTERCHANGE FORMATS Dimitris Grimanis2 1Ionian University, Corfu, Greece Presenter: Mark Yonge, AES Standards Manager 2University of Patras, Patras, Greece As the broadcasting world is switching to file-based pro- Real time, multichannel audio content delivery duction, it is vital to get to grips with current formats one over the air is expected to significantly simplify is likely to encounter as well as to know the different fla- the interconnection complexity required for set- vors, strengths, and weaknesses. The presenter has an ting up typical home theater applications. How- intimate insight into file-formats, not at least through his ever, despite the technological advantages of work as AES standards manager. wireless networking standards related to high transmission rates and Quality-of-Service sup- Session P12 Sunday, May 18 port, a number of issues has to be additionally 14:00 – 15:30 Topaz Lounge addressed, such as multiple loudspeaker syn- chronization and packet delay/losses containing POSTERS: AUDIO ARCHIVING, STORAGE, compressed quality and multiplexed audio data. RESTORATION, AND CONTENT MANAGEMENT In this paper further developments in the area of AND AUDIO NETWORKING wireless audio delivery are presented by considering in detail multichannel reproduction for wireless home theater applications. Using both sub- 14:00 jective and objective performance evaluation P12-1 Drift, Wow, and Flutter Measurement and criteria, it is shown that cable-free multichannel Reduction in Shrunken Movie Soundtracks— audio playback is feasible under specific net- Przemek Maziewski, Adam Kupryjanow, Andrzej working and audio coding conditions. Czyzewski, Gdansk University of Technology, Convention Paper 7394 Gdansk, Poland 14:00 The paper presents the method and algorithms used to determine and reduce drift, wow, and P12-4 Adaptive Playout for VoIP Based on the flutter in shrunken movie tapes. The idea behind Enhanced Low Delay AAC Audio Codec— the algorithms is to use image processing for Jochen Issing, Nikolaus Färber, Manfred Lutzky, calculating the local tape shrinkage, which is one Fraunhofer Institute for Integrated Circuits IIS, of the reasons for drift, wow, and flutter. The Erlangen, Germany shrinkage can be calculated via analyzing the The MPEG-4 Enhanced Low Delay AAC (AAC- image height of: a movie frame, sprocket hole, ELD) codec extends the application area of the pitch, or another standardized movie tape ele- Advanced Audio Coding (AAC) family toward ment; and then it can be expressed as the drift, high quality conversational services. Through wow, and flutter characteristic. After the charac- the support of the full audio bandwidth at low teristic determination both the soundtrack and delay and low bit rate, it offers excellent support

24 Audio Engineering Society 124th Convention Program, 2008 Spring for enhanced VoIP applications. In this Historical Event paper we provide a brief overview of the AAC- M3 RECORDING TECHNIQUE ELD codec and describe how its codec structure Sunday, May 18, 14:00 – 15:00 can be exploited for IP transport. The overlap- Room N ping frames and excellent error concealment make it possible to use frame insertion/deletion Presenter: Hans Lauterslager in order to adjust the playout time to varying network delay. A playout algorithm is proposed that Hans Lauterslager, retired balance engineer with Poly- estimates the jitter on the network and adapts gram-Philips, gives a description of 3-microphone the size of the de-jitter buffer in order to mini- recording technique as applied by Mercury and Philips in mizes buffering delay and late loss. Considering the sixties. typical network conditions and the same average delay, it is shown that the playout algorithm can reduce the loss rate by more than one magni- Sunday, May 18 14:00 Room K tude compared to fixed playout. Technical Committee Meeting on Electro Magnetic Convention Paper 7395 Compatibility

Student Event/Career Development RECORDING COMPETITION—STEREO Master Class 1 Sunday, May 18 Sunday, May 18, 14:30 – 18:00 14:00 – 16:00 Room B Room L

LOUDSPEAKER PARAMETERS The Student Recording Competition is a highlight at each convention. A distinguished panel of judges par- Presenters: Richard Small, Harman-Motive Inc., USA ticipates in critiquing finalists in each of the categories Neville Thiele, University of Sydney, in an interactive presentation during the convention. Sydney, NSW, Australia Student members can submit stereo and surround recordings in the categories classical, jazz, folk/world The loudspeaker parameters provide a procedure that is music, and pop/rock. Meritorious awards will be pre- now established universally for specifying loudspeaker sented at the closing Student Delegate Assembly Meet- drivers and designing their enclosures and associated ing on Tuesday. equalizers. Their derivation and measurement involves Judges: simplifications and approximations to what is, in fact, a Classical (14:30–15:20): Ken Blair, Leo de Klerk, Theo complex acoustical/mechanical/electrical system. They Wubbolts apply most accurately to the response of the system to Jazz/Blues (15:20–16:10): Jim Anderson, Ulrike small signals but nevertheless allow measurements to be Schwartz, Johannes Wohlleben made with surprisingly simple equipment and the perfor- Folk/World Music (16:20–17:10): Eva Bauer-Oppel- mance of loudspeaker systems to be predicted with high land, Darcy Proper, Mark Derksen precision. Since the initial publications between 1961 Pop/Rock (17:10–18:00): Carlos Abrect, Alex Case, and 1973, their effectiveness has been increased Werner Penasaert through additions and refinements in understanding the mechanisms involved and in measuring and calculating Recording competition sponsors: Schoeps, PMC, them, made by a number of researchers and reported in Sennheiser, Neumann a number of places. This master class aims to present them in a unified whole, correct some miscon- Historical Event ceptions, and answer some frequently-asked questions 100 YEAR OF THE TRIODE that, over the years, have arisen and mystified some stu- Sunday, May 18, 15:00 – 16:00 dents and designers. Room N

Presenter: Guido Tent Live Sound Seminar 5 Sunday, May 18 14:00 – 18:00 Forum “100 years of the Triode” is a short journey through the LIVE MIXING WORKSHOP history of electronics, which started in 1907 with the invention of the triode. This lecture gives an overview of how electronics developed, shows a documentary of Presenter: Gregor Zielinsky, Sennheiser Electronic the (re)making of the triode, and ends with an audio Corporation demonstration. This session presents a live sound check with a band including In Ear Check Live on Stage. Similar workshops Tutorial 9 Sunday, May 18 have been presented in many places worldwide, e.g., at 15:00 – 18:00 Room O AES conventions in Barcelona and Vienna, a.o. The focus this time will be on playback sound quality by loud- ROOM ACOUSTICS MODELING speakers for drums, guitar, and piano. The drum sound, especially, is probably the most ambitious —live or in the Presenters: Tapio Lokki, Helsinki University of studio. A team of experienced presenters will show many Technology, Espoo, Finland professional tricks. Lauri Savioja, Helsinki University of Technology, Espoo, Finland • How to record percussion • How to get a piano loud This tutorial will cover commonly applied wave-based • How to get highest sound quality from a guitar and ray-based methods in room acoustics modeling. The main emphasis is to explain basic principles of physi- ¯

Audio Engineering Society 124th Convention Program, 2008 Spring 25 Technical Program cally-based modeling methods. In addition, pros and Source Separation techniques applied to music cons of each method are discussed. Examples are given mixtures are able to extract relevant information with several modeling methods. The modeling results, that can be very useful for many applications, namely room acoustical parameters and auralizations, such as music remixing and reprocessing, lyrics are also discussed in the light of each method. recognition or music information retrieval. Among all the sources present in modern music Sunday, May 18 15:00 Room K themes, the singing voice has a especial interest Technical Committee Meeting on Microphones because it is the only one that combines music, and Applications lyrics, and expression. In this paper we propose a system designed for extracting the singing Sunday, May 18 15:30 Room H voice from stereo recordings in different steps. Standards Committee Meeting on SC-03-02 Transfer This system combines panning information and Technologies pitch tracking, allowing the refinement of the time-frequency mask applied for extracting a Session P13 Sunday, May 18 vocal segment, and thus, improving the separa- 16:00 – 17:30 Topaz Lounge tion. An application example is discussed. Convention Paper 7397 POSTERS: SIGNAL PROCESSING, SOUND QUALITY DESIGN 16:00 P13-3 The Downsampling Dilemma: Perceptual 16:00 Issues in Sample Rate Reduction—Brett Leonard, New York University, New York, NY, USA P13-1 Time-Alignment of Multiway Loudspeakers with Group Delay Equalization: Part I—Sunil Many options currently exist for sample rate con- Bharitkar, Chris Kyriakakis, Tom Holman, version. With sample rate reduction playing an University of Southern California and Audyssey integral part in the modern production world, down- Labs, Los Angeles, CA, USA sampling algorithm quality is more important than ever. This paper presents data exploring the differ- In this paper, a first of two-parts, a technique for ences in sample rate reduction algorithms. While time-aligning the driver responses (viz., woofer, certain tests clearly display differences in the quali- mid-range, and tweeter responses) in a multiway ty of the algorithms, listening test data shows the loudspeaker system is presented. Generally, average listener is unable to repeatedly discern the woofers exhibit a much larger time-of-arrival delay difference in sample rate reduction methods. at a listening position, compared to the mid-range Convention Paper 7398 and high-frequency drivers due to the presence of crossover networks. Moreover, the time-of-arrival 16:00 delay for all drivers is frequency dependent ex- hibiting a large variation over the audible frequen- P13-4 NU-Tech: The Entry Tool of the hArtes cy domain. Due to these differences a two-part Toolchain for Algorithms Design—Ariano study was undertaken to understand the effects of Lattanzi,1 Ferruccio Bettarelli,1 Stefania Cecchi2 these variations, quantitatively and qualitatively, in 1Leaff Engineering, Porto Potenza Picena (MC), the direct as well as reverberant field in typical lis- Italy tening rooms with diverse content. Various multi- 2Univerista Politecnica delle Marche, Ancona, way loudspeakers, measured in one of the ane- Italy choic chambers at Audyssey, were selected to provide a diverse corpus of responses. In this first The aim of the hArtes project is to facilitate and part we present the motivation behind the system automate the rapid design and development of used for applying all-pass filters to process audio heterogeneous embedded systems, targeting a signals being delivered to the multiway speaker combination of a general purpose embedded and propose a time-delay difference equalization processor, digital signal processing, and recon- technique, between the drivers of the multiway figurable hardware. In this paper we present the loudspeakers showing group-delay equalization NU-Tech platform, the main entry tool from the while retaining a flat magnitude response. Clearly, hArtes toolchain, which has the role of assisting applying all-pass filters will result in temporal- the designers in tuning and possibly improving smearing of the measured response. Further- the input algorithm at the highest level of more, an investigation using perceptually abstraction. A brief description of the project motivated variable-octave complex smoothing of itself will be given and its vocation to audio high- responses, and designing all-pass filters based on lighted through a case study application. this phase-smoothed data, will also be undertak- Convention Paper 7399 en. Quantitative results obtained will be presented in this paper whereas the next part of the two-part 16:00 paper will present results from listening tests. P13-5 Recovery of Missing Signals Utilizing (GHA) Convention Paper 7396 Generalized Harmonic Analysis—Applied Interpolation—Teruo Muraoka, Takahiro Miura, 16:00 Tohru Ifukube, University of Tokyo, Tokyo, Japan P13-2 Singing Voice Separation Combining For archiving damaged historical recordings, Panning Information and Pitch Tracking— recoveries of missing portions are as essentially Maximo Cobos, Jose J. Lopez, Technical important as noise reduction. Conventional University of Valencia, Valencia, Spain counter-measures with functional interpolation

26 Audio Engineering Society 124th Convention Program, 2008 Spring are not effective when the missing interval is spectral subtraction. In the proposed system, we long. Inharmonic frequency analysis GHA is make M combinations of observed signals and profitable for this purpose, because the recom- apply them to the correlation-based blind decon- posed signal with frequency components volution. The deconvolved signals are then used obtained by GHA exhibits very long periods. as inputs to the multi-microphone spectral sub- Length of the period is given as an inverse num- traction. These spectral subtractions with the ber of the least common multiple of the rendered multiple deconvolved signals estimate the rever- inharmonic frequency’s periods. This feature is berant energy by using both a frame delay and a very advantageous for signal recovery, and the frequency-dependent weight. Due to the accurate authors devised an extrapolation simply estimation of the reverberant energy, the combi- extending a re-synthesizing waveform obtained nation of correlation-based blind deconvolution through GHA analysis/re-synthesis. The authors with the multi-microphone spectral subtraction got satisfactory results as a whole by applying provides improved dereverberation performance. interpolation combined with forward and back- Performance improvement of the proposed sys- ward extrapolations based upon the abovemen- tem has been confirmed through experiments. tioned method. Results of recovery highly Convention Paper 7402 depend upon characters of signals (such as music), and the authors did not find definite rules 16:00 for setting GHA’s analyzing conditions. Those are given through auditory examination this time. P13-8 Harmonic and Intermodulation Analysis of Convention Paper 7400 Nonlinear Devices Used in Virtual Bass Systems—Nay Oo, Woon-Seng Gan, Nanyang 16:00 Technological University, Singapore P13-6 Combination of Warped and Linear Filter Nonlinear Devices (NLD) are used in virtual bass Structures for Loudspeaker Equalization— system. NLD generates harmonics which in turn German Ramos,1 Jose J. Lopez,1 Basilio Pueo2 create the pitch perception and are used in 1Technical University of Valencia, Valencia, Spain audio bass enhancement systems using psy- 2University of Alicante, Alicante, Spain choacoustics. This paper presents the mathematical derivations and analysis of five different The warping filters were introduced years ago NLD devices, together with intermodulation for loudspeaker equalization in order to solve the analysis of harmonics generated by these NLDs. lack of resolution of the linear filters at low fre- The five NLDS are half-wave rectifier, full-wave quencies, and also to follow the frequency reso- rectifier, square wave, polynomial function, and lution of psycho-acoustic scales like the Bark exponential function. The derivation of harmonic scale, with a more logarithmic than linear behav- analysis equations are based on Fourier Theo- ior. However, this improvement in the frequency rems, Chebyshev Polynomials, and Taylor resolution at low and mid frequencies is done at Series expansions. Besides the harmonics, the expense of loosing resolution at high fre- intermodulation components also resulted from quencies and increasing the complexity of the fil- NLDs. Both mathematical analysis and simula- ter and its implementation computational cost. In tion results are presented for the intermodulation this paper a smart combination of linear and effects of harmonics generated by NLDS. warped filter structures previously developed by Convention Paper 7403 the authors for FIR filters is presented with new contributions and extended to IIR filters. This Tutorial 10 Sunday, May 18 combination saves computational cost and 16:00 – 18:00 Room B obtains a proper frequency resolution at the whole frequency band, obtaining better results TINNITUS. JUST ANOTHER BUZZ WORD? for the same computational cost than when using linear or warped filters alone. The results have been subjectively tested using the ABX Presenters: Neil Cherian, Center for Performance methodology with successfully results. The pre- Medicine, Cleveland Clinic, Cleveland, OH, sented filter structures, methodology, and appa- USA ratus to do the filtering are patent pending. Michael Santucci, Sensaphonics Hearing Convention Paper 7401 Conservation, Inc., Chicago, IL, USA Jan Voetmann, Voetmann-Acoustics, Copenhagen, Denmark 16:00 Tinnitus is a common, yet poorly understood, disorder P13-7 Multichannel Dereverberation System Using where sounds are perceived in the absence of an exter- Modified Correlation-Based Blind nal source (phantom). Significant sound exposure with or Deconvolution and Multi-Microphone without hearing loss is the single most common risk fac- Spectral Subtraction—Jae-Woong Jeong,1 2 3 tor. Tinnitus can be debilitating, affecting quality of life or Young-Cheol Park, Seok-Pil Lee, Dae-Hee even one’s ability to function. Given the potential harm of Youn1 1 sound in the development of tinnitus, more aggressive Yonsei University, Seoul, Korea and proactive attitudes must be taken. In-ear monitoring 2Yonsei University, Wonju, Korea 3 strategies further necessitate meaningful conversations Korea Electronics Technology Institute (KETI), regarding hearing awareness, hearing protection, safe Sungnam, Korea standards for listening, and appropriate safeguards for This paper presents a new multichannel derever- products. This tutorial introduces the concept of tinnitus, beration system combining modified correlation- the pertinent anatomy and physiology, the audiologic pa- ¯ based blind deconvolution with multi-microphone rameters of tinnitus, current research, guidelines for

Audio Engineering Society 124th Convention Program, 2008 Spring 27 Technical Program identifying high risk behaviors, and how to determine that engineering and its related fields. The series is featured you have a problem. twice annually at both the United States and European AES conventions. Established in May 1999, The Richard Historical Event C. Heyser Memorial Lecture honors the memory of RECORDING HISTORY Richard Heyser, a scientist at the Jet Propulsion Labora- Sunday, May 18, 16:00 – 17:00 tory, who was awarded nine patents in audio and com- Room N munication techniques and was widely known for his ability to clearly present new and complex technical Presenter: Bert van der Wolf ideas. Heyser was also an AES governor and AES Silver Medal recipient. The Richard C. Heyser distinguished lecturer for the “The importance of technical specifications of recording 124th AES Convention is A. J. (Guus) Berkhout. He has equipment, and the sound quality of a recording, in the written several hundred scientific papers and a number communication of the Emotion in a recorded Musical of books in the fields of acoustics, geophysics, and inno- Performance . . . and how our brain and ears respond to vation. In the late eighties he introduced the concept of reproduced acoustical information in several different wave field synthesis (WFS) and wave field analysis audio formats.” A personal report of 20 years practical (WFA) in audio engineering. In 2003 he received the experience in Record Producing, and development of highest international award in the field of Exploration recording equipment through research of, and fascina- Geophysics. tion for the human audio/music perception. Berkhout is a member of the Royal Netherlands Acad- emy of Arts and Sciences (KNAW) and the Netherlands Sunday, May 18 16:00 Room K Academy of Engineering (AcTI). He received honorary Technical Committee Meeting on Loudspeakers memberships of the Society of Exploration Geophysicists and Headphones (SEG) and the European Association of Geophysicists and Engineers (EAGE). He also serves on a number of Workshop 14 Sunday, May 18 Governing Boards in the scientific and business commu- 16:30 – 18:00 Room N101 nities. The title of his lecture is, “The Big Challenges in Audio: A Glance into the Future.” Sound is an important information carrier. In speech CHANGING TO FILE-BASED PRODUCTION and music the source carries the message, but in acoustical imaging the message is given by the medium. Chair: Florian Camerer, ORF – Austrian TV In all these applications, it is important to realize that sound is a wave phenomenon, often with complex wave Panelists: Steinar Bjørlykke, NRK – Norwegian fronts and complex time signals. For example, in Broadcasting enclosed spaces wavefields represent an intricate inter- Dominique Brulhart, Merging Technologies ference pattern of multi-source signals and multi-bound- Jean-Christophe Liechti, TSR – Télévision ary reflections. Here, audio systems have the important Suisse Romande task to enhance the social function of these spaces. If we Brendan Mallon, BBC Scotland want to make the next big step in improving audio solutions for demanding listening environments, we should While the radio world has already several years of expe- challenge traditional believes and rethink current design rience with file-based production and transmission, the methods. television side is just at the beginning of the transition On the one hand, there is the dimension of new tech- toward a fully file-based workflow. This is a significant nological capability. Wavefield knowledge should have a change—one that poses many challenges on the engi- major impact on the way we develop the next generation neers, the infrastructure, the equipment, and the system of audio products. This technological expedition into the that ties it all together. Buzzwords like Content Manage- future will lead us to the exciting world of transducer ment System, MXF (Material Exchange Format), Buffer arrays and matrix processors, both for analysis (WFA) Storage, or Playout Server enter the world of the audio and synthesis (WFS) purposes. New functionality will engineer—often causing considerable panic. In this include variable acoustics, focused sound delivery, and workshop representatives of broadcasters who have selective signal enhancement. already gained some experiences in this field will share On the other hand, there is the dimension of improved their viewpoints, introduce their specific solutions to com- user value. For commercial success, technological excel- mon problems, and generally raise the awareness and lence is necessary but the perceived value by the market urgency to dive into this world. It will soon be the daily is of overriding importance. Knowledge of the changing life of us all. soft values in society should inspire the new solutions. This human-centered expedition into the future will bring Sunday, May 18 17:00 Room H us to the innovative world of integrated audio-optical sys- Standards Committee Meeting on SC-03-04 Storage tems. Examples are the combination of variable and Handling of Media acoustics with variable lighting (WFS-plus), the integration of hearing aids with spectacles (“hearing glasses”), Special Event and the connection of optical cameras with highly direc- OPEN HOUSE OF THE TECHNICAL COUNCIL AND tional microphones (“forensic audio products”). THE RICHARD C. HEYSER MEMORIAL LECTURE In conclusion, the message of the 2008 Richard C. Sunday, May 18, 18:00 – 19:30 Heyser Memorial Lecture is: Do not try to predict the Room B future, but have the ambition to create the future. Berkhout’s presentation will be followed by a reception Lecturer: A. J. (Guus) Berkhout hosted by the AES Technical Council. The Heyser Series is an endowment for lectures by emi- nent individuals with outstanding reputations in audio

28 Audio Engineering Society 124th Convention Program, 2008 Spring Special Event 4Katholieke Universiteit Leuven, Leuven, Belgium ORGAN RECITAL BY GRAHAM BLYTH Sunday, May 18, 20:30 – 21:30 One of the research topics within the HearCom Thomas Aquinen Kerk project, a European project that studies the Hunzestraat 87, Amsterdam impact of hearing loss on communication, is to find methods with which the speech quality as Graham Blyth’s traditional organ concert will be given at perceived by the hearing impaired can be mea- Thomas Aquinen Kerk (walking distance fom the conven- sured objectively. ITU-T Recommendation P.862 tion center). The organ was built by the local firm of Flen- PESQ and its wideband extension P.862.2, are trop in 1959, one year after the famous organ they built obvious candidates for this despite the fact that for the Bush Reisinger Museum at Harvard University, they were developed for normal hearing sub- and very similar in stoplist. jects. This paper investigates the extent to which The organ is a so-called straight organ with wind PESQ and possible simple extensions can be chests and mechanical tracker action. There are 30 used to measure the quality of speech signals as stops totally, divided over Great Organ (in the middle), perceived by hearing impaired subjects. Choir Organ (the little organ in front), Brustwerk (above Convention Paper 7404 the manuals), and Pedal Organ (on both sides of the Great Organ). The organ case was made in the style 09:30 common the fifties and sixties. The featured works Toccata, Adagio, and Fugue in C P14-2 Subjective Evaluation of Speech Quality in a by Bach; Variations on Mein junges Leben hat ein End, Conversational Context—Emilie Geissner,1 by Sweelink, 2nd Organ Sonata by Mendelssohn, plus Valérie Gautier-Turbin,1 Marie Guéguin,2,3 works by Dubois, Vierne, and Mushel. Laetitia Gros1 Graham Blyth was born in 1948, began playing the 1France Télécom R&D, Lannion, France piano aged 4 and received his early musical training as a 2Laboratoire Traitement du Signal et de l’Image, Junior Exhibitioner at Trinity College of Music in London, Rennes, France England. Subsequently, at Bristol University, he took up 3Université de Rennes, Rennes, France conducting, performing Bach’s St. Matthew Passion Within the framework of ITU-T, an objective con- before he was 21. He holds diplomas in Organ Perfor- versational model is developed to predict the mance from the Royal College of Organists, The Royal impact of network impairments on the conversa- College of Music and Trinity College of Music. In the late tional quality experienced by a end-user. To train 1980s he renewed his studies with Sulemita Aronowsky and validate such a model, subjective scores are for piano and Robert Munns for organ. He gives numer- required. Assuming that a conversation is made ous concerts each year, principally as organist and of talking, listening, and inter-action activities, a pianist, but also as a conductor and harpsichord player. subjective test protocol is specially designed to He made his international debut with an organ recital at take into account these multidimensional St. Thomas Church, New York in 1993 and since then aspects of the speech quality in a conversation. has played in San Francisco (Grace Cathedral), Los Subjects are asked to evaluate speech quality in Angeles (Cathedral of Our Lady of Los Angeles), talking, listening, and conversational contexts Amsterdam, Copenhagen, Munich (Liebfrauen Dom), separately during three successive tasks. The Paris (Madeleine and St. Etienne du Mont) and Berlin. analyses of several tests show that this method He has lived in Wantage, Oxfordshire, since 1984 where is valid for the assessment of listening, talking, he is currently Artistic Director of Wantage Chamber and conversational quality. Concerts and Director of the Wantage Festival of Arts. Convention Paper 7405 He divides his time between being a designer of professional audio equipment (he is a co-founder and Tech- nical Director of Soundcraft) and organ related activities. 10:00 In 2006 he was elected a Fellow of the Royal Society of P14-3 Contribution of Interaural Difference to Arts in recognition of his work in product design relating Obstacle Sense of the Blind While Walking— to the performing arts. Takahiro Miura, Teruo Muraoka, Shuichi Ino, Tohru Ifukube, University of Tokyo, Tokyo, Japan Session P14 Monday, May 19 09:00 – 12:00 Room C-D Most blind people can recognize some measure of objects existing around them only by hearing. This ability is called “obstacle sense” or “obsta- PSYCHOACOUSTICS, PERCEPTION, cle perception.” It is known that this ability is AND LISTENING TESTS, PART 1 facilitated while the subjects are moving, however, the exact reason of the facilitation has been Chair: Jan de Laat, LUMC, Leiden, The Netherlands unknown. It is apparent that some differences of sounds reaching between both ears significantly 09:00 change while approaching the obstacles. We focused on this phenomenon called interaural P14-1 Speech Quality Measurement for the Hearing difference in order to analyze the facilitation Impaired on the Basis of PESQ—John G. mechanism of the obstacle sense. We investi- Beerends,1 Jan Krebber,2 Rainer Huber,3 gated how the interaural differences change Koen Eneman,4 Heleen Luts4 depending on the head rotation while walking 1TNO Information and Communication and then measured the DL (Difference Limen) of Technology, Delft, The Netherlands the interaural difference. Furthermore, we com- 2Technische Universität Dresden, Dresden, pared the measurement data and the DL with Germany, now with Sysopendigia PLC, the relationship between the subject-to-obstacle Helsinki, Finland distance and then discussed one of the factors ¯ 3HörTech GmbH, Oldenburg, Germany

Audio Engineering Society 124th Convention Program, 2008 Spring 29 Technical Program of the facilitating the obstacle sense. be studied. Stimuli with combinations of two-sylla- Convention Paper 7406 ble words will be presented simultaneously in speakers to subjects, and the number of correct 10:30 identifications will be measured. In one category of stimuli speech, melody will be removed and P14-4 The Accuracy of Localizing Virtual Sound replaced with a monotonous pitch, equal for all Sources: Effects of Pointing Method and words. One category will have all words presented Visual Environment—Piotr Majdak, Bernhard from one speaker only. Conclusions will be related Laback, Matthew Goupell, Michael Mihocic, to earlier studies and common theories, the Cock- Austrian Academy of Sciences, Vienna, Austria tail party effect among others. The ability to localize sound sources in a 3-D- Convention Paper 7409 space was tested in humans. The subjects lis- tened to noises filtered with subject-specific Session P15 Monday, May 19 head-related transfer functions. In the experi- 09:00 – 12:00 Room E-F ment using naïve subjects, the conditions included the type of visual environment (darkness or SIGNAL PROCESSING, SOUND QUALITY DESIGN structured virtual world) presented via head mounted display and pointing method (head and Chair: Jan Abildgaard Pedersen, Lyngdorf Audio, manual pointing). The results show that the Skive, Denmark errors in the horizontal dimension were smaller when head pointing was used. Manual pointing 09:00 showed smaller errors in the vertical dimension. P15-1 Characterization of the Multidimensional Generally, the effect of pointing method was sig- Perceptive Space for Current Speech and nificant but small. The presence of structured vir- Sound Codecs—Thierry Etame,1,2 Laetitia tual visual environment significantly improved Gros,1 Catherine Quinquis,1 Gérard Faucon,2,3 the localization accuracy in all conditions. This Régine Le Bouquin Jeannes2,3 supports the benefit of using a visual virtual envi- 1France Télécom R&D, Lannion Cedex, France ronment in acoustic tasks like sound localization. 2University of Rennes, Rennes, France Convention Paper 7407 3INSERM, Rennes, France

11:00 The purpose of our work is to produce a reference system that can simulate and calibrate P14-5 Perceived Spatial Distribution and Width of degradations of speech and audio codecs which Horizontal Ensemble of Independent Noise are currently used on telecommunications net- Signals as Function of Waveform and Sample works, for subjective assessment tests of voice Length—Toni Hirvonen, Ville Pulkki, Helsinki quality. At first, 20 wideband codecs are evaluat- University of Technology, Espoo, Finland ed through subjective tests with the general goal This paper investigates the perceived sound dis- of producing the multidimensional perceptive tribution and width of a horizontal loudspeaker space underlying the perception of current ensemble as a function of signal length, as all degradations. Then, from a verbalization task, it loudspeakers emit simultaneous, white Gauss- appears that the identified attributes are clear/ ian noise bursts. In Experiment 1, subjects indi- muffle, high-frequency noise, noise on speech, cated the perceived distribution of 10 frozen and hiss. Finally, these dimensions are charac- cases where signal length was 2.5 ms. In Exper- terized with correlates such as spectral centroid, iment 2, two cases from the previous test were spectral flatness measure, Mean Opinion Score, investigated with signal lengths of 5-640 ms. The and correlation coefficient. results indicate that (1) ensembles consisting of Convention Paper 7410 different short noise bursts vary in perceived distribution between cases and (2) when the length 09:30 of the signal is increased, the produced sound P15-2 An Automatic Maximum Gain Normalization event is generally perceived more wide. In per- Technique with Applications to Audio ceiving such cases, the hearing system possibly Mixing—Enrique Perez Gonzalez, Joshua utilized some temporal integration and/or adap- D. Reiss, Queen Mary, University of London, tive processes. London, UK Convention Paper 7408 A method for real-time magnitude gain normal- 11:30 ization of a changing linear system has been developed and tested with a parametric filter P14-6 Effect of Minimizing Spatial Separation design. The method is useful in situations where and Melodic Variations in Simultaneously the maximum gain before feedback is needed. Presented Two-Syllable Words—Jon Allan, The method automatically calculates the appro- Jan Berg, Luleå University of Technology, Piteå, priate gain that should be applied in order to Sweden maintain maximum unitary gain. The method This paper will examine two important factors for uses an impulse measurement of a mathemati- the conception Auditory Streaming defined by cal model of the system to be normalized. This is Bregman, pitch, and localization. By removing one particularly useful for mixing engineers, who or two of these factors as possible identifiers to have to continually revise their gain structure in separate sound sources, the importance of each order to maximize gain before feedback. The of them and the effect of reducing both of them will system is also useful in many other situations where solving the analytical solution from the

30 Audio Engineering Society 124th Convention Program, 2008 Spring mathematical model is not possible. development environment the user can concen- Convention Paper 7411 trate on real-time audio algorithm development and performance evaluation in the workstation 10:00 environment. We present the proposed algorithm design method and environment, and its P15-3 An Alternative Approach for the Convolution application to an experimental Voice over Inter- in Time-Domain: The Taches-Algorithm— net Protocol (VoIP) system development. Laurent Millot, Gérard Pelé, ENS Louis-Lumière, Convention Paper 7414 Noisy-le-Grand cedex France We present an alternative temporal approach for 11:30 convolution, providing a new algorithm, called P15-6 New Enhancements to the Automatic Noise the taches-algorithm. Based on interferences Removal (ANR) System Utilizing Improved between the successive delayed and amplified Noise Statistics and Multi-Band Processing— output signals associated respectively with the Shamail Saeed,1 Harinarayanan E. V.,1 Deepen impulses constituting the input signal, the tach- Sinha,2 Anibal Ferreira2,3 es-algorithm can give access immediately to the 1ATC Labs, Noida, India new output sample and have a low latency 2ATC Labs, Chatham, NJ, USA response using vector-based optimization of the 3University of Porto, Porto, Portugal calculation. With the taches-algorithm it is easy to change (even in real time) the impulse We recently introduced a novel Automatic Noise response while running the calculation, simply Reduction (ANR) algorithm for the removal of by updating the impulse response to use it for wideband stationary/nonstationary noise from au- next samples, a task rather difficult to achieve dio. Current noise reduction techniques exhibit using FFT convolution. Real time audio demon- certain undesirable characteristics. Distortion strations using Pure Data and simple explana- and/or alteration of the audio characteristics is a tions of the taches-algorithm will be given. common problem. User intervention in identifying Convention Paper 7412 the noise profile is sometimes necessary. ANR uses a novel framework employing dominant com- 10:30 ponent subtraction and restoration and performs better than conventional techniques in subjective P15-4 Performance of Independent Component tests. Here we describe three enhancements to Analysis when Used to Separate Competing ANR. The first of these increases the level of noise Acoustic Sources in Anechoic and removal for the special case of stationary back- Reverberant Conditions—Ben Shirley, Paul ground noise. The second is a new tool for improv- Kendrick, University of Salford, Salford, Greater ing the temporal envelope coherence and yields Manchester, UK additional noise removal. The third is a multi-band A review of existing methods for independent com- processing tool for conditioning time-frequency en- ponent analysis was carried out and a series of velope for reduced listener fatigue. experiments conducted assessing the use of exist- Convention Paper 7415 ing independent component analysis (ICA) methods to separate microphone sources in varied Workshop 15 Monday, May 19 acoustic environments. Specifically the 09:00 – 10:30 Room B research looked at how effectively ICA could perform in a broadcast context using standard microphone techniques such as spaced omni and coin- PRACTICAL USES OF ACOUSTIC MODELING cident crossed cardioid pairs. Experiments were carried out in an anechoic chamber and also in a Chair: Ben Kok, Nelisse Ingenieursbureau B.V., listening room conforming to the ITU-R BS.1116-2 Eindhoven, The Netherlands standard. Results clearly indicate the limitations of ICA when performed on audio material recorded in Panelists: Peter Mapp, Peter Mapp Associates, a reverberant environment; however it was still Colchester, Essex, UK shown possible to achieve separation of signals of Henrik Møller, Akukon Oy Consulting up to 12 dB even in these conditions. Engineers, Helsinki, Finland Convention Paper 7413 Dirk Noy, WSDG, Basel, Switzerland What can be achieved by acoustic simulations for vari- 11:00 ous types of environments? The panel members will P15-5 A Cross-Platform Audio Signal Processing explain how they used software simulation in their con- Environment for Real-Time Audio Algorithm sulting and design practice. Examples will be given for Development—Mika Ristimäki,1 Matti performance spaces, small rooms (like studios), and Hämäläinen,2 Julia Turku,1 Riitta Väänänen2 harsh environments (like stadiums and tunnels). 1Nokia Research Center, Helsinki, Finland 2Nokia Research Center, Tampere, Finland Tutorial 11 Monday, May 19 09:00 – 11:00 Room L This paper presents a real-time audio algorithm development environment for experimental audio system research. The backbone of the THE GAME AUDIO PROCESS system is Pure Data audio signal processing platform, which enables flexible implementation Presenters: Mario Lavin ¯ of real-time audio systems. With the proposed Lucas Van Tol

Audio Engineering Society 124th Convention Program, 2008 Spring 31 Technical Program Andreas Varga using different bit rates, sample rates, codecs, Anton Woldhek etc. We evaluate the effect of various lossy audio encodings on the application of audio This tutorial provides an overview of the process of spectrum projection features to the automatic developing audio for a modern computer game, from ini- genre classification tasks. We show that tial conception through to final release. The session is decreases in mean classification accuracy, while presented by the audio team from Guerrilla Games who small, are statistically significant for bit-rates of are responsible for a number of PlayStation titles. The 96-kbps or lower. Also, a heterogeneous collec- talk illustrates the various aspects of game audio by con- tion of audio encodings has statistically signifi- sidering the development of a single title from the point cant decreases in mean classification accuracy of view of the audio director, sound designer, and pro- compared to a pure PCM collection. grammer. The panel will also draw on their separate and Convention Paper 7417 collective experience from other titles to contrast different approaches to game audio. The tutorial offers a detailed but technical introduction to game audio and is suitable 09:30 for newcomers and experienced professionals alike. P16-3 Loop Region Detection in Music Signals— Bee Suan Ong, Sebastian Streich, Centre for Monday, May 19 09:00 Room H Advanced Sound Technologies, Yamaha Standards Committee Meeting on SC-03-06 Digital Corporation, Japan Library and Archive Systems Spotting loops within a music recording seems to be an easy task for human listeners. Never- Monday, May 19 09:00 Room K theless it becomes highly time and effort con- Technical Committee Meeting on Acoustics suming when loop segments are to be identified and Sound Reinforcement from a large music collection. The process can be greatly facilitated with an audio editing tool Session P16 Monday, May 19 that highlights regions where loops appear and 09:30 – 11:00 Topaz Lounge suggests loop durations respectively. This paper proposes a method for computing both types of POSTERS: ANALYSIS AND SYNTHESIS information from the music signals. Our OF SOUND, PART 1 approach is based on identifying sequential and regular repetitions of tonal features. In addition, 09:30 we present a prototype implementation featuring the proposed method to facilitate the audio P16-1 A Channel Vocoder Using Wavelet Packets browsing and searching process. Finally, we dis- over a Reconfigurable Device—César Daniel cuss other possible applications of this technolo- Salvador Castañeda, Pontificia Universidad gy in the audio content description context. Católica del Perú, Lima, Peru Convention Paper 7418 A channel vocoder using wavelet packets for computer music applications is proposed. The 09:30 inputs are a modulating signal, which is choice P16-4 Music-Inspired Harmony Search Algorithm to be voice, and a carrier signal, which can be Applied to Feature Selection for Sound music or noise. The Wavelet Packets Channel Classification in Hearing Aids—Javier Amor, Vocoder transforms windowed frames of both Enrique Alexandre, Roberto Gil-Pita, Lorena signals to a symmetric multiresolution represen- Álvarez, Ester Huerta, Universidad de Alcalá, tation, mixes the envelope of the modulating sig- Alcalá, Spain nal with the carrier, and transforms back the result to the original domain. Simulations run This paper explores the application of the music- with Simulink. Real time implementations are inspired harmony search algorithm to the problem presented for Pure Data and Xilinx Virtex II Pro of feature selection for sound classification in digital FPGA. Appropriate choices of window, overlap, hearing aids. The importance of this problem is giv- wavelets, decomposition levels, and envelope en by the strong computational constraints inherent detector are presented to achieve different to the DSPs used in modern digital hearing aids. sound effects. Finally, new ideas to improve The goal of the feature selection algorithm is to transmission and compression rates in future select a subset of features in order to reduce the works are also proposed. computational complexity of the system while Convention Paper 7416 maintaining a low probability of error. A set of experiments will be performed to test the perfor- 09:30 mance of the proposed system, using a total of 74 different features. The results will be compared P16-2 The Effects of Lossy Audio Encoding on with those obtained using other widely-used algo- Genre Classification Tasks—Kurt Jacobson,1 rithms, such as a genetic algorithm, a sequential Ben Fields,2 Mark Sandler,1 Michael Casey2 search algorithm or random search. 1Queen Mary University of London, London, UK Convention Paper 7419 2Goldsmith’s College, University of London, London, UK 09:30 In large audio collections, it is common to store P16-5 Analysis of the Effects of Finite Precision in audio content using perceptual encoding. How- Sound Classifiers for Digital Hearing Aids— ever, encoding parameters may vary from col- Ester Huerta, Enrique Alexandre, Roberto lection to collection or even within a collection— Gil-Pita, Lorena Álvarez, Javier Amor, Universidad

32 Audio Engineering Society 124th Convention Program, 2008 Spring de Alcalá, Alcalá de Henares, Madrid, Spain NEW WAYS TO BROADCAST YOUR MESSAGE This paper deals with the analysis of quantization effects in an automatic sound classification Chair: Eirik Solheim, NRK Oslo, Oslo, Norway system for DSP-based hearing aids. The results Co-Chair: Karl Petermichl, ORF Radio obtained in this work will be used to find out the impact of finite accuracy determined by the digi- Eirik Solheim and Karl Petermichl will go through the lat- tal signal processor (DSP) on the users of hear- est trends on the internet. And they’ll show you case ing aids. The DSP has a finite word length that studies that include the use of YouTube, Facebook, Twit- affects the main ability of these systems: the ter, and a slightly controversial distribution method called automatic adaptation to the changing acoustic BitTorrent. environment. The goal of this work is to model a And possibly one more thing . . . quantized Neural Network-based classifier in order to compare the probability of error obtained with those nonfinite precision systems. Live Sound Seminar 6 Monday, May 19 Convention Paper 7420 09:30 – 12:00 Forum

09:30 FROM RIDER TO SHOW, PART 1: LOGISTICS P16-6 A Constructive Algorithm for Multilayer Presenter: Eberhard Müller, Neumann & Müller, Perceptrons for Speech/Non-Speech Hamburg, Germany Classification in Hearing Aids—Lorena Álvarez, Enrique Alexandre, Raúl Vicen, Lucas • About the rider Cuadra, Manuel Rosa, Universidad de Alcalá, • How to choose the correct equipment by using a soft- Alcalá de Henares, Spain ware program Constructive learning algorithms offer an attrac- • How to calculate within a given budget tive approach for the incremental construction of near-minimal neural-network architectures for pattern classification. This paper explores the Monday, May 19 10:00 Room K feasibility of using a constructive algorithm for Technical Committee Meeting on Audio Forensics multilayer perceptrons (MLPs) applied to the problem of speech/non-speech classification in Workshop 17 Monday, May 19 hearing aids. When properly designed and 11:00 – 12:30 Room B trained, MLPs are able to generate an arbitrary classification frontier with a relatively low compu- AUDIO NETWORKING FOR THE PROS tational complexity. The paper will focus on the design of a constructive algorithm for MLPs that Chair: Umberto Zanghieri, ZP Engineering srl, attempts to converge to the minimum complexity Rome, Italy network for the given problem. The results obtained will be compared with those cases in Panelists: Steve Gray, Cirrus Logic, Inc., Austin, TX, USA which the constructive algorithm is not considered. Greg Shay, Axia Audio, Cleveland, OH, USA Convention Paper 7421 Jérémie Weber, Auvitran, Meylan, France Aidan Williams, Audinate, Broadway, Australia 09:30 Several solutions are available on the market today for P16-7 Seeing the Inaudible. Descriptors Used for digital audio transfer over conventional data cabling, but Generating Objective and Reproducible Data only some of them allow usage of standard networking in Real-Time for Musical Instrument Playing equipment. This workshop presents some commercially Standard Situations—Tobias Grosshauser, available solutions (Cobranet, Livewire, Ethersound, Diemo Schwarz, Norbert Schnell, IRCAM, Paris, Dante), with specific focus on noncompressed, low-laten- France cy audio transmission for pro-audio and live applications, This paper describes a method to generate using standard IEEE 802.3 network technology. The objective and reproducible data to assist instru- main challenges of digital audio transport will be outlined, ment teaching and practicing. the method is including compatibility with common networking equip- based on using audio descriptors and their effi- ment, reliability, latency, and deployment. Typical sce- cient visualization that assist in the perception of narios will be proposed, with panelists explaining their musical parameters difficult to hear. To aid com- own approaches and solutions. parison, we defined and recorded a comprehen- sive database of positive and negative sound Exhibitor Seminar 3 Monday, May 19 examples from the violin that encompasses fre- 11:00 – 12:00 Room P quent mistakes made by students and a wide variety of playing styles. SENNHEISER ELECTRONIC CORPORATION Convention Paper 7422 Presenters: Sebastian Schmitz Gregor Zielinsky Workshop 16 Monday, May 19 MKH 800 Twin with Remote Controllable 09:30 – 11:00 Room O Directivity IS THE CHANNEL DEAD? MKH 800 Twin The MKH 800 Twin is a Doublecapsule ¯ Microphone with remote-controllable pattern for stereo

Audio Engineering Society 124th Convention Program, 2008 Spring 33 Technical Program and surround recording. Each of the capsules can be groups in order to solve a sound separation recorded separately as they have separate outputs. This problem. enables you to seamlessly adjust the pickup pattern at Convention Paper 7424 post-production stage. The presentation shows several examples of how to use this technique in stereo and sur- 11:30 round examples. P17-3 Analysis and Synthesis of Audio Vibrato Monday, May 19 11:00 Room H Using Harmonic Sinusoids—Wen Xue, Mark Standards Committee Meeting on SC-03-07 Audio Sandler, Queen Mary, University of London, Metadata London, UK This paper introduces the analysis and synthesis Monday, May 19 11:00 Room K of vibrato in music audio. The analyzer sepa- Technical Committee Meeting Advisory rates frequency modulators from their carriers on Regulations using a demodulation process. It then describes the frequency variations of a vibrato using a period-synchronized parameter set and the Session P17 Monday, May 19 accompanying amplitude variations using a 11:30 – 13:00 Topaz Lounge source-filter model, both of which can be regard- ed slow-varying. The synthesizer, on the other hand, reconstructs a vibrato from a given set of POSTERS: ANALYSIS AND SYNTHESIS parameters. Using this system we are able to OF SOUND, PART 2 retrieve specific characteristics of vibratos, or modify them to implement various audio effects. 11:30 Convention Paper 7425

P17-1 Structural Segmentation of Music Using Set 11:30 Accented Tones—Cillian Kelly, Mikel Gainza, David Dorran, Eugene Coyle, Dublin Institute P17-4 Distortion Analysis and Reduction for the of Technology, Dublin, Ireland Parametric Array—Ee-Leng Tan,1 Woon-Seng Gan,1 PeiFeng Ji,2 Jun Yang2 An approach that efficiently segments Irish Tra- 1Nanyang Technological University, Singapore ditional Music into its constituent structural seg- 2Chinese Academy of Sciences, Beijing, China ments is presented. The complexity of the segmentation process is greatly increased due to In this paper distortion analysis and reduction for melodic variation existent within this music type. the parametric array loudspeaker is being pre- In order to deal with these variations, a novel sented. The parametric loudspeaker has been method using “set accented tones” is introduced. found useful in generating a highly directional The premise is that these tones are less suscep- sound beam. However, due to the nonlinear tible to variation than all other tones. Thus, the interaction of ultrasonic wave in air, several un- location of the accented tones is estimated and desired harmonic distortions have been generat- pitch information is extracted at these specific ed. Conventional approaches in reducing the locations. Following this, a vector containing the distortion have not created satisfying solutions. pitch values is used to extract similar patterns A new approach capable of further reducing the using heuristics specific to Irish Traditional distortion has been proposed in this paper. Sev- Music. The robustness of the approach is evalu- eral simulation results are being carried out in ated using a set of commercial Irish Traditional this work to test and compare the effectiveness recordings. of this proposed solution with conventional Convention Paper 7423 approaches. Convention Paper 7426 11:30 11:30 P17-2 AnClaS3: A Blackboard-Based Cooperative Framework for Sound Separation—Antonio P17-5 Piano “Forte Pedal” Analysis and Detection Pena,1 Norberto Degara-Quintela,1 Manuel —Antony Schutz,1 Valentin Emiya,2 Dirk T. M. Sobreira-Seoane,1 Soledad Torres-Guijarro2 Slock,1 Bertrand David,2 Roland Badeau2 1Universidade de Vigo, Vigo, Spain 1EURECOM Institute, Valbonne Sophia- 2Laboratorio Oficial de Metroloxía de Galicia Antipolis, France (LOMG), Tecnópole, Ourense, Spain 2Telecom Paris Tech, Paris, France Blackboard modeling provides a great flexibility In this paper we describe some features of the in structuring complex problems and a robust Forte Pedal piano effect and propose a method adaptation to the conditions of the signal to be for detecting it through signal analysis. The analyzed, adding both bottom-up and top-down detection method is applied to single tones capabilities to the system. AnClaS3 (Analysis, recorded for this purpose. The Forte Pedal is Classification, and Synthesis for Sound Separa- found to increase the decay time of partials. In tion) is a cooperative project where five research fact, this effect dominates the behavior of the par- groups collaborate integrating algorithms and tials, in not only the duration, but also the evolu- developing new separation methods. This contri- tion. When the sustain pedal is used, a floor bution defines a blackboard-based framework noise appears for all the notes of the piano. Here, where four blackboard-based systems interact to after the analysis of some relevant characteristics integrate the expertise of independent research we provide a method based on harmonic plus

34 Audio Engineering Society 124th Convention Program, 2008 Spring noise decomposition for analyzing the residual Students are invited to sign-up for an individual meeting and decide if the pedal is pressed or not. with a distinguished mentor from the audio industry. The Convention Paper 7427 opportunity to sign up will be given at the end of the opening SDA meeting. Any remaining open spots will be posted in the student area. All students are encouraged Tutorial 12 Monday, May 19 to participate in this exciting and rewarding opportunity 11:30 – 14:00 Room L for individual discussion with industry mentors.

BROADCAST CASE STUDIES OF PRODUCTIONS Exhibitor Seminar 4 Monday, May 19 MADE IN CHALLENGING CIRCUMSTANCES 12:00 – 13:00 Room P

Presenters: Florian Camerer, ORF Vienna MINNETONKA AUDIO SOFTWARE (GERMANY) Gaute Nistov, NRK Oslo Teemu Tanskanen, YLE Helsinki Presenters: Markus Hintz A follow up from the last European AES convention in Jayson Tomlin Vienna in 2007, this workshop assembles 3 sound SurCode for Dolby E Workflow on Pro Tools, engineers each presenting an especially demanding tele- VST, and Other Platforms vision production needing tailor-made solutions for unusual challenges. An in-depth look to the nuts and bolts Dolby E has become a de facto standard in surround of these productions will be provided, from initial planning, audio even outside of broadcast facilities. Minnetonka to diverse microphone setups, signal flow, and redundan- Audio will present how the SurCode for Dolby E software cy and also including recording concepts for possible products improve existing workflows on Pro Tools, VST, postproduction. The programs are: the Nobel Peace Prize and on Minnetonka’s AudioTools Audio Workflow Engine Concert in Oslo, the Eurovision Song Contest in Helsinki, (AWE), helping studios to be productive and efficient. and the New Year’s Concert 2008 in Vienna.

Tutorial 13 Monday, May 19 Session P18 Monday, May 19 11:30 – 13:30 Room N101 12:30 – 17:00 Room C-D

ELECTROACOUSTIC MEASUREMENTS, PART 2 ROOM AND ARCHITECTURAL ACOUSTICS AND SOUND REINFORCEMENT Presenter: Christopher J. Struck, CJS Labs, San Francisco, CA, USA Chair: Jan Voetmann, DELTA Acoustics, Hoersholm, Denmark This tutorial focuses on applications of electroacoustic measurement methods, instrumentation, and data interpretation as well as practical information on how to 12:30 perform appropriate tests. Linear system analysis and P18-1 Diffusing Boundary Implementations in the 2- alternative measurement methods are examined. The D Digital Waveguide Mesh—Simon Shelley,1 topic of simulated free field measurements is treated in Damian Murphy2 detail. Non-linearity and distortion measurements and 1Technische Universiteit Eindhoven, Eindhoven, causes are described. Last, a number of advanced tests The Netherlands are introduced. This tutorial is intended to enable the 2York University, Heslington, York, UK participants to perform accurate audio and electroacoustic tests and provide them with the necessary tools The digital waveguide mesh is a wave-based to understand and correctly interpret the results. time-domain approach to the simulation of sound wave propagation in an acoustic system. The Tutorial 14 Monday, May 19 implementation of diffuse reflection is an impor- 11:30 – 13:00 Room O tant consideration in such an application, as the presence of diffuse reflection has a significant effect on an acoustic environment. The scatter- SMALL ROOM ACOUSTICS ing effect of diffuse boundaries on reflected sounds, both in simulation and the real world, Presenter: Ben Kok, Nelissen Ingenieursbureau B.V., can be described using a technique that results Eindhoven, The Netherlands in the formulation of frequency dependent diffu- Acoustic basics of small rooms will be discussed. Specif- sion coefficients. In this paper a number of differ- ic issues related to the size of the room (room-modes) ent approaches to modeling diffuse reflection in will be addressed. Absorption, reflection, diffraction, dif- a 2-D digital waveguide mesh are presented as fusion and how to use it, and low frequency treatment. well as a detailed analysis and comparison of Although this will not be a studio design class, specifics the local scattering effect of the diffuse boundary and differences of recording rooms and control rooms models using this technique. will be identified, including considerations for loudspeak- Convention Paper 7428 er and microphone placement. 13:00 Student Event/Career Development P18-2 RenderAIR—Room Acoustics Simulation MENTORING SESSIONS Using a Hybrid Digital Waveguide Mesh— Monday, May 19, 12:00 – 14:00 Damian Murphy,1 Mark Beeson,1 Simon Topaz Lounge Shelley,2 Alex Southern,1 Alastair Moore1 1University of York, Heslington, York, UK ¯

Audio Engineering Society 124th Convention Program, 2008 Spring 35 Technical Program 2Eindhoven University of Technology, ly with and without a total of eight different Eindhoven, The Netherlands absorbing devices. Results are compared and conclusions are presented. The digital waveguide mesh (DWM) is a numeri- Convention Paper 7431 cal simulation technique used to model signal propagation through a regular grid of spatio-temporal sampling points, and has been demon- 14:30 strated as appropriate for modeling the acoustics P18-5 Modeling Frequency-Dependent Boundaries of an enclosed space, particularly at low fre- as Digital Impedance Filters in FDTD and quencies. The RenderAIR DWM application K-DWM Room Acoustic Simulations—Konrad allows intuitive definition of parameters associat- Kowalczyk, Maarten van Walstijn, Queen’s ed with geometry, boundary surface, and University Belfast, Belfast, Northern Ireland, UK source/receiver parameters required to generate spatially encoded Room Impulse Responses This paper presents a new method for modeling (RIRs). In this paper the expectations and limita- frequency-dependent boundaries in finite differ- tions of DWM-based room acoustics modeling ence time domain (FDTD) and Kirchhoff variable are explored through the use of the RenderAIR digital waveguide mesh (K-DWM) room application in a number of situations. ISO3382 acoustics simulations. The proposed approach metrics are used as the main benchmark for the allows direct incorporation of a digital impedance results obtained, which compare well with both filter (DIF) in the multi-dimensional (i.e., 2-D or real-world measurements and more traditional 3-D) FDTD boundary model of a locally reacting geometric acoustic approaches. surface. An explicit boundary update equation is Convention Paper 7429 obtained by carefully constructing a suitable recursive formulation. The method is analyzed in terms of pressure wave reflectance for different 13:30 wall impedance filters and angles of incidence. P18-3 Volumetric Diffusers—Richard Hughes, Jamie Results obtained from numerical experiments A. S. Angus, Trevor Cox, Olga Umnova, confirm the high accuracy of the proposed digital University of Salford, Salford, Greater impedance filter boundary model, the reflectance Manchester, UK of which closely matches locally reacting surface (LRS) theory. Furthermore, a numerical bound- Although many types of diffusers have been pro- ary analysis (NBA) formula is provided as a posed, they are predominantly surface treat- technique for analytic evaluation of the numeri- ment. This paper places the diffuser in the vol- cal reflectance of the proposed digital imped- ume of the room rather than on the surfaces, ance filter boundary formulation. forming a volume-based diffuser. In particular, Convention Paper 7430 we examine suitable sequences for their imple- Winner of the Student Paper Award mentation. We also consider suitable metric’s to evaluate their performance. At first single layer volumetric diffusers are examined, and then mul- 15:00 ti-layer volumetric diffusers are investigated. In P18-6 Loudspeaker Time Alignment Using Live particular, the effects of varying the spacing, and Sound Measurements—Wolfgang Ahnert, number of layers, is more closely examined. The Stefan Feistel, Thorsten Maier, Alexandru Radu Boundary Element Method (BEM) model is used Miron, Ahnert Feistel Media Group, Berlin, to gain accurate predictions of the diffuser’s per- Germany formance. Finally, we demonstrate a diffusion structure that has a similar performance to that The authors previously introduced the measure- of a Primitive Root Diffuser (PRD). ment software EASERA SysTune, which can be Convention Paper 7432 used for measurements with live music and speech signals. In this paper we discuss specifically the use of real-time measurements for the 14:00 time alignment of loudspeaker arrays and distrib- P18-4 Commercial Low Frequency Absorbers uted systems and for the optimal adjustment of —A Comparative Study—Gabriel Hauser,1 their phase relationships. Being capable of deriv- Dirk Noy,1 John Storyk2 ing impulse responses of up to 12 seconds 1Walters-Storyk Design Group, Basel, length, the measuring process with EASERA Switzerland SysTune is simpler and more accurate as the 2Walters-Storyk Design Group, Highland, NY, USA real-time function provides a more immediate view on the tuning process. Because measure- This paper ties in to a previous Convention ments can be performed with standard stimulus Paper by the same authors (AES 115th Conven- signals as well as with external speech and tion, 2003, #5944) and presents a current set of music signals, fine-tuning loudspeaker settings commercially available passive and active low becomes possible even during the rehearsal frequency absorbing devices. One item in partic- time of the musicians. Required measurement ular is of an experimental nature—a wood box conditions and limitations are given loaded with conventional membrane loudspeak- Convention Paper 7433 ers. These are not connected to an amplifier, but to a variety of different passive electronics networks (parallel, serial). Reproducible acoustical 15:30 measurements have been taken in a completely P18-7 INR as an Estimator for the Decay Range untreated rectangular concrete room, sequential- of Room Acoustic Impulse Responses—

36 Audio Engineering Society 124th Convention Program, 2008 Spring Constant Hak,1 Jan Hak,2 Remy Wenmaekers3 talker are discussed as is the need to the need 1Eindhoven University of Technology, to adequately assess the effects of both rever- Eindhoven, The Netherlands beration and noise. The findings discussed in 2Acoustics Engineering, Boxmeer, the paper are also relevant to the installation and The Netherlands testing of classroom “sound field” systems and 3Level Acoustics, Eindhoven, The Netherlands also boardroom type reinforcement systems and conferencing / teleconferencing systems. A room acoustic impulse response can be used Convention Paper 7436 to derive the reverberation time and other parameters. To this end a certain minimum energy decay range or effective signal to noise ratio is required, which relates to the difference between Session P19 Monday, May 19 the integrated signal level and the noise level. 12:30 – 15:30 Room E-F An impulse response parameter called INR is presented as an estimator for the decay range SOFTWARE, INSTRUMENTATION, and shown to be a useful qualifier in practical AND MEASUREMENT measurements. Convention Paper 7434 Chair: John Vanderkooy, University of Waterloo, [Paper presented by Remy Wenmaekers] Waterloo, Ontario, Canada

16:00 12:30 P18-8 Musical-Inspired Features for Automatic P19-1 Graphical Control of a Parametric Equalizer Sound Classification in Digital Hearing —Jörn Loviscach, Hochschule Bremen Aids—Pedro Vera-Candeas,1 Francisco J. (University of Applied Sciences), Bremen, Germany Cañadas-Quesada,1 Enrique Alexandre,2 Manuel Rosa2 Graphic equalizers allow the user to define a fil- 1University of Jaén, Linares, Jaén, Spain ter’s magnitude response virtually free of restric- 2University of Alcalá, Alcalá, Spain tions. Parametric equalizers are much more limited. However, they offer some vital advan- This paper proposes the use of some musical- tages over graphic equalizers, such as consum- inspired features for the automatic classification of ing less computational power and operating sounds in digital hearing aids. This kind of appli- minimally invasively with naturally soft magni- cation is characterized by very strong constraints tude and phase responses. This paper aims at in terms of computational complexity. The pro- combining the best of both worlds. It presents a posed features are based on fundamental fre- range of methods to control a digital parametric quency detection and exhibit a low computational equalizer graphically through a curve or a collec- complexity while providing good results in terms tion of anchor points. While the user is editing of probability of correct classification. The perfor- the graphical input, an optimization process runs mance of the system will be tested using a 1-NN in the background and adjusts the equalizer’s classifier, the goal being to distinguish among parameters to reflect the input. In addition, the speech, noise, and music. For the experiments a number of bands and their type (shelving/peak) sound database, obtained using a hearing aid can be adjusted automatically to produce a sim- simulator, will be used. ple solution. Convention Paper 7435 Convention Paper 7437

16:30 13:00 P18-9 Assessing the Potential Intelligibility of P19-2 Audio Software Development—An Audio Assistive Audio Systems for the Hard of Quality Perspective—Jonas Ekeroot, Jan Berg, Hearing and Other Users—Peter Mapp, Luleå University of Technology, Piteå, Sweden Peter Mapp Associates, Colchester, Essex, UK When developing audio applications, different Around 14% of the European population suffer choices on software implementation aspects from a noticeable degree of hearing loss and influence the total audio software signal path and would benefit from some form of hearing assis- can be of importance from an audio quality per- tance or deaf aid. Recent DDA legislation and spective. The field is not well documented in the requirements mean that many more hearing literature. A study was carried out aiming at iden- assistive systems are being installed—yet there tifying relevant questions that must be consid- is evidence to suggest that many of these sys- ered. The general development perspective was tems fail to perform adequately and provide the on audio software written in C++ to be run on benefit expected. This paper reports on the general purpose CPUs. A research review, com- results of some trial acoustic performance test- prising literature from different fields such as ing of such systems. In particular the effects of audio engineering, computer science, and soft- system microphone type, distance, and location ware engineering was conducted to summarize are shown to have a significant effect on the and integrate an overview of the field. The result resultant performance. The potential of using the can be viewed as a map of questions for future Sound Transmission Index (STI) and in particu- research activities, consisting of further literature lar STIPa, for carrying out installation surveys studies and experiments with software prototypes. has been investigated, and a number of practical Convention Paper 7438 problems are highlighted. The requirements for a suitable acoustic test source to mimic a human ¯

Audio Engineering Society 124th Convention Program, 2008 Spring 37 Technical Program 13:30 This paper presents a Multiple Input Single Out- put (MISO) nonlinear model in combination with P19-3 Multi Carrier Modulator for Switch-Mode 1,2 sine-sweep signals as a method for nonlinear Audio Power Amplifiers—Arnold Knott, system identification. The method is used for Gerhard Pfaffinger,1 Michael A. E. Andersen2 1 identification of loudspeaker nonlinearities and Harman/Becker Automotive Systems GmbH, can be applied to nonlinearities of any audio Straubing, Germany 2 components. It extends the method based on Technical University of Denmark, Lyngby, nonlinear convolution presented by Farina, pro- Denmark viding a nonlinear model that allows simulation While switch-mode audio power amplifiers allow of the identified nonlinear system. The MISO small implementations and high output power model consists of a parallel combination of non- levels due to their high power efficiency, they are linear branches containing linear filters and very well known for creating electromagnetic memory-less power-law distortion functions. interference (EMI) with other electronic equip- Once the harmonic distortion components are ment, in particular radio receivers. Lowering the identified by the method of Farina, the linear fil- EMI of switch-mode audio power amplifiers while ters of the MISO model can be derived. The keeping the performance measures to excellent practical application of the method is demon- levels is, therefore, of high general interest. A strated on a loudspeaker. modulator utilizing multiple carrier signals to Convention Paper 7441 generate a two level pulse train will be shown in this paper. The performance of the modulator 15:00 will be compared in simulation to existing modulation topologies. The lower EMI as well as the P19-6 Junction Identification Using Acoustic preserved audio performance will be shown in Reflectometry—Adam Kestian, Agnieszka simulation as well as measurement results of a Roginska, New York University, New York, NY, prototype. USA Convention Paper 7439 Acoustic reflectometry is a non-invasive, time- 14:00 domain method of identifying the geometry of an acoustical space. A sound pulse is injected into P19-4 A Comparison of Theoretical, Simulated, a space and the resulting impulse response and Experimental Results Concerning the details particular changes of impedance. In the Stability of Sigma Delta Modulators—Georgi present paper acoustic reflectometry is utilized Tsenov,1 Valeri Mladenov,1 Joshua D. Reiss2 to identify scattering junctions of geometric 1Technical University of Sofia, Sofia, Bulgaria spaces. Most notably, the four most common 2Queen Mary, University of London, London, UK types of scattering junctions are identified: a cross-sectional increase, cross-sectional Sigma delta modulation is a popular form of decrease, L-intersection, and T-intersection. audio analog-to-digital and digital-to-analog con- Convention Paper 7442 version, but suffers from stability problems for many designs and many input signals. A general theory of stability in sigma delta modulators has been developed that predicts the stability of a Workshop 18 Monday, May 19 high order one-bit sigma-delta modulator (SDM) 12:30 – 14:30 Room B under a variety of designs. In this paper the theoretical approach to stability as it applies to AUDIO OVER IP SYSTEMS FOR REPORTERS boundedness of states is explained. Several low pass SDM designs are developed that are Chair: Lars Jonsson, Swedish Radio intended for audio analog to digital conversion, and predicted results for stability of these Panelists: Lee Chaundy, BBC, UK designs are given. Stability is examined both in Gerald List, Tranxtel, Germany terms of the maximum allowable DC input ampli- Ulrich Loebbert, WDR, Germany tude and the theoretical sufficient conditions for stable behavior. Theoretical results are com- This workshop will present case studies on the use of pared with simulated results, and where possi- audio over IP systems for reporters out in the field. Audio ble, with experimental results from a realization over IP units have gradually become an alternative to of a third order SDM with adjustable parameters. ISDN and other means of reporting. Sometimes well- Practical observations are then made concern- managed IP-networks can be used, or the unreliable but ing the effect of noiseshaping, pole/zero place- widely spread Internet. Wireless alternatives for reporters ment, and cut-off frequency on the stability. will also be discussed and evaluated. Examples of using Convention Paper 7440 the SIP protocol for connecting between AoIP units and VoIP telephones will be discussed. 14:30 Live Sound Seminar 7 Monday, May 19 P19-5 A New Method for Identification of Nonlinear 13:00 – 17:00 Forum Systems Using MISO Model with Swept-Sine Technique: Application to Loudspeaker Analysis—Antonín Novák,1,2 Laurent Simon,2 FROM RIDER TO SHOW, PART 2: LIVE MIXING Pierrick Lotton,2 Frantisek Kadlec1 EVENT 1Czech Technical University in Prague, Prague, Czech Republic Presenter: Eberhard Müller, Neumann & Müller, 2Université du Main, Le Mans, France Hamburg, Germany

38 Audio Engineering Society 124th Convention Program, 2008 Spring • Stage set up, calibrating Session P20 Monday, May 19 14:00 – 15:30 Topaz Lounge • Microphone set up and sound check • Show POSTERS: PSYCHOACOUSTICS, PERCEPTION, AND LISTENING TESTS This session includes practical demonstrations with a jazz combo (piano, percussion, guitar, bass, vocal) and is presented in combination with an OB-Van. Attendees 14:00 can move between auditorium and OB-van to calculate P20-1 Loss of Subjective Localization Cues in Virtual the different sound patterns. Acoustic Opening—Elena Blanco-Martín, Exhibitor Seminar 5 Monday, May 19 Francisco Javier Casajus-Quiros, Juan Jose 13:00 – 14:00 Room P Gomez-Alfageme, Luis I. Ortiz-Berenguer, Universidad Politécnica de Madrid, Madrid, Spain ODEON The reproduced sound event quality is very Presenters: Claus Lynge Christensen important in a WFS configuration that is used for Jens Holger Rindel an acoustic opening. One way of checking the subjective quality perceived by a listener is the ODEON Room Acoustic Simulations ITU R. BS.1387 “Method for objective measurements of perceived audio quality,” but this ODEON is the state-of-the-art software for room acoustic method does not provide information about the simulations and auralizations, used in major concert hall listener’s ability to localize the sound. A Matlab and theater projects. The seminar will show examples of application has been implemented (SEL, Sound modeling including CAD-import, source directivity, array Event Localization) simulating an acoustic open- modelling, tools for reflection analysis, and ODEON’s ing configuration. The number of microphones multichannel high fidelity auralization. and loudspeakers in the arrays is selectable, just as the sound source position, the gap between Monday, May 19 13:00 Room H array transducers and the listener position. This Standards Committee Meeting on SC-02-12 Audio simulation has been verified against a real con- Applications of IEEE 1394 figuration of acoustic opening. Moreover, the loss of localization cues has been analyzed with Monday, May 19 13:00 Room K different multichannel codifications. Technical Committee Meeting on Multichannel Convention Paper 7443 and Binaural Audio Technologies 14:00 Student Event/Career Development EDUCATION FORUM PANEL P20-2 Effect of Interaural Differences on Loudness Monday, May 19, 13:30– 15:30 of Narrowband Noise Bursts—Toni Hirvonen, Room P Ville Pulkki, Helsinki University of Technology, Espoo, Finland Moderator: Jason Corey, University of Michigan, MI, This paper investigates the effects of interaural USA time and level differences (ITDs and ILDs, Panelists: Mark Drews, University of Stavanger, Norway respectively) on loudness. Dichotic samples con- Theresa Leonard, The Banff Centre, Canada taining various amounts of interaural differences Konrad Strauss, Indiana University, USA were compared to a diotic reference. The sub- Cornelis Van der Gragt, The Hague Royal jects adjusted the relative threshold gain of the Conservatoire, The Netherlands test sample using a two-alternative, forced Guiding and Evaluating Student Learning choice adaptive procedure (2AFC). The test sig- in Audio Engineering nals were Gaussian noise samples with a bandwidth of one critical band and center frequencies As educators we are concerned with aiding and facilitating of 150, 600, and 2400 Hz. The results imply that students' acquisition of knowledge and experience. We ILD is prominently responsible for changes in also usually need to find some method of evaluating directional loudness, which is in agreement with whether students have gained the appropriate skill level present binaural loudness models that consider and knowledge base for a given course. As audio is some- only ILD. The experiments revealed significant what unique in terms of how students increase their knowl- individual differences between subjects even edge of the field, we as educators are faced with unique when matching two identical signals. challenges as well. During the discussion the panel will Convention Paper 7444 consider questions such as: What are the best ways to evaluate how well students are learning audio engineer- 14:00 ing? How should lectures, labs, and projects be structured to facilitate student learning in this field? How should we P20-3 Perception of Movements of a Focused best deal with differences in learning among students in a Sound Generated with a Linear Loudspeaker class? How much should curricula change as technology Array System—Ichiki Manon,1 Daiki Sato,1 and the industry changes? Tomoaki Tanno,1 Kaoru Ashihara,2 Shogo Kiryu1 1Musashi Institute of Technology, Setagaya-ku, Tokyo, Japan 2National Institute of Advanced Science and Technology, Tsukuba, Japan ¯ A special loudspeaker array system was

Audio Engineering Society 124th Convention Program, 2008 Spring 39 Technical Program developed for an experiment on perception of 14:00 movements of a focused sound. The spatial patterns of the sound pressure level for the focused P20-6 How to Widen the Sweet Spot in Monitoring sounds were measured. The patterns were 5.1—Julien Bassères, Patrick Thevenot, Taylor improved compared to the previous preliminary Made System, Nangis, France experiment using commercial devices. A Generally speaking, sound reproduction tends to psychoacoustic experiment on perception of achieve the widest sweet spot. But it's seldom movements of the focused sound was conduct- realized and more than that, the restricted sweet ed using the developed system. spot has become rather usual and well accepted Convention Paper 7445 by the audio community. This paper proposes to find a new approach in order to get a wider 14:00 sweet spot, up to a certain extent, in multichannel. By optimizing the directivity of each loud- P20-4 Subjective Evaluation for Music Recording speaker in order to compensate the position of Positions in a Coherent Region of a the listener, this method aims at creating a Reverberant Field—Yoshifumi Hara,1 Hiroaki 2 3 1 coherent and homogeneous acoustic field. Spe- Nomura, Mikio Tohyama, Kazunori Miyoshi cial care will be given to the directivity pattern 1Kogakuin University, Shinjuku-ku, Tokyo, Japan 2 (amplitude and phase) of the loudspeaker system. Kure College of Technology, Kure-city, Convention Paper 7448 Hiroshima, Japan 3Waseda University, Shinjuku, Tokyo, Japan 14:00 In this paper we describe the most preferable P20-7 Auditory Modeling via Frequency Warped frequency characteristics of the early reflections Transforms—Alexey Petrovsky,1 Marek for music recording positions. We recorded short Parfieniuk,2 Adam Borowicz,2 Alexander passages from two pieces of music (Handel, Petrovsky1,2 “Water Music Suite” and Brahms, “Symphony 1Belarusian State University of Informatics and No. 4”) at various distances from a sound source Radioelectronics, Minsk, Belarus in a coherent region in a reverberation chamber. 2Bialystok Technical University, Bialystok, Poland Subjects evaluated the preference and the subjective loudness through headphones under the The goal of this paper is to show and compare diotic condition by paired comparison tests. As a four different versions of auditory modeling result, we found that the most preferable dis- based on frequency warped transforms: bark- tance indicated the distance where the loudness scaled wavelet packet decomposition, bark- became maximum. The preferable recording scaled adapted wavelet packet decomposition, condition could be also characterized by narrow- warped discrete Fourier transform, and four- band envelope spectrum analysis. band wavelets paraunitary filter bank, useful for Convention Paper 7446 perceptual audio coding, speech enhancement, and parametric audio coding matching pursuit procedure based on the psychoacoustic optimized wavelet packet dictionary. A practical 14:00 implementation of the audio signal processing based on the given auditory modeling approach- P20-5 Efficient Individualization of HRTF Using es are in details considered and analyzed from Critical-Band-Based Spectral Cues Control— positions: depth of a compression, perceptual 1 2 Yoomi Hur, Young-Cheol Park, perception, a structural realizability, an opportu- 1 3 Dae-Hee Youn, Seok-Pil Lee nity to build embedded systems. 1 Yonsei University, Seoul, Korea Convention Paper 7449 2Yonsei University, Wonju, Korea 3Korea Electronics Technology Institute, 14:00 Bundang-Gu, Sungnam-Si, Korea P20-8 The Role of Spectral Features in Sound Recently, 3-D audio technologies have been Localization—Daniela Toledo, Henrik Møller, commonly implemented through headphones. A Aalborg University, Aalborg, Denmark major problem of the headphone-based 3-D Spectral components of head-related transfer audio is in-the-head localization, which occurs functions (HRTFs) are highly dependent on the due to the inaccurate head-related transfer func- anthropometric characteristics of subjects. In the tion (HRTF). Since the individual measurements low frequency range, a common structure is of HRTFs are impractical, there have been sev- often found in HRTFs from different subjects. eral researches for HRTF customization. In this However, individual differences are seen at high paper we propose an efficient method of cus- frequencies. In binaural synthesis with non-indi- tomizing HRTFs. In the proposed method, spec- vidual HRTFs, localization errors occur if the tral notches and envelopes are controlled based spectral characteristics of the directional filters on a critical-band rate. Thus, the structure of the used do not match the individual characteristic of proposed algorithm is much simpler than that of the listener. This investigation is focused on the previous methods, but still effective. The spectral characteristics of HRTFs that are proposed method was evaluated in the problem relevant as localization cues and how to para- of externalization, and the results showed that meterize them. This is done by cross-matching the customized HRTF using the proposed individual and non-individual HRTFs from differ- method could greatly improve the externalization ent subjects according to the results of localiza- performance. tion experiments. Convention Paper 7447 Convention Paper 7450

40 Audio Engineering Society 124th Convention Program, 2008 Spring 14:00 fessional digital repositories. The time slot left for this transfer is estimated to be 20 years only. P20-9 Multichannel Loudness Listening Test—Ian The tutorial describes the global situation and surveys Dash,1 Luis Miranda,2 Densil Cabrera2 1 standards, guidelines, and recommended practices relat- Australian Broadcasting Corporation, Sydney, ed to principles and practical aspects of sound archiving, NSW, Australia 2 issued by AES, IASA (International Association of Sound University of Sydney, Sydney, NSW, Australia and Audiovisual Archives), and UNESCO. As part of ongoing research for ITU Recommen- dation BS.1770 Algorithms to measure audio Historical Event programme loudness and true-peak audio level, VINTAGE LOUDSPEAKERS listening tests were conducted using a standard Monday, May 19, 14:00 – 15:00 five-channel geometry in a standard listening Room N room to confirm the channel gains and the spectral weightings for equal loudness contribution. Presenter: Garry Margolis Most ITU-related work to date has used broadcast program as a test signal. In this test, octave An informal review of recording studio monitoring prac- band noise was used as a test signal. Twenty- tices in North America from the 1950s through the 1970s seven listeners participated. Results were ana- as seen and heard by a recording engineer who joined a lyzed for statistical consistency as well as for loudspeaker company and learned how studio monitors average and variance. Agreement between the really work. test results and various broadband loudness models, including ITU-R BS.1770, is examined. Convention Paper 7451 Monday, May 19 14:00 Room K Technical Committee Meeting on Automotive Audio

Master Class 2 Monday, May 19 Workshop 19 Monday, May 19 14:30 – 16:00 Room B 14:00 – 15:30 Room L CONCEPTS IN SOUND QUALITY BRINGING DYNAMICS BACK INTO THE MIX Presenter: Jens Blauert, Ruhr-Universitaet Bochum, Chair: Ronald Prent, Galaxy Studios Bochum, Germany

Panelists: Thomas Lund, TC Electronic A/S The lecture presents a fundamental consideration of the Darcy Proper, Galaxy Studios nature of sound quality and offers ways for structuring different aspects of it. It aims at making audio engineers Pop music and the fight for level—this is an ongoing bat- more aware of the various components and complex tle of even a hint of dynamics. Peak measurement and processes involved in the formation of sound-quality normalization has caused the fight for “who is the loudest” judgments. Thinking of sound quality means thinking of and with it intersample peaks well beyond 0dBFS. Severe percepts, which brings subjectivity to the fore. To deal distortion and listener fatigue is the result. The switch of a with subjectivity scientifically, the first part of the lecture few broadcasters from peak normalization to loudness will present a line of epistemological arguments that is normalization is the straw on which proponents of a stringently based on actual perception and not on fiction. dynamic, lively, and “breathing” mix cling to in order to In this way, the place and role of subjectivity in acoustics bring this topic center-stage again. Key figures from the and audio engineering will be determined. To further elu- mixing and mastering side as well as from the manufac- cidate the formation process of sound quality, the second turers and like-minded initiatives will discuss ways out of part of the talk starts with the introduction of the charac- the loudness race, good practices, spreading the word, ter of sounds and then moves on to different aspects of etc., and will illustrate their points through examples. sound quality, starting from sound quality (as such) through sound-transmission quality and auditory-scene Tutorial 15 Monday, May 19 quality to product-sound quality. A generalized, system- 14:00 – 15:30 Room N101 oriented approach toward sound-quality evaluation is thereupon presented. Finally, the third part of the talk SOUND ARCHIVING: STANDARDS RELATED deals exemplarily with sound quality in spaces for musi- TO THE LONG-TERM PRESERVATION cal performances. The question of proper references OF A SOUND RECORDING turns out to be crucial for any further analysis. Conse- quently, efforts are taken to explore and assess these Presenter: Dietrich Schüller, Phonogrammarchiv, references systematically. To this end, a hierarchical Austrian Academy of Sciences, Vienna, order for references, based on the amount of abstraction, Austria is proposed and discussed in some detail. Issues like typicalness, functional adequacy, aesthetic form, and tra- The stocks of sound recordings held in repositories dition are touched upon. worldwide are estimated to amount to 100 million hours. Even if we want to keep only a fraction of this legacy, Historical Event logistical and financial challenges to preserve these doc- COMPACT CASSETTE uments of artistry, history, and cultural diversity are of Monday, May 19, 15:00 – 16:00 significant dimensions. Original carriers, whether analog Room N or digital, are prone to decay, and—more importantly— replay equipment becomes increasingly unavailable for Presenter: Willem Andriessen those holdings that have not yet been transferred to pro- ¯

Audio Engineering Society 124th Convention Program, 2008 Spring 41 Technical Program The “Compact Cassette,” an incredible success story with Bergweiler,2 Manfred Neumann,2 Stefan an incredible sudden end. The technical and market envi- Holzhäuser,2 Andreas Walther1 ronment in the late fifties and early sixties, which led to its 1Fraunhofer Institute for Integrated Circuits IIS, birth, its technical and product development over the years, Erlangen, Germany its impact on magnetic tape development, its “explosive” 2Lear Corporation GmbH, Kronach, Germany marketing development, its rivals, such as DC international, Elcassette, Microcassette, DAT, and DCC, and at last its Audio systems with high quality sound reproduc- ultimate killing factor: the total integrated digitizing of private tion capabilities are becoming more and more and personal audio consumption,thanks to CDR for down- popular in the car. The need to create a pleasant loading and MP3 for “ideal” portability and mobility will be sound field has lead to an increased number of discussed by Willen Andriesse, a retired tape specialist at loudspeakers combined with digital signal pro- BASF. cessing. To benefit from the advantages of surround sound reproduction also for two-channel legacy content an upmixing algorithm is Exhibitor Seminar 6 Monday, May 19 required. In this paper challenges and require- 15:00 – 16:00 Room P ments for high quality surround sound reproduction and upmixing are first introduced separately TRINNOV AUDIO (FRANCE) and then discussed jointly with the specific focus on the automotive environment. Finally a test Presenters: Arnaud Laborie method for the evaluation of different upmixing Felipe Avila-Reyes algorithms in the car is suggested. Convention Paper 7452 Digital Room Correction for the Control Room The main topic of this presentation is the challenge of 16:00 achieving consistent audio monitoring in the control room and throughout the production chain. Is digital room cor- P21-2 A General Approach to Methods for rection relevant to professional control rooms? How can Loudspeaker Array Synthesis—Juan Miguel it help to correct acoustical/ monitoring problems? What Navarro Ruiz, Universidad Católica San Antonio are its limitations? Real world examples and a short de Murcia, Guadalupe, Murcia, Spain demonstration are provided. Loudspeaker arrays are often used as sound reinforcement in large concert halls and outdoor events to provide increased directivity. Contrary Monday, May 19 15:00 Room K to what happens in loudspeaker systems, there Technical Committee Meeting on Transmission is an entrenched theory in antenna array synthe- and Broadcasting sis, which has been used extensively over the past few years. This paper focuses on dis- Student Event/Career Development cussing several consolidated antenna array syn- RECORDING COMPETITION—SURROUND thesis methods. Then, a simulation software is Monday, May 19, 15:30 – 18:30 implemented to show pros and cons of using on Room L loudspeaker arrays. Finally, an efficient synthesis method is proposed to achieve the required The Student Recording Competition is a highlight at each characteristics. convention. A distinguished panel of judges participates Convention Paper 7453 in critiquing finalists in each of the categories in an interactive presentation during the convention. Student members can submit stereo and surround recordings in the 16:00 categories classical, jazz, folk/world music, and pop/rock. P21-3 On Large Multiactuator Panels for Wave Field Meritorious awards will be presented at the closing Stu- Synthesis Applications—Basilio Pueo,1 José dent Delegate Assembly Meeting on Tuesday. Escolano,2 José Javier López,3 Germán Ramos3 Judges: 1University of Alicante, Alicante, Spain Sound for Picture (15:30–16:30): Florian Camerer, 2University of Jaén, Linares (Jaén), Spain Bernhard Maisch, Wilfried van Baelen 3Technical University of Valencia, Valencia, Spain Classical (16:30–17:30): David Griesinger, Jared Sacks, Cornelis Van der Gragt Wave Field Synthesis (WFS) is a spatial sound Non-Classical (17:30–18:30): Ronald Prent, Bosse, rendering technique that generates a true sound Ternstrom, Wilfried van Baelen field using loudspeaker arrays. Multiactuator Panels (MAPs) are an alternative technology to Recording competition sponsors: Schoeps, PMC, the dynamic piston loudspeakers, based on the Sennheiser, Neumann distribute mode operation. Because of its low visual profile and negligible vibration of the panel, MAPs are very suitable for WFS reproduc- Session P21 Monday, May 19 tion. However, the size of current prototypes 16:00 – 17:30 Topaz Lounge does not allow its use for real immersing environments in which the loudspeaker must be POSTERS: MULTICHANNEL SOUND integrated as walls or as projection screens. In addition, the extra area of a large panel can be 16:00 used to accommodate extra exciters with which to generate sound fields at another elevation lev- P21-1 Challenges in Reproduction and Evaluation el. In this paper a very large MAP prototype is of Upmixed Audio in an Automotive 1 presented that has been designed and built to Environment—Oliver Hellmuth, Steffen fulfill the requirements of immersive audio appli-

42 Audio Engineering Society 122nd Convention Program, 2007 Spring cations. It represents a step forward in the appli- Tutorial 17 Monday, May 19 cations of MAPs for immersing scenarios. The 16:00 – 18:00 Room E-F panel size enhances its acoustic behavior in the low frequency range. Also, it can be employed MUSIC-INDUCED HEARING DISORDERS for relatively large projection screens for video- conferencing and for virtual reality. Presenter: Jan de Laat, LUMC, Leiden, The Convention Paper 7454 Netherlands

16:00 Frequent exposure to loud sounds often causes hearing damage. Recent studies in Sweden demonstrated that P21-4 Temporal Changes of Psychological 74% of all musicians complain of hearing disorders such Impressions Regarding Microphone Arrays as: hearing loss, tinnitus (whistling or noise in the ear), for Multichannel Recording—Toru Kamekawa, hyperacusis (suffering from sounds that are too loud), Atsushi Marui, Tokyo University of the Arts, distortion (attentive) and diplacusis (perception of differ- Adachi-ku, Tokyo, Japan ent pitches on the left and right sides). It appears that the Microphone technique for surround sound sound volume of concerts has gradually become louder recording of an orchestra is discussed. Seven and louder over the last decades, and this corresponds types of surround microphone sets recorded in a to the extent of hearing problems in musicians, even concert hall were compared in subjective listen- attended by unfitness for work. ing test on attributes such as powerfulness and In The Netherlands these troubles are becoming rec- spaciousness using a method inspired by ognized little by little. A few years ago the government MUSHRA (MUltiple Stimuli with Hidden Refer- (Department of Social Affairs) established an agreement ence and Anchor). To minimize temporal change with the employers (administrations of professional sym- in music, Phase Randomized Signal (PRS) was phony orchestras) and employees (union of musicians) proposed. From the average score of the listen- aimed to manage these problems. Attention and consid- ing test, the impression difference between eration will be focused on awareness, professional infor- original source and PRS was found in some mation (even at early ages, starting at music colleges), microphone arrays consisting of directional prevention of and protection from hearing loss (not only microphones at some pieces. It means that the by hearing protectors such as personal ear moulds), and impression of these arrays depend on temporal rehabilitation of hearing handicaps in musicians. changes in music. The data from the listening In a research project with 259 musicians of three pro- test between the original source and PRS fessional symphony orchestras in Amsterdam, we com- showed that impressions of powerfulness had bined results of the outcomes of questionnaires, pure slightly higher correlation. The relations of the tone audiograms, speech audiometry, transient and dis- physical factors of each array were also com- tortion-product oto-acoustic emissions, tinnitus matching, pared, such as SC (Spectral Centroid), LFC and frequency selectivity measures in order to achieve (Lateral Fraction Coefficient), and S/O more insight in the individual hearing sensitivity related to (Side/Omni Ratio) of each array. The correlation the time of exposure to loud music. Remarkable of these physical factors and the attribute scores results of this investigation will be discussed. show that the contribution of these physical factors depends on music and its temporal change. Monday, May 19 16:00 Room K Convention Paper 7455 Technical Committee Meeting on Perception and Subjective Evaluation of Audio Tutorial 16 Monday, May 19 16:00 – 18:00 Room N101 Workshop 20 Monday, May 19 16:30 – 18:00 Room B QUANTIZATION EFFECTS IN AUDIO SIGNAL PROCESSING WOW, THIS SOUNDS GREAT!

Presenter: Jamie Angus, University of Salford, Salford, Chair: Christian Hugonnet, Consultant Greater Manchester, UK In this tutorial we will first present the principles of quanti- Panelists: Morten Lindberg, 2L Oslo, Oslo, Norway zation and the basic effects of finite precision in digital fil- George Massenburg, GML Inc. ter implementations (both fixed and floating point) on the Eirik Solheim, NRK – Norwegian Broadcasting audio signal. We explain the effects of finite precision on Helmut Voitl, Director frequency response (using an audio equalization task as an example). We shall see how different forms of filter What parameters do we have and do/will our children structures offer advantages when finite word length is have to say this in the age of MP3, Skype, and general considered. We shall also look at how different structures data-reduction? Does high-quality audio have a future? affect the audio signal. Finally we will discuss the Will new high-resolution music-only formats ever be a effect of different types of dither and the various pitfalls commercial success? Is generally the sensitivity for that can occur. intricacies and detail, for dynamics and resolution deteriorating? Do social developments have an impact on this situation? What is the effect of increasing environmental acoustic pollution? What are our criteria for “quality”? These and other questions will be discussed by a panel of experts with an acute sensitivity for those changes and a fascinating insight into this development. ¯

Audio Engineering Society 122nd Convention Program, 2007 Spring 43 Technical Program Special Event 09:00 NEW REVENUE STREAMS FOR STUDIOS P22-1 Manufacturing Recordings from 100-Year-Old Monday, May 19, 16:30 – 17:30 Masters—Sean Davies,1 Rinus Hooning2 Room O 1S.W. Davies Ltd., Aylesbury, UK In this special event, the European Sound Directors’ 2Record Industry bv, Haarlem, The Netherlands Association, ESDA in collaboration with the APRS have Most work on the 78 rpm analog recording for- invited a distinguished panel of prominent European stu- mat concentrates on pressings made near to the dio owners and producers including Robin Millar, Mal- time of the recording and the best ways to colm Atkin, and Dave Harries from the UK; Eric van Tijn retrieve the information from these for future and Chris Pilgram from the Netherlands; and others to storage and reproduction. However, a consider- discuss the potential opportunities to establish new busi- able number of metal master plates have been ness models and revenue streams for the recording preserved from the earliest days to the end of communities. the format’s active period. This paper describes Studios a project to manufacture new pressings from the original plates, the reasons for doing so, and the Times are hard for studios and producers and, for that technical challenges involved. matter, professional equipment manufacturers. We all Convention Paper 7456 wallow toward the bottom of the music business food chain suffering the rebound from the commercial uncer- 09:30 tainties that are attacking the record companies and the music business in general. Is the time right for studios to P22-2 Replay of Digital Original Tapes: Practical become participants in the royalty structure? What possi- Experiences with Video Tape Based PCM bilities are there for new studio-based services and other Adapters and R-DAT—Nadja Wallaszkovits,1 diversification such as digital distribution, video record- Heinrich Pichler,2 Johannes Spitzbart1 ing, DVD authoring and extended deliverables. 1Austrian Academy of Sciences, Vienna, Austria 2Audio Consultant, Vienna, Austria Producers As many of the early digital formats are already Producers have long waged a campaign to share in obsolete and support of these formats cannot be performance royalties. Progress in the UK, a few EU guaranteed much longer by the manufacturers, territories, and the USA is encouraging, however, the archives should presently give priority to the treatment of producers is inconsistent and needs to be replay of original recordings on such material. harmonized. Chairman of ESDA, Peter Filleul, will report Based on a short theoretical discussion and out- on progress across the territories. lining the format-specific characteristics, the paper discusses a variety of practical problems Monday, May 19 17:00 Room K of signal retrieval from PCM (Pulse Code Modu- Technical Committee Meeting on Semantc Audio lation) encoded signals on a VTR (video tape Analysis recorder) and R-DAT (Rotary-Head Digital Audio Tape), such as mechanical problems, tracking Special Event problems and playback incompatibilities, data BANQUET integrity checking, extraction and incompatibility Monday, May 19, 20:00 – 23:30 of sub-code-information, pre-emphasis, as well Kompaszaal as other problems occurring from irregular recording conditions (typically with field record- The traditional AES banquet offers an excellent opportu- ings produced on portable devices) or format nity to meet and/or invite your colleagues, (potential) peculiarity customers, officials, and all who are involved in audio. Convention Paper 7457 This year’s event will take place in the so-called Kom- paszaal, an Arrival/Departure hall of the KNSM, a ship- ping company in the east harbor region of Amsterdam. 10:00 This hall was built in 1956 under the supervision of archi- P22-3 A New System for File-Based Audio tect Johan van Tienhoven and radiates the style of the Recording, Preservation, and Access at fifties. The company no longer exists. If the weather Indiana University’s Jacobs School of Music cooperates, we have the possibility to have a wide look —Konrad Strauss, Travis Gregg, Indiana over the water while drinking our aperitifs. Here we also University, Bloomington, IN, USA would like to remember the 60 years existence of the AES! The Indiana University Jacobs School of Music 75 Euros for AES members; 80 Euros for nonmembers has been making live concert recordings on a Tickets will be available at the Special Events desk. variety of formats since the 1940s, and we continue to record approximately 500 concerts each year. Recent industry trends and changes in technology have led us to investigate the possi- Session P22 Tuesday, May 20 bility of creating high-resolution digital files rather 09:00 – 11:00 Room E-F than continuing to use physical media as the archival format for our recordings. Our goal was AUDIO ARCHIVING, STORAGE, RESTORATION, to develop a system for the creation, access, AND CONTENT MANAGEMENT and long-term preservation of high-resolution audio recordings and associated metadata that Chair: Tin Jonker, NOB, Hilversum, The Netherlands conformed to emerging standards for digital audio preservation. We began building such a system in July of 2006 and reached full imple-

44 Audio Engineering Society 124th Convention Program, 2008 Spring mentation in February of 2007. This paper gives product value high the information gap between an overview of the development process, pre- the producers and the lay clients must be nar- sents hardware and software solutions, and dis- rowed. The sub-set of this thesis is that this nar- cusses workflow and data management issues. rowing can be achieved through even a singular Convention Paper 7458 simulator training session. What is presented in this paper is the conceptual framework and some 10:30 preliminary tests. The tests are not substantial as studies are slow. AVAI is not a core area of the P22-4 A Fast Feature Extraction System on work of the Multimedia Section handling this study. Compressed Audio Data—Tobias Friedrich, Nevertheless, it is important to have some re- Matthias Gruhne, Gerald Schuller, Fraunhofer sponse from AES on this preliminary presentation. Institute for Digital Media Technology, Ilmenau, Convention Paper 7460 Germany [This paper was not presented but may be We describe an efficient system that directly purchased] extracts features from compressed audio material. It consists of a time/frequency conversion 10:00 method and a feature extraction algorithm. The P23-2 Nonexistence of Frontal Signal Unmasking conversion method provides the feature extrac- from Spatially Wide Masker—Ville Pulkki, tion algorithm with a suitable complex spectral Jukka Ahonen, Helsinki University of representation directly from the compressed Technology, Espoo, Finland domain. It further allows a trade-off between computational complexity and conversion accu- The masking of a frontal signal by spatially wide racy. Several operating points using different noise sources was investigated in a listening conversion accuracies were tested with an experiment. The noise sources consisted of a sin- MPEG audio identification system in order to gle or multiple symmetrically positioned loud- evaluate the identification confidence. Based on speakers in the frontal horizontal plane in anechoic these results it is possible to reduce the compu- conditions. It is shown that the detection threshold tational complexity from O(N log N) to O(N) com- of the signal does not depend on masker width, pared to the conventional approach (complete which suggests that frontal unmasking does not decoding followed by a frequency analysis). exist in loudspeaker listening. In additional tests Convention Paper 7459 with signal source positioned in side it is shown that moderately small binaural unmasking occurs in that case from wide masker, and that increasing Workshop 21 Tuesday, May 20 the width of masker source decreases binaural 09:00 – 10:30 Room N101 unmasking effect. Convention Paper 7461 JOAN OF ARC [CANCELLED] 10:30 Session P23 Tuesday, May 20 P23-3 Reaction Times and Performances in 09:30 – 12:00 Room C-D Recognition Tasks to Assess Speech Quality —Virginie Durin,1 Laetitia Gros,1 Gilles Hericher2 PSYCHOACOUSTICS, PERCEPTION, 1Orange Labs, Lannion, France AND LISTENING TESTS, PART 2 2Laboraoire Psychologie et Neurosciences de la cognition, Mont Saint Aignan, France Chair: Joerg Bitzer, University of Applied Science This paper deals with perceptive test methodolo- Oldenburg, Oldenburg, Germany gies to assess speech quality of telecommunica- tion systems. Faced with drawbacks of typical 09:30 methodologies recommended by ITU-T, a new way to assess speech quality is investigated, by P23-1 A Proposed Audio Visual Product Measure— collecting reaction times and performances Joe Peters, National University of Singapore, when subjects are achieving tasks involving Singapore degraded speech signals. A duel task with a digit The Multimedia Section at the Centre for Instruc- recognition memory task and a letter recognition tional Technology at the National University of Sin- task is proposed. Three different quality levels gapore has developed an audio visual are applied to audio signals describing digits and assessment index (AVAI) to serve as a tool for letters. The results show significant differences clients to measure their evaluation of audio and of performances and reaction times between the video products. AVAI is based on a listing of three quality levels. indicators and variables that make up the funda- Convention Paper 7490 mental elements in the capture and processing of AV products (video production): image, color, light, 11:00 audio, form, aesthetics, and delivery. AVAI is cur- P23-4 Evaluation of Stereophonic Images with rently being used by professionals for internal Listening Tests and Model Simulations— evaluation. A series of simulator-based AVAI Munhum Park,1 Philip Nelson,1 Kyeong Ok Kang2 courses are also underway, the purpose of which 1University of Southampton, Highfield, is to enable lay persons to understand the indica- Southampton, UK tors and variables through simulated 2Electronics and Telecommunications Research explanations. The thesis is, that in order to keep Institute (ETRI), Daejeon, Korea ¯

Audio Engineering Society 124th Convention Program, 2008 Spring 45 Technical Program A binaural hearing model has recently been sug- Session P24 Tuesday, May 20 gested for the evaluation of the performance of 09:30 – 11:00 Topaz Lounge virtual acoustic imaging systems. The model considers excitation-inhibition (EI) cell activity patterns as the internal representation of sound POSTERS: ROOM AND ARCHITECTURAL localization cues and a pattern-matching proce- ACOUSTICS AND SOUND REINFORCEMENT dure with a frequency-weighting scheme produces the estimate of source location in the 09:30 horizontal plane. Given the reasonable prediction of some important features in human sound P24-1 Objective Evaluation of a Non-Environmental localization and lateralization, this paper pre- Control Room for 5.1 Surround Listening— sents a further verification and application of the Soledad Torres-Guijarro,1 Antonio Pena,2 model in actual listening tests. In this paper par- Norberto Degara-Quintela2 ticipants' responses to stereophonic images 1Laboratorio Oficial de Meroloxía de Galicia have been compared with the predictions of the (LOMG), Tecnópole, Ourense, Spain model, individually established from the subject's 2Universidad de Vigo, Vigo, Spain own HRTF. Model predictions have been found The control room of the Universidad de Vigo was to be both qualitatively and quantitatively consis- built for the purpose of assessing small audio tent with the test results, and in particular, the artifacts, such as listening analytically to coded agreement between 2 and 3 kHz gave a good material with different data rates. It follows a indication that, unlike some similar models, the non-environment design that minimizes the influ- current model can effectively incorporate both ence of the room. The use of such a room as a ITD and ILD information according to their rela- 5.1 surround listening room will be analyzed tive importance. according to international recommendations. Convention Paper 7463 This research includes the study of the electroacoustic behavior of loudspeakers, geometric 11:30 and acoustic properties of the room, and sound field conditions. A discussion of some diver- P23-5 The Sound Character Space of Spectrally gences and implications for its use when per- Distorted Telephone Speech and its Impact forming surround listening tests follows the on Quality—Marcel Wältermann, Alexander measurement results. Raake, Sebastian Möller, Berlin University of Convention Paper 7465 Technology, Berlin, Germany Spectral distortions of speech transmitted over a 09:30 telephone channel may stem from linear channel P24-2 A Case Study of Sound Reproduction and filtering, codecs, electro-acoustic properties of Acoustic Enhancement in Concert Halls end-user terminals, or the acoustic environment Using Wave Field Synthesis—Clemens at send side. In this paper a study is presented Kuhn-Rahloff, Matthias Rosenthal, Max Cas- that aims at revealing the perceptual space of dorff, Roger Moser, sonic emotion ag, Obergltt spectrally distorted telephone speech and estab- (Zurich), Switzerland lishing a link to the overall quality of the speech. Two dimensions were identified as relevant for This paper presents the wave fields synthesis explaining the perceived quality: indirectness system under construction at the National and brightness. Whereas brightness is related to Conservatory of Music Detmold (Germany). The the center frequency of a transfer function, indi- system is dedicated to sound reproduction for rectness is correlated with the equivalent rectan- artistic purposes at the Tonmeister department gular bandwidth and constitutes the dominating (Erich Thienhaus Institute) and to an enhance- factor in the perceptual space in terms of cov- ment of room acoustics. The system comprises ered variance. The concept of the bandwidth im- 346 independent loudspeaker channels, includ- pairment factor that fits into the framework of the ing a horizontal loudspeaker array all around the so-called E-Model and that is based on these auditory (500 seats) and ceiling loudspeakers. simple parameters for computing the integral Since the hall is used for a broad repertoire com- quality of spectrally distorted speech could suc- prising chamber music, romantic orchestra cessfully be applied to the given data. instrumentations, organ concerts, contemporary Convention Paper 7464 music, etc., the hall will be equipped with a variable room acoustic system. The paper presents perceptual aspects of system design concerning Tuesday, May 20 09:00 Room H the direct sound and diffuse field as well as prac- Standards Committee Meeting: AESSC Plenary tical implementations for WFS rendering. Convention Paper 7466

09:30 P24-3 Small Studios with Gypsum Board Sound Insulation: A Review of Their Room Acoustics, Details at the Low Frequencies— Lorenzo Rizzi, Francesco Nastasi, Rizzi Acustica, Lecco LC, Italy At the present time most music preproduction

46 Audio Engineering Society 124th Convention Program, 2008 Spring and production is often carried out in very small, Dependent Boundary Conditions in a Digital privately owned rooms, which are called “project Waveguide Mesh—José Escolano,1 Basilio studios.” Gypsum board technology is very com- Pueo,2 José J. López,3 Máximo Cobos3 mon in the construction of these rooms because 1University of Jaén, Linares, Jaén, Spain of its high insulation capabilities compared to low 2University of Alicante, Alicante, Spain monetary and time costs. The paper discusses 3Technical University of Valencia, Valencia, Spain sweet spot impulse response measurements that have been carried out in three different but The digital waveguide mesh is a popular method acoustically small rooms built with gypsum board for time domain acoustic system simulation such sound insulating structures comparing it to a as room acoustics. One of the main reasons to masonry built one. The room modal behavior is choose this paradigm relies in the ease to underlined, continuing with the analysis of include boundary conditions in the simulation. decaying in time at low frequencies related to This paper is focused on the comparison of the insights on perception and analysis. A different simulation with real-world measurements, where methodology of study is proposed. a particular scenario is physically built and the Convention Paper 7467 corresponding simulation, according to their physical parameters, is carried out. The main scope of this paper is the validation and discus- 09:30 sion of a boundary condition model and their P24-4 On the Measurement of Electro Acoustic correspondence with the measurements through Enhanced Sound Fields—Florian Walter,1 an example. Frank Melchior2 Convention Paper 7470 1Mueller-BBM GmbH, Munich, Germany 2Fraunhofer Institute for Digital Media 09:30 Technology, Ilmenau, Germany P24-7 Subjective Effects of Dispersion in the The installation and optimization of acoustic Simulation of Room Acoustics Using Digital enhancement systems require a large amount of Waveguide Mesh—Jose J. Lopez,1 José experience. The verification in terms of mea- Escolano,2 Maximo Cobos,1 Basilio Pueo3 surement is most of the time done using conven- 1Technical University of Valencia, Valencia, Spain tional reverberation and acoustic parameter 2University of Jaén, Linares (Jaén), Spain measurements according to ISO 3382. This is a 3University of Alicante, Alicante, Spain good solution for diffuse sound analysis and general examination of early reflections, but in The simulation of room acoustics using the Digi- terms of direction dependant analysis the results tal Waveguide Mesh method has gained interest are not satisfying. In this paper a room equipped in the last few years. One of the problems of this with an acoustic enhancement system was mea- method is the frequency and angle dependent sured using a circular array. The effects of dispersion. In order to reduce this effect, an adding specific early reflections and direction- oversampling is usually employed but at the cost dependent diffuse energy generated by the of highly increasing the resulting computational acoustic enhancement system are investigated. cost and restricting the simulation to lower fre- The results are compared to standard measure- quencies. In this paper a subjective analysis is ments according to ISO 3382. carried out; where different oversampling factors Convention Paper 7468 in voice band simulations have been performed and evaluated by a set of listeners. Some listening tests employing ABX methodology have 09:30 been used to evaluate the subjective effects, P24-5 Applying Cochlear Modeling and obtaining some preliminary results that, although Psychoacoustics in Room Acoustics—Jasper not being conclusive, they represent a first van Dorp Schuitman, Diemer de Vries, Delft approach to the problem. University of Technology, Delft, The Netherlands Convention Paper 7471 The acoustical qualities of a concert hall or any other room are generally expressed using acoustical parameters. These parameters are Workshop 22 Tuesday, May 20 determined from impulse responses, as mea- 09:30 – 11:00 Room B sured from single positions in a room or along a line array. However, from array measurements it SOS—SAVE OUR SPECTRUM! turned out that parameters can fluctuate severe- ly between small distance steps, something Chair: Malcolm Johnson, Institute of Broadcast which does not agree with human perception. Sound, UK Applying cochlear modeling and psychoacoustics in this process seems a promising tech- Panelists: Matthias Fehr, Sennheiser – Initiative Digital nique to reach results that do not suffer from Dividend these fluctuations and thus are much closer to Jürgen Hanelt, VDT Germany human perception compared with conventional Jan Outters, IRT Mnich techniques. Convention Paper 7469 The sell-out of the UHF radio spectrum poses a serious threat not only to the broadcast industry but on most 09:30 other audio fields as well. Coordinated international action is required in a situation where time is running P24-6 Empirical Evaluation of the Frequency- out. The legal situation will be examined and possible

Audio Engineering Society 124th Convention Program, 2008 Spring 47 Technical Program and necessary steps will be discussed. Current activi- Panelists: Karlheinz Brandenburg, Fraunhofer Institute ties will be presented, hoping to raise the importance of James Cridland, BBC Audio and Music immediate action to still counteract this dramatic Interactive development. Jurgen Jaron, mufin GmbH Liz Rice, Last.fm

Workshop 23 Tuesday, May 20 The continuing rapid evolution of the online digital music 09:30 – 10:00 Room P scene has meant that a wide variety of independent music sources now exist, many catering to indie artists or specific genres or niches, and often with individual- MUSICAL TELEPRESENCE WITH NARROW BAND ized business models and distribution methods. In the NETWORKS process, the exploration, discovery, and critiquing of music has become decidedly more complex, if potential- Chair: Alexander Carôt ly more interesting, for the consumer. Online music sources range from massive data bases, some available Panelists: Ulrich Kramer only by membership or subscription and tied to hard- Gerald Schuller ware or ISPs; to social networking sites; to home web- Christian Werner sites of labels, bands, orchestras, artists, and radio stations; to blogs; a/v streaming; internet radio, among In this presentation Alexander Carôt will demonstrate a many others. Methods to adequately search out music live music performance with professional musicians in among all of these, to judge quality, and to find the bit Germany. He will play bass guitar live and in real time rates, formats, and software compatible with one’s play- over the Internet as if in the same room. In that context back system are still evolving. This workshop looks at he will use his Soundjack software in combination with current approaches to the search and evaluation prob- the ULD codec. While Soundjack offers the appropriate lem, including algorithmic methods, the exchange of low delay audio streaming conditions, the ULD codec opinion via social networks, online reviews, etc., and (developped at Fraunhofer IDMT) introduces a total also considers whether online reviews, blogs, and chat encoding/decoding delay of only 6 ms and provides rooms are adequate as the new determinants of suc- decent bit rates for the use in narrow band networks. cess and quality.

Historical Committee Meeting Session P25 Tuesday, May 20 11:30 – 16:00 Room E-F The Historical Committee meeting will be held Tues- day, May 20, at 11:00 in Room N. MULTICHANNEL SOUND

Chair: Günther Theile, Institut für Rundfunktechnik, Tutorial 18 Tuesday, May 20 Munich, Germany 10:30 – 12:30 Room L 11:30 WHAT IS THE MATRIX? P25-1 Bitstream Format for Spatio-Temporal Wave Field Coder—Francisco Pinto, Martin Vetterli, Presenters: Geoff Martin, Bang & Olufsen a/s, Ecole Polytechnique Fédérale de Lausanne, Struer, Denmark Lausanne, Switzerland Helmut Wittek, Schoeps Mikrofone GmbH, We present a non-parametric method for com- Karlsruhe, Germany pressing multichannel audio data for reproduc- This tutorial will be an introduction to the use of matrixing tion through Wave Field Synthesis. The method in microphone and mixing techniques for stereo and mul- consists of applying a two-dimensional filterbank tichannel. The basics of microphone polar patterns will to the input multichannel signal, in both time and be explained, followed by the fundamentals of tech- channel dimensions, and coding the two-dimen- niques such as "textbook" M/S, double M/S, and sound- sional spectra using a spatio-temporal frequency field recording. Included in the tutorial will be a discus- masking model. The coded spectral data is sion of how to use matrixing to "re-mix" an existing organized into a bitstream together with side recording, to modify microphone configurations in post- information containing scale factors and Huff- production, and to manipulate spatial characteristics of man codebook information. We demonstrate stereo mixes. In addition, information on the exciting pos- how this coding method can be applied to any sibilities in the fast-paced world of karaoke soundtrack smooth distribution of loudspeakers in space, production will be presented. while obtaining a stable bit rate that is 15% lower compared to coding each channel independently Convention Paper 7472

Workshop 24 Tuesday, May 20 12:00 11:00 – 13:00 Room N101 P25-2 The Design of Ambisonic Decoders for the ITU 5.1 Layout with Even Performance FINDING AND EVALUATING MUSIC FOR Characteristics— David Moore, Jonathan DOWNLOADING OR STREAMING Wakefield, University of Huddersfield, Huddersfield, West Yorkshire, UK Chair: Steve Harris, BridgeCo AG All previously published Ambisonic decoders for irregular loudspeaker layouts have localization

48 Audio Engineering Society 124th Convention Program, 2008 Spring performance that varies significantly by angle 13:30 around the listener. This contrasts with decoders designed for evenly spaced arrangements of P25-5 Frequency-Dependent Signal-Correlation in loudspeakers where performance characteristics Surround- and Stereo-Microphone Systems and the Blumlein-Pfanzagl-Triple (BPT)— are isotropic. Furthermore, even localization per- 1 2 formance around the listener is desirable for a Edwin Pfanzagl-Cardone, Robert Höldrich 1Salzburg Festival, Salzburg, Austria number of application areas of 5.1 surround 2 sound. New decoder design criteria are present- Institute of Electronic Music and Acoustics, ed that aim to reduce this variation in localization Graz, Austria performance. These criteria are added to a mul- With the aim to recreate the original concert-hall ti-objective fitness function, based on auditory sound field as faithfully as possible in the con- localization theory, which guides a heuristic trol- or living-room, recordings were made simul- search algorithm to derive decoder parameter taneously with an artificial head and several sets for the ITU5.1 layout. The derived decoders surround microphone techniques (among them exhibit a significant improvement in localization the new BPT method). The surround recordings performance variation by angle around the 360- were rerecorded using the same dummy-head degree sound stage. as in the concert hall. The results of subjective Convention Paper 7473 listening tests (loudspeaker as well as binaural) were assessed using ANOVA and correlation 12:30 analysis. Acoustical analysis of the dummy-head recordings was performed by measuring the Fre- P25-3 Methods for Sharing Stereo and Multichannel quency-Dependent Inter Aural Cross-Correlation Recordings among Planetariums—Leslie Coefficient (FIACC): the low-correlation AB-PC Gaston, Peter Dougall, Erick D. Thompson, microphone system was capable of reproducing University of Colorado at Denver, Denver, CO, the original sound field better than any of the USA other systems under test (DECCA, KFM, OCT). There is a demand for research on the transfer- A microphone systems Critical Frequency, ability of surround sound audio from one plane- below which correlation raises toward 1, is tarium to another, so that (1) audiences have defined. similar experiences and (2) audio engineers can Convention Paper 7476 easily create this experience. This paper will consider: acoustics, production, delivery, equip- 14:00 ment, and seating arrangements. Our recent survey of over 100 planetariums worldwide in P25-6 Holographic Design of Source Array for the fall of 2007 will provide a look at current Achieving a Desired Sound Field—Wan-Ho Cho,1 Jeong-Guon Ih,1 Marinus M. Boone2 practices. The University of Colorado Denver 1 and Gates Planetarium have collaborated in Korea Advance Institute of Science and Technology (KAIST), Daejeon, Korea order to explore the potential of current audio 2 technology, and to discover what similarities and Delft University of Technology, Delft, The differences exist between planetariums in Netherlands order to achieve this goal of transferability. For realizing a desired complicated sound field, Convention Paper 7474 an acoustic source array should be designed appropriately to obtain the acoustic source para- 13:00 meters. To this end, we suggest a method utilizing the acoustical holography technique based P25-4 Optimal Hierarchical Bandwidth Limitation of on the inverse boundary element method. Surround Sound—Yu Jiao, Slawomir Zielinski, Acoustical analogy between the problems of Francis Rumsey, University of Surrey, Guildford, source reconstruction and source design was Surrey, UK the initial motivation of the study. In the design of In order to save the transmission bandwidth of the source array, the pressure distribution at surround sound, a technique named Hierarchical specific field points is the constraint of the prob- Bandwidth Limitation (HBL) was proposed by the lem and the signal distribution at the source sur- authors. In HBL, a psychoacoustically hierarchi- face points is the object function of the problem. cal transform is used as the preprocessing algo- The whole procedure of the application consists rithm prior to bandwidth limitation. In our former of three stages. First, a condition of the desired experiments we found that the Karhunen-Lòeve sound field should be set as the constraint. Sec- transform (KLT) is a suitable hierarchical trans- ond, the geometry and boundary condition of the form for HBL. Besides the hierarchical transform, source array system and the target field, i.e., the choice of an appropriate strategy for band- points in the sound field of concern, are modeled width allocation is also essential from the point by the boundary elements. Actual characteristics of view of the resultant audio quality. In order to of source and space can be considered to gen- find the optimal bandwidth allocation strategy erate the accurate condition of the target field. that achieves the best audio quality, the authors Finally, the source parameters are inversely attempted to build up the mathematical relation- calculated by the backward projection. As an ship between audio quality and the bandwidth al- example, a source array to fulfill the plane wave location strategy using a MUSHRA listening test. propagating zone and another quiet zone near The experiment design and results of this listen- the propagation zone was designed and tested ing test are reported in this paper. by simulation and measurement. Convention Paper 7475 Convention Paper 7477 ¯

Audio Engineering Society 124th Convention Program, 2008 Spring 49 Technical Program 14:30 (MAGIC arrays). Convention Paper 7480 P25-7 New Dimensions for Ambisonics—Michael Chapman, Culoz, France Session P26 Tuesday, May 20 Both two-dimensional (pantophonic) and three- 11:30 – 13:00 Topaz Lounge dimensional (periphonic) representations of soundfields are common place in ambisonics. POSTERS: LOW BIT-RATE CODING Reproducing either on rigs essentially designed for the other is common place. What though if 11:30 one synthesizes a four (or more) dimensional soundfield and reproduces this on a standard P26-1 Autoregressive Modeling of Hilbert rig? As there appears to be no source on hyper- Envelopes for Wide-Band Audio Coding— spherical harmonics applicable to ambisonics, Sriram Ganapathy,1,2 Petr Motlicek,1 Hynek the mathematical basis is first set out. The Hermansky,1,2 Harinath Garudadri3 manipulation of hyperambisonic soundfields (ro- 1IDIAP Research Institute, Matigny, Switzerland tation, mirroring, dominance) is then discussed. 2Ecole Polytechnique Fédérale de Lausanne During that discussion various “proofs” are (EPFL), Lausanne, Switzerland advanced as to the finite range of transforma- 3Qualcomm Inc., San Diego, CA, USA tions that can be applied to ambisonic soundfields, of whatever dimension. Frequency Domain Linear Prediction (FDLP) Convention Paper 7478 represents the technique for approximating temporal envelopes of a signal using autoregressive models. In this paper we propose a wide-band 15:00 audio coding system exploiting FDLP. Specifical- ly, FDLP is applied on critically sampled sub- P25-8 Improving Spherical Microphone Arrays— bands to model the Hilbert envelopes. The Nicolas Epain, Jérome Daniel, France Télécom residual of the linear prediction forms the Hilbert R&D, Lannion, France carrier, which is transmitted along with the enve- Spherical microphone arrays are useful for numer- lope parameters. This process is reversed at the ous applications, such as spatial audio capture decoder to reconstruct the signal. In the objec- and beamforming. However, these sensor arrays tive and subjective quality evaluations, the are known to have a limited frequency range, due FDLP-based audio codec at 66 kbps provides to poor directivity at low frequencies and spatial competitive results compared to the state-of-art aliasing at high frequencies. In this paper we study codecs at similar bit-rates. two methods aiming at enhancing the frequency Convention Paper 7481 range of spherical microphone arrays without using more sensors. First, the benefit of locating 11:30 the sensors at the end of cavities within the sphere is assessed through measurements and simula- P26-2 On Locality of Spectral Oriented Tree for Bit- tions. Second, we study the influence of using Plane Based Low-Bit Rate Audio Coding— large membrane microphones. Finally, results Yu-Lin Wang, Alvin W. Y. Su, National Cheng- show that the frequency range could be increased Kung University, Tainan, Taiwan in both cases studied. For Spectral Oriented Trees (SOT) based Convention Paper 7479 coders such as SPIHT and CEIHT, locality is usually related to the locations of coefficients 15:30 within a SOT and its effect to coding efficiency. How to construct a SOT to achieve better locality P25-9 Migration of 5.0 Multichannel Microphone is very important. This paper presents a diag- Array Design to Higher Order MMAD (6.0, 7.0, nostic aspect of the localities of different order- and 8.0) with or without the Inter-Format ing techniques for low bit-rate audio coding. We Compatibility Criteria—Michael Williams, used several coefficient ordering schemes to Sounds of Scotland, Paris, France construct SOTs with the same set of MDCT The severe limitations of the 5.0 Multichannel coefficients and observed their effects. Both Reproduction Standard in reproducing good objective and subjective results are presented. quality audio-visual or stand-alone audio sur- Convention Paper 7482 round sound reproduction has increased the pressure on recording and reproduction system 11:30 designers to increase the number of channels in P26-3 Perceptual Matching Pursuit for Audio an attempt to give an even more satisfactory Coding— Hossein Najaf-Zadeh, Ramin envelopment experience. This paper extends the Pichevar, Hassan Lahdili, Louis Thibault, MMAD process to show how higher order chan- Communications Research Centre Canada, nel array designs (6.0, 7.0, and 8.0) can be Ottawa, Ontario, Canada developed from the existing data on 4.0 or 5.0 Multichannel Front Sound Stage Coverage Array This paper introduces a Perceptual Matching Designs with almost perfectly seamless and lin- Pursuit (PMP) algorithm for audio coding. A ear surround sound reproduction. Designing for masking model has been developed and inte- inter-format compatibility can also be accommo- grated into the matching pursuit algorithm to dated from the existing multi-format array design account for the characteristics of the hearing data described in a previous paper on Multichan- system. By doing so, only an audible kernel is nel Arrays Generating Inter-format Compatibility extracted at each iteration. Moreover, contrary to

50 Audio Engineering Society 124th Convention Program, 2008 Spring the matching pursuit algorithm, PMP will stop We propose a novel technique of using bit allo- decomposing an audio signal once there is no cation for linear prediction coefficients in asym- audible part left in the residual. We have used metric lossless audio compression. We show ITU_R PEAQ to compare audio materials how to determine the optimal bit allocation using decomposed by PMP and by matching pursuit. a new closed-form formula for the excess error Objective scores for PMP increase by up to 1 unit. from quantization, and describe a recently intro- A semi-formal listening test has verified the objec- duced algorithm (Optimization-Quantization tive scores and shown the perceptual superiority Least Squares), which computes the optimal of PMP over the matching pursuit algorithm. quantized prediction coefficients applied for the Convention Paper 7483 allocation. The proposed method, implemented as a modified asymmetrical OptimFROG, 11:30 obtains small (but consistent) signal dependent compression improvements with virtually no de- P26-4 A Unifying Approach to Transform and coder complexity increase (on a 847 MB audio Sinusoidal Coding of Audio—Maciej corpus, up to 0.27%, on average 0.06%). Com- Bartkowiak, Poznan University of Technology, pared to MPEG-4 ALS, it obtained 0.38% better Poznan, Poland compression, while being at the same time The paper describes a new scenario for low bit approximately 5 times faster at decoding. rate audio compression that combines two clas- Convention Paper 7486 sical techniques: transform coding and sinusoidal coding into a united framework. The main 11:30 idea is to adaptively decompose the audio signal P26-7 Design of Framing in MPEG Surround Based into subbands whose central frequencies follow on Dynamic Programming Algorithm— continuously the local instantaneous frequencies Chi-Min Liu, Chung-Han Yang, Han-Wen Hsu, of certain signal components (formants or indi- Wen-Chieh Lee, National Chiao Tung University, vidual harmonic partials). The content in each Hsinchu, Taiwan subband is encoded in the baseband after frequency shift toward DC. The technique may be MPEG Surround (MPS) defined by ISO/IEC is considered either as modified transform coding, the audio coding standard of multichannel sig- i.e., coding along instantaneous frequencies or nals based on the down-mixed signal and the as extended sinusoidal coding, i.e., modeling spatial parameters. In MPEG Surround, the with partial envelopes that are represented by time-frequency tiles decide the units to share the transform coefficients. In other words, it is a same spatial parameters among the multichan- hybrid scheme offering a continuous operating nel signals. Hence, the decision of the tiles is the mode between purely transform and purely sinu- critical module deciding the required quality and soidal compression. bits. However, the large number of combination Convention Paper 7484 in the time regions, frequency bands, and multichannel signal statistics has spanned the huge 11:30 search space for deciding the tiles. Our previous work at AES 119 has proposed the dynamic pro- P26-5 Low Bit Rate Audio Coding for Digital gramming method to efficiently decide the time- Wireless Systems—Stephen Wray, APT Ltd., frequency units for the parameter stereo coding Belfast, Northern Ireland, UK in HE-AAC. This paper will extend the dynamic With the transition from analog to digital televi- programming method to the MPS coding. sion, the available spectrum for wireless micro- Convention Paper 7487 phones, in-ear monitors, and other wireless devices could be under threat. Spectrum is a 11:30 valuable commodity, and it is the responsibility P26-8 New Enhancements to the Audio Bandwidth of governments to manage it appropriately. Extension Toolkit (ABET)—Harinarayanan E. Much has been made recently of the Spectrum V.,1 Raghuram Annadana,1 Deepen Sinha,2 Squeeze both sides of the Atlantic with discus- Anibal Ferreira2,3 sions on White Spaces and the Digital Dividend. 1ATC Labs, Noida, India With bandwidth at such a premium the audio 2ATC Labs, Chatham, NJ, USA industry has been forced to consider new tech- 3University of Porto, Porto, Portugal nologies that make efficient use of spectrum without sacrificing quality or service. Within this Audio bandwidth extension has emerged as a context, we need a new revolutionary approach key low bit rate coding tool. In continuation with to maximizing bandwidth efficiency. The author our on going research on audio bandwidth will present a new and novel coding solution to extension, this paper presents new enhance- overcome the prevailing technical limitations and ments to the Audio Bandwidth Extension Toolkit industry requirements for wireless applications. (ABET). ABET consists of three primary tools Convention Paper 7485 Accurate Spectral Replacement (ASR), Fractal Self Similarity Model (FSSM), and Multi-band 11:30 Temporal Envelope Amplitude Coding (MBTAC). Additionally we have also introduced a blind P26-6 Bit Allocation for Linear Prediction bandwidth extension mode into ABET. We dis- Coefficients with Application to Lossless cuss several new ideas / improvements to Audio Compression —Florin Ghido, Ioan ABET. Specifically, enhancements to the blind Tabus, Tampere University of Technology, bandwidth extension architecture that allow it Tampere, Finland to work with signals with only 3.5–4.0 kHz au- ¯

Audio Engineering Society 124th Convention Program, 2008 Spring 51 Technical Program dio bandwidth are described. We also elaborate 13:00 on a new tool for efficient coding of time-frequency envelope that cuts the overhead by P27-1 Perceptual Evaluation of Numerically 0.75–1.0 kbps/channel. We also address a prac- Simulated Head-Related Transfer Functions tical issue, i.e., the computational complexity —Julia Turku, Miikka Vilermo, Eira Seppälä, and describe a new low decoder complexity Monika Pölönen, Ole Kirkeby, Asta Kärkkäinen, mode of ABET Leo Kärkkäinen, Nokia Research Center, Convention Paper 7488 Helsinki, Finland Head-related transfer functions (HRTFs) produced by numerical simulations were compared Workshop 25 Tuesday, May 20 to measured HRTFs through two listening tests. 11:30 – 13:30 Room O The purpose was to determine whether the numerically simulated HRTFs, which do not con- 10 THINGS TO GET RIGHT IN PA AND SOUND tain any of the artifacts associated with acoustic REINFORCEMENT measurements, capture the detail necessary for reproducing convincing 3-D sound. The results suggest that when virtual sound sources are pre- Chair: Peter Mapp, Peter Mapp Associates, sented to listeners binaurally over headphones, Colchester, UK the measured and modeled HRTF sets perform Panelists: Mark Bailey, Harman/JBL equally well in terms of perception of direction. Luke Jenks, Meyer Sound Laboratories, Regarding preference of binauralization meth- Berkeley, CA, USA ods, the simulated HRTFs performed slightly Evert Start, Duran Audio better. Convention Paper 7489 The workshop will discuss the 10 most important things to get right when designing / operating sound reinforcement 13:30 and PA systems. However, as attendees at the workshop will learn there are many more things to consider than just P27-2 Evaluating Perception of Salient Frequencies: the 10 golden rules, and that the order of importance of Do Mixing Engineers Hear the Same Thing? these often changes depending upon the venue and type —Joerg Bitzer,1 Jay LeBoeuf,2 Uwe Simmer1 of system. The workshop aims to provide a practical ap- 1University of Applied Science Oldenburg, proach to sound systems design and operation and will Oldenburg, Germany be illustrated with many practical examples and case his- 2Imagine Research, Inc., San Francisco, CA, USA tories. Each workshop panelist has many years of practical experience and between them can cover just about In this paper we analyze the agreement of mix- any aspect of sound reinforcement and PA systems de- ing engineers when finding salient frequencies in sign, operation, and technology. Come along to a work- recorded audio tracks. Twenty-two mixing engi- shop that aims to answer questions you never knew you neers were asked to use an equalizer with a had—but of course, to find out the ten most important high-Q and high-gain setting. Using this tool to ones, you will have to attend the session! sweep through the files’ frequencies, they analyzed sixteen audio tracks and reported the most perceptually salient frequencies. The results Master Class 3 Tuesday, May 20 show that the agreement depends on the analy- 12:00 – 14:00 Room B sis bandwidth. Most mixing engineers agree with a wide frequency range. However, only a few LINEAR AUDIO POWER AMPLIFICATION engineers agree if the matching bandwidth is below or equal to one-third octave. In this paper Presenter: Douglas Self we try to explain these results and give a detailed analysis. Audio power amplifiers are one of the few fields where Convention Paper 7462 analog still has the upper hand when high quality is required. In this presentation I will review some of the configurations available and examine their advantages 14:00 and otherwise. The classical amplifier configuration still P27-3 Influence of Visual Appearance on remains extremely useful and is capable of very good Loudspeaker Sound Quality Evaluation— results in terms of linearity when the design details are Alex Karandreas, Flemming Christensen, handled properly. I hope to show that this can be done Aalborg University, Aalborg, Denmark easily. Some new findings in minimizing distortion will be presented. I shall also be looking at the issue of noise in Product sound quality evaluation aims to identify power amplifier circuits, and how that too can be reduced relevant attributes and assess their influence on as much as possible. the overall auditory impression. Extending this sound specific rationale, the present paper eval- Session P27 Tuesday, May 20 uates overall impression in relation to audition 13:00 – 16:00 Room C-D and vision, specifically for loudspeakers. In order to quantify the bias that the loudspeaker appearance has on the sound quality evaluation of a PSYCHOACOUSTICS, PERCEPTION, naive listening panel, audio stimuli of varied AND LISTENING TESTS, PART 3 degradation are coupled with actual loudspeakers of different visual appearance. Chair: John Beerends, TNO Information and Convention Paper 7491 Communication Technology, Delft, The Netherlands

52 Audio Engineering Society 124th Convention Program, 2008 Spring 14:30 tering. The resulting analyses include determination of sounds that were judged as inappropriate, P27-4 Comparison of Loudspeaker/Room independent of condition. Equalization Preferences for Multichannel, Convention Paper 7493 Stereo, and Mono Reproductions: Are Listeners More Discriminating in Mono?— Sean Olive, Sean Hess, Allan Devantier, 15:30 Harman International Industries, Inc., Northridge, P27-6 Loudness Calculation for Individual CA, USA Acoustical Objects within Complex Automated digital loudspeaker/room correction Temporally Variable Sounds—Cornelius products are more popular than ever despite the Bradter, Klaus Hobohm, Hochschule für Film general lack of perceptual studies on their per- und Fernsehen, Potsdam, Germany formance measured over a range of different Models used for loudness calculation normally playback conditions. This paper describes the treat their input signal as an integral whole. For first of several perceptual experiments designed sounds consisting of two or more distinguishable to explore how different loudspeaker-room cor- acoustical objects this contradicts the listening rection methods affect the sound quality of experience. Auditory perception analyzes and reproduction given a range of different listening identifies acoustical objects and may treat them rooms, loudspeakers, setups, and programs that differently. By expanding principles used in exci- might influence their perceived performance. A tation synthesis-based loudness models, we panel of trained listeners gave comparative pref- developed a procedure to calculate loudness of erence ratings for three different loudspeaker a time-varying acoustical object while a second equalizations evaluated in a semi-reflective room object is simultaneously present. When signals using three multichannel music recordings repro- of both objects are available individually and in duced in surround, stereo, and mono playback combination, the procedure reflects effects of modes. The three equalizations were based on one object on the other as well as changes of either anechoic or in-room measurements with loudness perception due to signal features of different perceptual weighting given to the direct one or both objects. versus the direct + reflected sounds radiated by Convention Paper 7494 the loudspeaker. The different equalizations were identical below 400 Hz to focus on perceptual effects occurring above the room’s transition frequency. The results are summarized as Workshop 26 Tuesday, May 20 follows: all three equalizations were equally pre- 13:00 – 14:30 Room L ferred over the unequalized system; the difference in preference increased monotonically as MIX YOUR OWN NEW YEAR’S CONCERT—PART 1 the number of playback channels was reduced from 5 (surround) to 1 (mono). Chair: Florian Camerer, ORF – Austrian TV Convention Paper 7492 Panelists: Jean-Marie Geijsen, Polyhymnia Morten Lindberg, 2L Oslo - Oslo, Norway 15:00 Ronald Prent, Galaxy Studios P27-5 Caution and Warning Alarm Design and Josef Schütz, ORF Radio Evaluation for NASA CEV Auditory Displays 1 2 A selection of experienced mixers with different back- —Durand Begault, Martine Godfroy, Aniko grounds (classical music, pop, rock) have about 5-7 min- Sandor,3 Kritina Holden4 1 utes to set up a 5-channel mix of the current New Year’s NASA Ames Research Center, Moffett Field, Concert with the original multitrack recording. The audi- CA, USA 2 ence will be able to witness the individual approaches to San Jose State University Foundation, NASA that task. A second session (30 minutes later) will com- Ames Research Center, Moffett Field, CA, USA 3 pare those mixes all aligned so that on-the-fly switching LZ Technology, NASA Johnson Space Center, is possible. The different approaches and results will be Houston, TX, USA discussed. 4Lockheed Martin Corporation, NASA Johnson Space Center, Houston, TX, USA Session P28 Tuesday, May 20 The design of caution-warning signals for NASA’s 13:30 – 15:00 Topaz Lounge Crew Exploration Vehicle (CEV) and other future spacecraft will be based on both best practices POSTERS: SOFTWARE, INSTRUMENTATION, based on current research and evaluation of cur- AND MEASUREMENT rent alarms. A design approach is presented based upon cross-disciplinary examination of psychoa- 13:30 coustic research, human factors experience, aero- space practices, and acoustical engineering re- P28-1 An Anatomy of Graph-Based User Interfaces quirements. A listening test with thirteen for Media Processing—Christopher Schultz,1 participants was performed involving ranking and Jörn Loviscach,2 Shailendra Mathur,3 Jay grading of current and newly developed caution- LeBoeuf4 warning stimuli under three conditions: (1) alarm 1Universität Bremen, Bremen, Germany, now at levels adjusted for compliance with ISO 7731, mediaclipping, Bremen, Germany “Danger signals for work places—Auditory Danger 2Hochschule Bremen, Bremen, Germany Signals”; (2) alarm levels adjusted to an overall 15 3Softimage Corp., Avid Technology, Inc., dBA s/n ratio; and (3) simulated codec low-pass fil- Montreal, Quebec, Canada ¯

Audio Engineering Society 124th Convention Program, 2008 Spring 53 Technical Program 4Digidesign, Daly City, CA, USA, now at Imagine 13:30 Research, Inc., San Francisco, CA, USA P28-3 Delta-Sigma DAC Topologies for Improved Graph-based user interfaces are employed in a Jitter Performance—Ivar Løkken, Anders Vinje, variety of software such as audio synthesizers, Trond Sæther, Norwegian University of Science video compositing tools, and database applica- and Technology, Trondheim, Norway tion builders. All of these uses afford the graphi- Specifications for audio digital-to-analog-con- cal metaphor of a graph: “Nodes” such as sound verters (DACs) place requirements on the ana- generators or filters are tied together by “links,” log circuit design that contradict physical design which may represent signal flow or conceptual conditions in a modern, digital-oriented system relations. Focusing on media production tools, on a chip process. Because of low supply volt- we have examined a large range of current soft- ages, use of current-steering DACs has become ware products to find out which de-facto stan- the dominant choice for high resolution applica- dards have evolved in the field of graph-based tions. Fed by a delta-sigma modulator that interfaces and which features can be considered requantizes the digital signal to a manageable unique. We categorize a multitude of interface number of bits, the current-steering DAC is a concepts employed in actual graph-based inter- continuous time type converter without any dis- faces and describe differences in their imple- crete time filtering. This makes it very suscepti- mentation. The findings provide guidelines for ble to sampling clock jitter. In this paper jitter dis- developers of media production software. tortion is addressed at a topology level, Convention Paper 7495 investigating design choices for the delta-sigma requantizer and the possible use of semidigital 13:30 multi-bit current-steering filter DACs to reduce P28-2 A Framework for Automatic Mixing Using problems with jitter susceptibility. Timbral Similarity Measures and Genetic Convention Paper 7497 Optimization—Bennett Kolasinski, New York University, New York, NY, USA 13:30 A novel method is introduced for automatic mix P28-4 New Measurement Methods for Anechoic recreation using timbral classification techniques Chamber Characterization—Juan Gómez- and an optimization algorithm. This approach Alfageme, José Luis Sánchez-Bote, Elena uses the Euclidean distance between modified Blanco-Martín, Universidad Politécnica de Spectral Histograms to calculate the distance Madrid, Madrid, Spain between a mix and a target sound and uses a genetic optimization algorithm to figure out the As a continuation of the work presented at the best coefficients for that mix. The implementa- 122nd AES Convention (Paper 7153), this paper tion has been shown to successfully recreate tries to study in depth the anechoic chambers multitrack mixes accurately and may pave the qualification. The purpose of this paper is to find way toward the automatic mixing of novel multi- parameters that allow the characterization of this track sessions based on a desired target sound. type of enclosure. The proposal tries to obtain Convention Paper 7496 data of the anechoic chambers absorption by means of the transfer functions between pairs of microphones or by means of the impulse response between pairs of microphones. The

124 th Convention Papers and CD-ROM Pres 2008 ented a May 17 t the 12 – May 4th Con 20 • Am vention sterdam Convention Papers of many of the presentations given at the , The N etherla 124th Convention and a CD-ROM containing the 124th Convention nds Papers are available from the Audio Engineering Society. Prices follow:

124th CONVENTION PAPERS (single copy) Member: $ 5. (USA) Nonmember: $ 20. (USA)

124th CD-ROM (187 papers) Member: $160. (USA) 124th Convention 2008 May 17– Nonmember: $200. (USA) May 20 Amerstdam, The Netherlands

54 Audio Engineering Society 124th Convention Program, 2008 Spring results of the transfer functions between pairs of can also show similar nonlinear behavior. microphones can be easily checked by the Convention Paper 7500 agreement of the inverse squared law, allowing determination of the chamber cut-off frequency. Making a band filtering confirmed the anechoic Workshop 26 Tuesday, May 20 chamber’s qualifications. 13:00 – 14:30 Room L Convention Paper 7498 MIX YOUR OWN NEW YEAR’S CONCERT—PART 1 13:30 P28-5 Acoustic Feedback Reduction Based on LMS Chair: Florian Camerer, ORF – Austrian TV and Normalized LMS Algorithms in WOLA Filters Bank Based Digital Hearing Aids— Panelists: Jean-Marie Geijsen, Polyhymnia Raúl Vicen-Bueno,1 Almudena Martínez-Leira,2 Morten Lindberg, 2L Oslo - Oslo, Norway Manuel Rosa-Zurera,1 Lucas Cuadra-Rodríguez1 Ronald Prent, Galaxy Studios 1Universidad de Alcalá, Alcalá de Henares, Josef Schütz, ORF Radio Madrid, Spain A selection of experienced mixers with different back- 2Dimetronic Signals, San Fernando de Henares, grounds (classical music, pop, rock) have about 5-7 min- Madrid, Spain utes to set up a 5-channel mix of the current New Year’s Acoustic feedback phenomenon can disturb a Concert with the original multitrack recording. The audi- digital hearing aid performance at high gains, ence will be able to witness the individual approaches to causing instability in the haring aid and degrada- that task. A second session (30 minutes later) will com- tion in the speech. In order to restore a stable pare those mixes all aligned so that on-the-fly switching situation, an acoustic feedback reduction (AFR) is possible. The different approaches and results will be subsystem using adaptive algorithms such as discussed. the least-mean square (LMS) algorithm is needed. This algorithm has a reduced computational Tutorial 19 Tuesday, May 20 cost, but it is very unstable. In order to avoid this 14:00 – 16:30 Room N101 situation, another feedback reduction system based on a modified version of the LMS algo- LARGE ROOM ACOUSTICS rithm is used. Such algorithm is: the Normalized LMS (NLMS). These two algorithms are tested in Presenter: Diemer de Vries, Delft University of two digital hearing aid categories: the In-The-Ear Technology, Delft, The Netherlands and the In-The-Canal. These categories are selected because they have great feedback In the tutorial the physical principles of room acoustics effects, so robust AFR subsystems are needed. will be explained. It will be explained how, by measuring The added stable gain (ASG) over the limit gain or calculating impulse responses along arrays of receiv- when an AFR subsystem is working in the digital er positions, the temporal and spatial properties of a hearing aid is obtained for each category. The sound field can be analyzed and understood, and how ASG is determined as a trade-off between two these properties are related to perceptual quality cues. measurements: the segmented signal-to-noise Several models to calculate impulse responses will be ratio (objective measurement) and the speech discussed. Architectural as well as electro-acoustical quality (subjective measurement). The results measures to correct for acoustical shortcomings will be show how the digital hearing aids working with a proposed. All these aspects will be illustrated with exam- feedback reduction adaptive filter adapted with ples from practice. the NLMS algorithm is able to achieve up to 18 dB of increase over the limit gain. Master Class 4 Tuesday, May 20 Convention Paper 7499 14:30 – 16:30 Room B

13:30 A UNIVERSAL GRAMMAR OF CLASS D AUDIO AMPLIFICATION P28-6 Nonlinear Distortions in Capacitors—Menno van der Veen,1 Hans van Maanen2 1ir.bureau Vanderveen bv, Zwolle, The Presenter: Bruno Putzeys, Hypex Electronics Netherlands and Grimm Audio, The Netherlands 2Temporal Coherence, The Netherlands At first sight, class D amplifiers exhibit a baffling array of Many people have claimed that capacitors have disjoint design philosophies and topologies. Yet, all of a notable influence on the audible quality of these can be seen as permutations of the same basic systems. We have identified one of the major structure. This master class explores the “universal causes of nonlinear distortions in capacitors. grammar of class D audio amplification” the building Charging the capacitor will result in an attractive blocks and design choices underlying all class D force acting on the conducting plates. As no designs, like power stage topology, loop control strategy, material is infinitely stiff, this force will reduce the and modulation method. This generalized view makes thickness of the dielectricum and thus increase the clear how certain designs turn out to be uncontroversially capacitance. This process occurs in both phases suboptimal choices while some other apparently mean- of an AC signal in the same way and is thus non- ingless rearrangements like localized error feedback can linear. In this paper the consequences of this be demonstrated to have real performance advantages. process are discussed. It should be noted that other passive components like resistors and inductors Student Event/Career Development STUDENT DELEGATE ASSEMBLY MEETING —PART 2 Audio Engineering Society 124th Convention Program, 2008 Spring 55 Technical Program Tuesday, May 20, 14:30 – 16:00 Room O At this meeting the SDA will elect a new vice chair. One vote will be cast by the designated representative from each recognized AES student section in the European and Inter- national Regions. Judges’ comments and awards will be presented for the Recording Competitions. Plans for future student activities at local, regional, and international levels will be summarized.

Workshop 27 Tuesday, May 20 15:00 – 16:30 Room L

MIX YOUR OWN NEW YEAR’S CONCERT—PART 2

Chair: Florian Camerer, ORF – Austrian TV Panelists: Jean-Marie Geijsen, Polyhymnia Morten Lindberg, 2L Oslo, Oslo, Norway Ronald Prent, Galaxy Studios Josef Schütz, ORF Radio A selection of experienced mixers with different back- grounds (classical music, pop, rock) have about 5-7 minutes to set up a 5-channel mix of the current New Year’s Concert with the original multitrack recording. The audience will be able to witness the individual approaches to that task. A second session (30 minutes later) will compare those mixes all aligned so that on-the-fly switching is possible. The different approaches and results will be discussed.

56 Audio Engineering Society 124th Convention Program, 2008 Spring