CAS JOURNAL

In This Issue:

" DOWN HOME LUTHERIE 111 " DIRECTIVITY PATTERNS " INNOVATION IN VIOLINMAKING " ON THE OF THE " E STRING

Catgut Acoustical Society

To increase and diffuse the knowledge of and to promote construction offine stringed instruments

Vol. 3, No. 7 (Series II) May 1999 To our Readers...

This issue ofthe CAS Journal is being published without the guiding hand of its former Editor A. Thomas King, who has indi- cated that he wants to spend more time making and less time writing about it. It is a choice all of us can well under- stand, and we look forward to some great instruments coming out of his shop.

Tom has done an outstanding job of rejuvenating the CAS Journal and in attracting two Associate Editors of eminent stature. Less visible but equally important was his initiation of desktop publishing, which brought flexibility and shorter leadtimes to the more mundane aspects of semi-annual production. With considerable personal effort Tom has put theJournal on a new footing; he has our sincere gratitude and we wish him well.

The departure of our Editor occasioned a review of the direction and positioning of the Journal. We serve two distinct audi- ences: makers and scientists, — and to some ofour friends that dichotomy at times has seemed the Journal's undoing. It also is its strength: there is no other publication that attempts, issue after issue, to report and translate scientific findings into prac- tical guidelines for the stringed instrument maker. We will need to get better at that process of translation: we know that for many the Journal is often still too technical. We will be grateful for suggestions, such as those of John Soloninka (November 1998 issue).

We are not merely interested in publishing end-products ofresearch. There is also a need to provide all those engaged in this wonderful business ofexploring the secrets ofsound a greaterawareness ofprojects that are ongoing or maystill be in the plan- ning stage. George Bissinger's article in the November 1998 issue is an example. We intend to be activist, by fostering new research and helping to create newopportunities for collaboration. Our only constraint will be a healthy dose ofpeer-review, which has served us well in preserving the credibility of the CAS Journal and which forms the basis for Editorial discretion.

We are most fortunate to have Dr. Robert Schumacher and Gregg Alf as Associate Editors. Each is a highly respected leader in his field.Pending the appointment of a new Editor, and even thereafter, they will need the help and support of us all. Specificallythey need to be alerted to work being done in the various corners ofour musical acoustics world, which might even- tually lead to an article in this Journal. A simple note will do the trick!

This May 1999 Issue contains the usual mix of technical and maker-oriented articles. We trust you will find it of interest.

The CAS Journal is published twice a year by the Acoustical Society Inc., a non-profit organizationwhich aims to increase and diffuse knowledge of musical acoustics and to promote the construction of fine stringed instruments.

The annual fee for membership in the Catgut Acoustical Society is ndividuals 50.00 Institutions US$ 75.00 Students US$ 50.005( (for two years) An additional postage & handling charge of US$ 10.00 applies to memberships outside the USA.

CAS Journal back issues are available to CAS members only. Reprints of specific articles are availableto all interested parties For membership, back-issues andreprints contact the CAS 112 Essex Avenue, Montclair, New Jersey 07042, USA. Tel: 973.744.4029 (Thursdays only) Fax: 973.744.9197 Website: www.marymt.edu/~cas E-mail: [email protected]

US$

Office, CAS JOURNAL May 1999

Vol. 3, No. 7 (Series II)

The Catgut Acoustical Society is known for pioneer research in acoustical principles and the application of these principles to the making of fine stringed instruments, including the . To fulfill its mission, the Society supports publications, meetings for researchers and makers, musical compositions, lectures, and concerts.

DOWN-HOME LUTHERIE A III: Loudness History of a Page 2 5 Newborn Carolyn W. Field

Associate Editors Directivity Patterns of Acoustic Page 3 Gregg T. Alf Radiation from Bowed Violins ([email protected]) 9 by Lily M. Wang and Robert T. Schumacher Courtney B. Burroughs

([email protected]) : '- a*' 7■*-* ■; ,, ; * 333. : ... *» -: -" - : Page 34 Managing Editor Innovation in Violinmaking Elizabeth McGilvray 18 by Joseph Curtin Editoriol Advisory Board Daniel W Haines Page 40 Carleen M. Hutchins On the Acoustics of the A. Thomas King Violin: or Body Hill John T. Randerson 23 by Erik V. Jansson and Oliver E. Rodgers Benedykt K. Niewczyk Page 42

E String Whistles 28 by Bruce Stough Page 44

CAS Journal (ISSN 0882-2212) is published semi-annually by the Catgut Acoustical Society, Inc., 112 Essex Avenue, Montclair, NewJersey 07042. Neither the Society nor the Journal's editorial staff is responsible for facts and opinions expressed in articles or other materials contained in the Journal. Copyright 1999

' -7 *, ;;:.:, ; Carolyn W Field is a graduate of Swarthmore College (BA 1948) and the University of Houston (MA 1973). She worked as a student andthen colleague of Carleen Hutchins between 1977 and 1989. Since 1980 she has made stringed instruments in her shop in Oak Ridge. She is in the process of completing instrument numbers 25-28.

Courtney B. Burroughs is an Associate Professor of Acoustics in the Graduate Program in Acoustics at Pennsylvania State University. His research interests include the acoustics of musical instruments, , structural acoustics and transportation acoustics.

Lily Wang recently completed her PhD studying the radiation mechanism from bowed violins at the Pennsylvania State University Graduate Program in Acoustics. She is currently pursuing postdoctoral research on architectural acoustics at the Technical University of Denmark in Lyngby, under the 1998-99 Acoustical Society of America Hunt Postdoctoral Fellowship.

Joseph Curtin studied music and philosophy at the University of Toronto before becoming a violinmaker under the guidance of Otto Erdesz. He specializes in building new violins and , as well as collaborating with the University of Michigan physicist Gabriel Weinreich on various research papers.

Erik V Jansson is Associate Professor of Musical Acoustics at the Royal Institute of Technology (KTH), Stockholm, Sweden. After receiving his PhD, he spent a year as a research assistant at Case Western Reserve University, where he worked with Arthur Benade. In cooperation with N. E. Molin, Jansson has recorded vibration modes of the violin body and developed practical material tests with wooden blanks. Major papers can be found in the books Benchmark Papers in Acoustics (1975) and Research Papers in (1993).

Benedykt K. Niewczyk is a fourth generation violinmaker and manager, with his father, of the oldest andlargest workshop in Poland. He studied physics at the Adam Mickiewicz University in Poznan with a Polish MSc degree in Applied Physics. Since graduating from the University, he has worked full time making andrepairing iunstruments.

Bruce Stough has a BA in Philosophy from the University of Wisconsin (Madison) and an MS in Computer Science from the University of Minnesota (Minneapolis). He is a software engineer for Unisys, and also makes new violins in his own shop. He studied violinmaking with Edward Campbell.

Paul R. Laird is Associate Professor of Music History at the University of Kansas. He holds a PhD in Music from the University of North Carolina at Chapel Hill. His article "The Life and Work of Carleen Maley Hutchins" appeared in the Fall 1993 issue of Ars Musica Denver.

2 CASJ Vol. 3, No. 7 (Series II), May 1999 THE BAROQUE ALFRED PLANYAVSKY

Dr. Paul Laird University of Kansas 907 Christie Ct Lawrence, KS 66049-4148

of history can be most unset- family or da gamba family was the early nineteenth century, after the Studytling. As one learns more about a intended for a particular line of music, viola da gamba had almost completed its subject, gradually peeling back layers of Nowhere is this inspecifity of nomencla- slide into obscurity. The assumed physi- the onion, what once seemed firm truths ture more acute than for the bass instru- cal attributes of the modern orchestral become less clear and new certainties do ments, which, as Planyavsky demon- "string bass" are nothing more than a not always emerge. Alfred Planyavsky strates, were called abewildering array of generalization because, as Planyavsky helps destroy what one might believe is names. What each of these names actu- notes, some have the square shoulders the neat history of the lowest-sounding ally refers to sometimes cannot be and arched back of the violin, of the modern orches- known, partly because of the paucity of Planyavsky addressed the instrument's tra. He rebuilds part of that history in surviving instruments. Compared to history in detail in his Die Geschichte des this useful book, but other areas (where smaller instruments that are more easily Kontrabaßes (2nd edition, 1984), but an interested reader might desire clarity) protected, precious few large basses sur- felt the need to present its early history in remain murky. This is not really vive, making their history often a matter more detail. Planyavsky's fault because in places of conjecture. A recurrent theme in Planyavsky's available evidence is scanty, but at other Planyavsky calls the object of his his- study is that the term violone referred to times one might wish he took a less tory the "human-sized" bass string a number of different human-sized bass- polemical approach and admitted that instrument, a useful distinction given the es with different tunings and physical evidence is anecdotal. One can, for frequent characterization of the cello and characteristics. His consideration of the example, cite several existing instru- bass viola da gamba as bass instruments, violone covers the sixteenth through ments from a particular era or place, but These instruments play 8-foot pitch, eighteenth centuries and music-making any conclusions about those instruments however, and Planyavsky for the most in Italy, Germany, France, and elsewhere, require speculation. part is interested in bass instruments that Topics of special interest include: the Much of the volume concerns itself play at 16-foot pitch. early history of the violone in Italy; with the violone of the seventeenth cen- What is always said about the string 's extensive commen- tury, a messy period in the history of all bass of the , of course, is that it tary on large bass stringed instruments bowed string instruments, including a is not a member of the at from the early seventeenth century; the number ofinstruments that later fell into all, but a remnant of the viola da gamba role of the violone as a melody bass disuse. During most of the century the family, seen in its tuning infourths rather instrument in trio sonatas; Corelli's use viola da gamba familyremained popular, than fifths, its sloping shoulders, and flat of the violone; its role in the music of especially in England, even as the violin back. Problems with this simplistic Handel and Bach; and the violone as a family began to dominate European description abound. are tuned in solo instrument in the late eighteenth orchestral and , fourths, but with a third between the century. Terminology for string instruments in middle of six strings. According to Planyavsky is a double bass player the seventeenth century was frustrating- Planyavsky, the human-sized bass's tun- and an advocate for the importance of ly vague, sometimes even leaving in ing in fourths existed earlier but only the human-sized string bass instrument doubt whether a member of the violin started to become the in throughout music history. He was a

CASJ Vol. 3, No. 7 (Series II), May 1999 3 Laird - The Baroque Double Boss Violone

member of the Vienna Boys, learn- instruments. He might cite what for "Berlioz, Wagner, Bruckner, Stravinsky, ing double bass as a youngster. He many was a preferred combination, but and Strauss developed an orchestral cul- played with the Vienna Philharmonic by the 1 670s the cello was emerging as a ture that challenged double bassists to and Vienna State Opera, taught double favored instrument in Bologna and other rethink their traditional role as thefoun- bass at the Institute of Vienna Choir cities, and it surely was also sometimes dation of the orchestra." (p. 143) With Boys, and played violone with Nikolaus heard on bass lines without a violone. no further explanation, one is left to puz- Harnoncourt's Concentus Musicus Wien. It is fascinating to read Planyavsky's zle why those particular are He is also one of the world's leading description of the many instruments that cited, and why names like Beethoven, authorities on the instrument's history, bore the name violone. One of the most Brahms, Mahler, and others do not author of the abovementioned book and commonwas a large member of the viola appear. over 200 articles, essays, and reviews. da gamba family with and six Still, this book has much to recom- His narrative is spirited and some of his strings tuned an below those of mend it. A rich history of the violone is conclusions seem almost predetermined the tenor : Gl-C-F-A-d-g. This developed both in prose and in 58 illus- to favor the side of human-sized bass large bass would often have been played trations. All are black-and-white, and a instruments, but he has researched the in tandem with a smaller bass, either a few are almost too small to see the subject extensively in iconography, trea- cello or viola da gamba, or possibly intended image, but just from the illus- tises, and other sources. another instrument closer in size to the trations one derives a rich appreciation of An example of one of these conclu- double bass. Any of these combinations the importance of the human-sized bass sions is Planyavsky's insistence that the would have produced both 16-foot and from the early sixteenth century. melody string bass instrument that 8-foot on a bass line. A rather Documentation is extensive and covers played the continuo in most Italian trio surprising instrument was the fourth- works in several languages. The book is sonatas of the seventeenth century was third violone, popular in Viennese music written in a dry but readable style, and the violone, not the cello. He also notes in the second half of the eighteenth cen- James Barket's translation seems to work this was the case in much of Germany tury. With five strings and tuned Fl-Al- well. It is an important work to have during the century, as well as elsewhere D-F#-A, the instrument was often used rendered in English, a step that fails to in Europe. Some will find this surprising in a solo capacity. Wolfgang Amadeus happen with many useful musicological because the cello's role as a Baroque con- Mozart, for example, wrote an obligato works written in German. This is an tinuo instrument has long been generally partfor the instrument in his concert aria important read for anyone interested in accepted, but many in the early music Per questa bella mano, K. 612. Many the early history of stringed instruments. community know that Planyavsky might praised the instrument's light and trans- Readers of this journal might especially be right. Planyavsky believes that the parent sound and how effective it was in benefit from the appendix, fifteen draw- cello was not integrated into Italian fast passages. ings of various forms of human-sized music until around 1700. It clearly hap- Planyavsky usually seems on safe basses from the sixteenth century ren- pened earlier than this in some places, historical and rhetorical grounds, impor- deredby Bernhardt van Kampen. but one cannot ignore all of the trio tant when he takes issue with the work of sonata publications advertising music for many previous scholars. Occasionally, two violins, violone, and harpsichord. however, he missteps. When trying to Publication Information Sometimes violoncello is listed as another demonstrate the presence ofhuman-sized Alfred Planyavsky. The Baroque Double possibility. Earlier generations ofscholars basses in Lully's orchestra, he cites Jean- Bass Violone. Transl. James Barket. have misunderstood what is meant by Jacques Rousseau, writing 81 years after Lanham, Maryland and London: the term violone, incorrectly assuming it Lully's death. In Chapter 8, after consid- The Scarecrow Press, Inc., 1998. meant the violoncello. Some might find ering a mountain of evidence from the 194 pp. $55.00 it jarring to think of their favorite trio eighteenth and nineteenth centuries, sonatas played by two violins, a human- Planyavsky delivers the conclusion that The opinions expressedare those ofthe review- sized bass instrument, and harpsichord, the three-string double bass was seldom er and are not necessaryily endorsed by the but it was surely common. Later, after used before 1700. This might be the Editorial Board of the CASJournal. about 1700, the cello and violone often intention of the passage, but one ques- doubled on the bass line, playing one tions the necessity of citing all the evi- octave apart. Planyavsky, however, dence of the instrument's use after 1700. seems to dismiss some worthwhile He opens Chapter 9, a description of the doubts, because such pieces were played violone/double bass since the nineteenth by many different combinations of century with a remarkable statement:

4 CASJ Vol. 3, No. 7 (Series II), May 1999 DOWN-HOME LUTHERIE III: A Loudness History of a Newborn Cello

Carolyn W Field 13 Moore Lane Oak Ridge, TN 37830

A newly constructed cello is testedfor loudness before and after varnishing.

a sociable violin maker in a town up stroke on each note lasted about 3 1/2 varnishing, the have been com- Asfull of cultured and scientifically seconds with an equal wait before the pared five times. In Table 1 the results of minded people I often have to answer next stroke. Each test consists of three the nine tests are given in one column, in questions about the craft. For the more scales on the instrument in question, another a description ofphysical changes common questions I have of course pre- alternating with three scales on another to the instrument and in the last column pared stock answers. When asked about cello which serves as a control for all cello some subjective comments. the effect ofvarnish on a new instrument tests. The loudnesses of the notes in the The loudness pattern revealed here is I usually reply, "When it is first varnished scales are recorded as the Voltage Sums clearly defined. Cello 29 was more pow- it is not as good as it was in the white but and compiled by computer. The voltages erful before varnishing than after. The after a couple of years when the varnish is ofthe notes in each scale are totaled and overall change in the ratios of the Voltage well dried the instrument returns to its the tesults for each set of three scales are Sums given in Table 1 is about 30%. original quality." I don't remember averaged. The ratio of the average This is much greater than the 3% - 5% where I got this answer, except for the Voltage Sums for the two instruments is precision of the method. drying time of varnish which came from then calculated. In the work referred to No normal adjustments have had the Hutchinses, and I don't even know above (Field & Field, November 1997) it any perceptible effect on the results. The that it is true. So when recently I com- was shown that the precision of such soundpost was moved several times, but pleted my third cello I realized that I had ratio values is 3% - 5%, as measured by if this had any objective or subjective a unique opportunity. I could test the the standard deviations ofreplicates. effect it was too subtle for our test and instrument before and after varnishing The new cello being investigated is my ear. Strings were changed without and collect some quantitative informa- my No. 29 (29th instrument, third and effect, though admittedly to others of the tion about at least one aspect of its qual- last cello.) It was assembled and strung same brand and presumably tension. ity, its loudness. up on 5 September 1997 and of this writ- The bridge was slightly trimmed with The test method has been thorough- ing is about four months old. The con- only temporary subjective effect. The ly described in recent articles. (Field & trol cello is my No. 17, completed in wolf suppressor made the player happier Field, May 1997 and November 1997) In 1991 and in continual use since then. but did not change the test results. The short, the experimenterplays on the test Cello 17 is in all opinions a very good is the only open question. It instrument a chromatic scale starting cello and rather deceptively powerful. was shortened, unwisely, immediately with the lowest open string, the goal Both cellos are made of similar Sitka after the first test and maybe involved in being to play as close to the bridge and as and Big Leaf . The arches the drop in output between the first and loudly as possible without breaking the are nearly identical although the outlines second tests. If it is, however, this is not sound. In the present work the range of differ somewhat. due to any mechanism that I know the scale was two and a fifth from The instruments werecompared four about. Ao and Bo modes (my bass singer C at 65 to G at 292 cps. Bow tension times before varnishing ofCello 29- (The gave up on W) were not very close on 6 was set at 300 grams, the bow length parts had been previously sealed and pro- September. They certainly did not cou- was marked at 15 inches and the down- tected with onecoat ofoil varnish.) Since ple between 9 September and 6 October

CASJ Vol. 3, No. 7 (Series II), May 1999 5 Field - A Loudness History

Figure 1 ■ Cello No. 29: 6 September 1997 and I doubt if they couple now. A major operation was performed on the finger- board during varnishing but no minor dressing and smoothing seems to have affected loudness since that time. Somewhat as an afterthought the results of the nine cello tests were com- piled and printed as spectra of the volt- age output of the individual notes. I was curious to see possible changes in distri- bution of loudness over the course of the experiment. The accuracy of these Voltage Sums for individual notes is somewhat less than that of the total Voltage Sum. Each number represents, Pitch after all, the average of only three hand fingered and bowed notes. But some general tendencies can be noted. The first spectrum, that of the test made on 6 Figure 2 ■ Cello No. 29: 18 September 1997 September, shows Cello 29 at its strongest relative to Cello 17. (Fig. 1) The test of 10 September is very similar except that output seems to be shifting from bass to treble. This is even more apparent in the test of 18 September. (Fig. 2) The spectrum recorded on 13 November, two days after completion of the varnishing process shows reduced peaks and an extremely smooth upper register. (Fig. 3) This spectrum may reflect the stiffness I felt in the instru- ment on that date. The wolf suppressor introduced confusion into the pattern of 3 December though it did not change the relationship of the Voltage Sums of Cello 29 and Cello 17. But by 30 December the expected peaks around the octave F's Figure 3 ■ Cello No. 29: 13 November 1997 had reestablished themselves as had the treble peaks seen in Figure 2. (Fig. 4). For comparison with Cello 29 in its rapidly improving state I include the spectrum of Cello 17, tested on 30 December at the same time as Cello 29- -(Fig. 5). Remember that Cello 17 is a seven-year-old instrument that has been much played. The of the two instruments differ but their strengths are practically identical. I hope in the future to see if any changes occur in Cello 29 as it matures. If possible we will test the two cellos again in a yearor two when the varnish is CC#DD#E FF#GG#AA#B CC#DD#E FF#GG#AA#B CC#DD#E FF#G Pitch presumably dry. In the meantimewe can

6 CASJ Vol. 3, No. 7 (Series II), May 1999 Field - A Loudness History be reasonably sure that, given my mate- Figure 4 ■ Cello No. 29: 30 December 1997 rials and methods the first half of my stock answer is correct. ■ CASJ

ACKNOWLEDGMENTS

I wish to thank Frank H. Field whose curiosi- ty, ingenuity and technical knowledge made this work possible.

REFERENCES

Field, Carolyn W and Frank H. Field. C Of D D# E FF#GG#AA#B CC#DD#E FF#GG#AA#B CC#DD#E FF#G 1997. "Down-Home Lutherie: a sim- Pitch ple technique for measuring loudness," CASJournal (May): 29-36. Field, Carolyn W and Frank H. Field. Figure 5 ■ Cello No. 17: 30 December 1997 1997. "Down-Home Lutherie II: a computer-based technique for measur- ing integral loudness and observation of a bow-velocity independent loudness quantity," CAS Journal (November): 44-51. Spear, Deena Z. 1987. "Achieving an Air/Body Coupling in Violins, Violas and Cellos." CASJournal(May): 4-7.

CC#DD#E FF#GG#AA#B CO#DD#E F F# G G# A A* B GC#DD#E FF#G Pitch

ERRATA

In the May 1997 issue of the CAS Journal two corrections should be made to the article by Oliver Rodgers on the "Effect of Adjustment":

On page 22, right hand column, sixteen lines from the bottom, "Figure 7" should read "Figure 6".

On page 24 the captions on Figures 5 and 6 should be reversed

CASJ Vol. 3, No. 7 (Series II), May 1999 7 Field - A Loudness History

Table 1 ■ Chronology of Cello No. 29

Date(1997) Ratios of Voltage Changes Comments Sums 29/17 6 September 1.31 First setup Now that is a cello! But it has a bad wolfontheD string as wellas on theG. 9 September Fingerboard shortened 18 mm, raising Bq Cello is smoother, wolfreceding mode from 94 to 100Hz 10September 1.14 13 September Expertcellist playedit andlovedit. 18 September 1.14 Experimentswith weightsonfingerboard. A0 is 87-90 Hz. 18 September to I played it in quartet and orchestra re- 5 October hearsals; likedit so muchI almost decided not tovarnishit at all. 5 Oct. 1.18

6 October to 11 A slave to convention, I applied 6 or 7 November coats of commercial oil varnish. Finger- board was trimmed until Bq reached 90/91. (See Spear 1987for a discussion of adjustingfingerboards.)

13 November 0.99 Secondsetup It is stiff in bass and shrill in treble, a real disappointment.

8 CASJ Vol. 3, No. 7 (Series II), May 1999 DIRECTIVITY PATTERNS OF ACOUSTIC RADIATION FROM BOWED VIOLINS

Lily M. Wang and Courtney B. Burroughs Graduate Program in Acoustics Pennsylvania State University EO. Box 30 State College, PA 16804 E-mail: [email protected]

Directivity patterns ofacoustic radiation have been measured in thefar-field ofa violin, excit- ed with an open-frame mechanical bowing machine. Analysis of the directivity patterns confirms that, atfrequencies below 600 Hz, the violin radiates omnidirectionally, while above 600 Hz, cer- tain trends are apparent as the patterns become increasingly complex. It is notedthat when differ- ent strings are excited, thefar-field radiation patterns observed at nearly the samefrequency are similar, even in higherfrequency ranges where modal overlap is high. When the difference infre- quency between two directivity patterns exceeds somefraction ofa semitone, though, the measured radiation patterns differ significantly. This is demonstrated quantitatively by computing an rms difference around thepolar plots betweenpatterns. The sensitivity of the directivity patterns toper- cent changes in increases withfrequency.

make valid experimentalmeasure- different excitation are expected to be percent change in frequency did not pro- Toments on a violin, the instrument minimal in the lower frequency range, duce a significant variation in the direc- must be excited in a manner which is where radiating eigenmodes of the tivity pattern. At higher , consistent,reproducible, and not harmful instrument are distinct in frequency and however, that same percent change lead to the instrument (Jansson et al. 1986). modal overlap is low (Rodgers 1991). to notable pattern differences. These An open-frame mechanical bowing Modal overlap becomes greater, though, results corroborate Weinreich's machine, introduced below, has been at frequencies above 1000 Hz; thus, "Directional Tone Color" theory, which constructed for the task. This paper doc- changes in the excitation may influence states that violin radiation patterns vary uments measurements of the far-field the weighting of modes in the summed rapidly with frequency above 1 kHz, radiation patterns in two planes around response generated at a particular fre- impacting the subjective perception of the violin which were performed while quency, and subsequently the sound radi- sound produced by violins (1997). using the bowing apparatus. Certain ation. trends with frequency are evident upon To quantify the similarity between The Open-Frame Mechanical closer study of the directivity patterns, two directivity patterns, the rms differ- Bowing Machine These are related to directivity results ence between the patterns was comput- Ideally, the excitation of a violin from previous researchers. ed. This quantity has moreoverbeen uti- under a testing situation would simulate The measured directivity patterns lized to study at what percent change in real playing conditions. Assorted tests have also been contrasted to each other, frequency do the directivity patterns have been made while hand-bowing vio- first at matching frequencies produced begin to alter significantly. After initial lins (Saunders 1937, Jansson 1976, from bowing different strings to assess studies, it became apparent that this per- Rodgers 1993). However, it is in gener- the effect of the difference in excitation, cent change is itself a function of fre- al difficult to make reproducible meas- Variances in the radiated patterns due to quency. At lower frequencies, a certain urements with violinists holding and

CASJ Vol. 3, No. 7 (Series II), May 1999 9 Wang and Burroughs - Directivity Patterns

playing the instrument, as they naturally Reciprocal techniques have also been Figure 2 ■ A close-up view of the driv- adapt their playing to what they hear. In used, during which loudspeakers sur- en pulley system. The arm rotates so the search for acceptable alternatives, rounding the violin stimulate vibrations that different violin strings may be excit- researchers have in the past tried various of the instrument body (Moral and ed. A screw on top of the pulley assem- mountings and excitations (Hutchins Jansson 1982, Weinreich 1985a, 1985b). bly increases the separation between the 1973). Violins have been hung by rubber Other scientists have sought a con- driving and idle pulleys, thereby increas- bands, clamped at edges, fit in Styrofoam trolled bowing excitation, as some main- ing the belt tension. casings, and held loosely at the neck and tain that there is no substitute for abow- with foam. Both transient and ing excitation (Rodgers 1993), which steady state excitations of the instrument produces proper content and have been applied. The most common mechanical forces on the string transient excitation employs impact (Hutchins 1973). Raman utilized a bow- hammers (Marshall 1985, Bork 1993, ing machine which moved the violin Bailey and Bissinger 1997). Steady state back and forth across a fixed bow, while excitation, on the other hand, often Saunders' bowing machine from the involves electromagnetic excitation of the 1930s had a number of celluloid disks at strings or electromechanical shakers driv- the end ofan arm which descended upon ing the bridge, with input signals which the instrument (Hutchins 1983). Other range from single frequencies and swept bowing machines have employed real sine signals to maximum length bows controlled by computers and sequences (MLS) (Morset et al. 1998) and motors (Pickering 1991, Schumacher those using digitally stored 1994). curves (Weinreich and Causse 1986, Many of the above methods have Muller and Lauterborn 1996). been slightly problematic, as concerns have been raised about changing bound- ary conditions, loading, accurate Figure 1 ■ The open-frame mechanical driving of the bridge, and the amount of bowing machine. is at its The violin held energy input. Jansson et al. (1986) com- neck and tailpiece in the center of the pared a group ofholding and mechanical apparatus and driven by abelt sewn with excitation schemes, each with pros and horsehairs on apulley system. cons, not including bowing excitations. They concluded that much of the deci- sion about which method is best to use depends on what the desired output is, such as input admittance, cavity or body Consequently, to generatemost accurate- mode shapes, or sound radiation. ly the sum of modes which result from The desired output in the authors' real playing situations, a bowing excita- ongoing investigation is the radiated tion was deemed necessary. sound field from a violin, which results A new open-frame mechanical bow- directly from the vibration of the instru- ing machine was designed and construct- ment. The instrument's vibration is itself ed for the authors' violin studies. The a sum of normal modes, and the final apparatus mounts the violin in the center sum is affected by the method of excita- of an open frame where it is loosely held tion. This fact has been documented by at the neck with a foam pad and stands a number of researchers, including on a foam pad at the base of the tailpiece, Weinreich through the differences in the same locations at which a human radiativity from two excitation directions player may hold it (Figure 1). The open (1985a), and Kondo et al. whose results frame, which measures three feet by showed different holographic shapes for three feet by five feet (0.9144 m x the top and back plates of a violin when 0.9144 m x 1.524 m), is constructed different angles of string excitation at the from 1/2 inch black iron pipe shaft tub- bridge were implemented (1980). ing, connected by prefabricated fittings.

10 CASJ Vol. 3, No. 7 (Series II), May 1999 Wang and Burroughs - Directivity Patterns

Meanwhile, a driven pulley system, Figure 3 ■ The two planes along which spectra measurements were made: hanging down from the top of the appa- the horizontal plane, parallel to the plane of the bridge; and the longitudinal plane, ratus frame, runs a belt which is hand- parallel to the plane of the strings. Each plane is marked with degree angles match- wovenwith horsehairs, creating a contin- ing the directivity plots presented in this paper. uous bowing excitation. The belt has a polyester carcass with a polyurethane covering, and dimensions of 600 mm x 10 mm x 1 mm. Approximately 30 horse hairs are handsewn evenly across the 10 mm width, with the ends of the hairs glued to the interior side ofthe belt. The entry and exit points of the hairs are staggered along the belt's length. Arub- ber-based glue is used so that the bond does not become brittle and break when maneuvering around the pulley. The number of hairs is much less than the 200 ordinarily found on a bow; however, the sound produced is consistent with that from a normally bowed violin. The open-frame mechanical bowing apparatus allows for adjustment of vari- ous bowing parameters. The pulley mechanism may be rotated to attack dif- ferent violin strings (Figure 2). On top of the pulley assembly is a screw which alters the belt tension by increasing the separation between the driving and idle and one parallel to the plane of the narrowband magnitude spectrum from pulleys. The bowing or belt velocity is strings again at the height of the excita- 100 to 5000 Hz at each location; no increased or decreasedby varying the DC tion area, referred to as the longitudinal phase information was obtained. For the voltage input to the motor which drives plane (Figure 3). The instrument used measurements in the longitudinal plane, the pulley system. One can adjust the was a machine-made full-size Scherl and the bowing apparatus was placed on its bow distance from the bridge, as well, by Roth student violin (base model side so that the boom circled the violin moving the violin fixture up and down. #R270E4), rented from a local music on a plane parallel to the strings. Meanwhile, the force of the belt against store. Measurements were conducted with the violin string changes as counter- The bowing apparatus with violin the bowing speed and force set at typical weights are moved forward or backward was situated on a wire mesh which runs values of 0.35 m/s and approximately 1 on the moment arms atop the apparatus. across the mid-section of an anechoic N (Askenfelt 1989). Representative This force may be calculated by balanc- chamber. The chamber has dimensions spectra are shown in Figures 4 and 5. ing the moments around the pivotpoint. ofroughly 25 ft (7.6 m) on each side and Figures 4(a-d) depict sample spectra a cutoff frequency below 100 Hz. A 1/2 measured at 180°in the horizontal plane, Power Spectra Measurements inch omnidirectional B&K microphone directly in front of the violin's top plate; A series of power spectra and direc- was mounted on a boom which circled Figures s(a-d)wereproduced at the 180° tivity measurements were conducted the center of the violin in the far-field, position in the longitudinal plane, out while using the open-frame mechanical approximately 2 m away. This distance from the tailpiece. The results demon- bowing machine. Data were taken for ensured that the measurement was in the strate that the excitation was steady, pro- excitation on each of the four open far-field for frequencies down to 196 Hz, ducing power spectrawhich have strong strings in two different planes: one paral- or the of the open narrow peaks at the fundamental fre- lel to the plane of the bridge at the G string (Bies and Hansen 1988). The quency of the string and height of the excitation area (approxi- boom, which was computer-controlled, which are 30-50 dB above the back- mately 2.5 cm above the bridge), hence- moved the microphone in 5° increments ground noise level. forth referred to as the horizontal plane; around the violin, measuring an average

CASJ Vol. 3, No. 7 (Series II), May 1999 Wang and Burroughs - Directivity Patterns

Figure 4 ■ The power spectraobtained from bowing the a) open G string, b) open Directivity Patterns of the D string, c) open A string, and d) open E string, as measured in the horizontal plane Bowed Violin at 180°, directly infront of the violin. From the power spectra measure- ments, directivity patterns were formed b) Open D string specificallyat each fundamental and har- monic frequency in the 100 to 5000 Hz measurement band. The directivity pat- terns for the lowest four partials from bowing the open D and A strings are shown in Figures 6(a-b) for the horizon- tal plane and Figures 7(a-b) for the lon- gitudinal plane. Although the directivi- ty patterns ofevery harmonic up to 5000 0 1000 2000 3000 4000 5000 Frequency (Hz) Hz have been studied, they are not all presented here. d) Open E string Upon closer study, one finds that there are certain trends in the directivity ■20 patterns in the horizontal plane. At fre- T3s quencies less than 600 Hz, a. -40 the radiated co patterns are omnidirectional. Above that > -60 frequency range, between 800 and 1000 DC Hz, ■80 the directivity patterns begin to exhibit a cardioid-like pattern, with a "100 maximum out 0 1000 2000 3000 4000 5000 0 1000 2000 3000 4000 5000 oriented from the top Frequency (Hz) Frequency (Hz) plate of the violin (Figure 8a). At even higher frequencies, ranging from 1 100to 1600Hz, thepatternsresemble a skewed Figure 5 ■ The power spectra obtained from bowing the a) open G string, b) open doublet pattern, with nulls that spin D string, c) open A string, and d) open E string, as measured in the longitudinal clockwise as the frequency increases plane at 180°, directly out from the tailpiece. (Figure 8b). Above 1700 Hz, the pat- terns become increasingly irregular, some a) Open string b) Open D string with deep nulls forming full lobes, and others with shallow nulls (Figure 8c). -20 These directivity plots are asymmetric, CD and the largest lobes are not necessarily a! 40 a! -40 co CO directedperpendicular to the top plate. CD = -60 ■% -60 The measurements made in the lon- CU :3 CC gitudinal plane also demonstrate some -80 -80 interesting trends. Again under about

-100 100 600 Hz, thepatterns look omnidirection- 0 1000 2000 3000 4000 5000 0 1000 2000 3000 4000 5000 al. Between 600 and Hz, Frequency (Hz) Frequency (Hz) 800 however, the directivity resembles an ellipsoidal or c) Open string d)Open A E string egg shape, with the maxima oriented towards the upper bassbar and lower ■20 ■20 soundpost quadrants (Figure 9a). From T3 -40 _j 800 to 900 Hz, the violinproduces a car- i £L -40 CO CO dioid a CD m radiation pattern, with maximum S -60 > -60 directed towards 0° (Figure 9b). These a: rr forms change the of to 1400 ■80 -80 in range 950 Hz to, become a doublet directivity pat- -100 100 tern. In the lower part of this range, the 0 1000 2000 3000 4000 5000 0 1000 2000 3000 4000 5000 Frequency (Hz) Frequency (Hz) maxima seem to lie in the same quad-

12 CASJ Vol. 3, No. 7 (Series II), May 1999

G

" Wang and Burroughs - Directivity Patterns

Figure 6 ■ The directivity patterns of the lowest four partials from bowing a) the plexity. The other plane on which Meyer open D string and b) the open A string, as measured in the horizontal plane. The made measurements is similar to the lon- labeledfrequencies are rounded to the nearest whole number. The patterns are nor- gitudinal plane used in this paper, but malized to the highest of the measured levels for each string. with the violin in an inclined position as it is when played. Nevertheless, his results and the trends noted here are sim- ilar. Below 500 Hz, patterns are omnidi- rectional, while the ellipsoidal or egg shape dominates from 550-700 Hz with peaks in the lower soundpost and upper bassbar quadrants. Then at 800 and 1000 Hz, a cardioid develops. Further similarities exist at 1250 Hz with the skewed doublet pattern; above this fre- quency, the plots become more complex. Bissinger (1995) has also measured far-field directivity patterns in the hori- zontal plane. His results are omnidirec- tional below 600 Hz, radiate symmetri- cally outward from the top plate around Figure 7 ■ The directivity patterns of the lowest four partials from bowing a) the 800 Hz, and show a skewed doublet pat- open D string and b) the open A string, as measured in the longitudinal plane. The tern at 1158 Hz. Comparing Bissinger's labeled frequencies are rounded to the nearest whole number. The patterns are nor- longitudinal plane results to these trends malized to the highest of the measured levels for each string. also shows some correlation. There is roughly an omnidirectional behavior at b) Open A string 90 low frequencies, which becomes ellip- soidal and doublet-like up to 1 kHz. At 1038 Hz, the doublet pattern clearly emerges with the peak lobes oriented in the lower soundpost and upper bassbar quadrants. Consequently there is evidence that the trends denoted in this paper may apply to violins globally and are not unique to the violin tested or method used here.

Changes in Directivity Patterns with Frequency From the large number of directivity rants as those of the ellipsoid, while nulls Cremer 1984, Bissinger 1995). Looking patterns accumulated in this study at fre- are oriented close to 0° and 180° (Figure more closely at Meyer's seminal work quencies between 196 Hz and 5000 Hz, 9c). Above 1400 Hz, the patterns (1972), which shows histograms depict- it is possible to learn more about how become increasingly complicated with ing the principal directions of radiation directivity patterns of violins change many bumps, nulls, and lobes, although for a number of violins, one finds pat- with frequency. Weinreich has formulat- nulls near 70° and 295° were common terns that follow the trends outlined ed the idea of "Directional Tone Color" for some ofthe patterns (Figure 9d). here. In the horizontal plane, Meyer's which is based on violin radiation pat- Other published work, despite using results reveal mostly omnidirectional terns above 1000 Hz changing drastical- different violins, different mountings and behavior below 600 Hz, a cardioid pat- ly with frequency (1997). He compared different excitations, have produced sim- tern at 1000 Hz, and a skewed doublet the radiativity at two locations around ilar patterns which match the trends pattern at 1500 Hz. Frequencies above the violin, measured using a reciprocal shown here (Meinel 1937, Meyer 1972, 2000 Hz exhibit many lobes and com- technique, and concluded that patterns

CASJ Vol. 3, No. 7 (Series II), May 1999 13 Wang and Burroughs - Directivity Patterns

Figure 8 ■ Directivity trends in the horizontal plane: a) cardioid-like patterns emerge between 800 and 1000 Hz; b) skewed doublet patterns appearbetween 1100 and 1600 Hz; c) patterns become increasingly complicated above 1700 Hz.

c) above 1700 Hz 90

D string, 1161 Hz —A string, 1311 Hz ■D string, 1742 Hz D string, 870 Hz —-E-- string, 1334 Hz A string, 2620 Hz ]-- G string, 1000 Hz G string, 1588 Hz E String,3334 Hz

Figure 9 ■ Directivity trends in the longitudinal plane: a) ellipsoidal or egg shape sure levels around two measured patterns patterns emerge between 600 and 800 Hz; b) cardioid patterns appear between 800 werecalculated, and the data setwith the and 900 Hz; c) doublet patterns develop between 950 and 1400 Hz; d) patterns lower average value was multiplied by a become increasingly complicated above 1400 Hz. normalization factor so that its average level matches the other's. a) 600-800 Hz b) 800-900 Hz 90 90 The rms difference between the normal- ized patterns was then calculated as fol- lows:

A rms — i X I" 2n f

where n is the number of measurement points around the directivity pattern (n =72 in this paper), and Lln and L2n are the relative sound pressure levels of the two patterns at the nth measurement location, after normalization. This quan- tity is essentially an average dB differ- ence taken around the polar plots. Before examining changes in direc- tivity patterns with frequency, considera- tion was given to differences between directivity patterns at nearly equal fre- quencies when different strings were bowed. Harmonics with nearly equal frequencies were selected from excitation of different strings (Figures 10(a-c))3 (The harmonic frequency values are very close to oneanother but not precisely the differ betweenfrequencies that are even a quantitatively also. same, because each string was tuned by semitone apart. From the current data- Since the relative and not absolute ear.) As onecan see from the examples in base of planar radiation patterns, the levels are of interest for comparison, the Figure 10, the differences in the patterns degree of similarity between patterns at directivity patterns were first normalized caused by bowing different -strings different frequencies can be compared to each other. The average sound pres- appear to be small, with calculated rms

14 CASJ Vol. 3, No. 7 (Series II), May 1999 Wong and Burroughs - Directivity Patterns

Figure 10 ■ Comparison of directivity patterns at similar frequencies, obtained from bowing different strings: a) 1177 Hz from bowing D string and 1178 Hz from bowing G string in longitudinal plane; b) 1742 Hzfrom bowing D string and 1747 Hz from bowing A string in horizontal plane; and c) 2648 Hzfrom bowing D string and 2658 Hz from bowing E string in longitudinal plane. The rms difference, Arms for each case is provided.

a) D string 1 177 Hzvs. string 1 178 Hz (Arms=l .2dB) b) D string 1742 Hz vs. A string 1747 Hz (Arms=l .7dB) c) D string2648 Hzvs. E string 2658 Hz

180 1 80 1801

"D string, 1177 Hz ■D string, 1742 Hz ■D string, 2648 Hz --G string, 1178 Hz -- A string, 1747 Hz -- E string, 2658Hz

Figure 11 ■ Therms differencebetween directivity patterns as afunction of the per- differences around 4 dB or less. cent change in frequency for the base frequency shown, in the a) horizontal plane and Consequently, even in higher frequency b) longitudinal plane. ranges where modal overlap is high, the far-field radiation response due to a sum a) Horizontal Plane of eigenmodes is barely altered. Finding that the string excitation has minimal effect on the far-field patterns, the rms difference was thencalculated for Base frequency a wide variety of frequency combinations 581Hz on the horizontal and the longitudinal —A—" 1742 Hz planes, to determine how large of a -■"!&-- 2612 Hz in —X—3334 Hz change frequency is required before 4645 Hz significant changes in radiation patterns emerge. After preliminary study, though, it was apparent that the fre- quency change that produced a signifi- cant alteration in the radiation pattern is itself a function percentchange in frequency of frequency. Computing the rms difference as a func- b) LongitudinalPlane tion of the percent change in frequency for different base frequencies produced results as shown in Figures ll(a-b) for the horizontal and longitudinal planes. Base frequency These demonstrate that, at base frequen- — — 589 Hz cies below 800 Hz, the deviation remains A—1765HZ" consistently — 2648 Hz low, under 3 dB. This phys- —X—3531Hz ically stands to reason, since below 800 " 4651Hz Hz, the violin patterns are essentially omnidirectional with no lobes or strong nulls. With base frequencies above 800 Hz, one finds that the deviation can

0 5 10 15 20 25 30 increase up to values of 10 dB. percentchange in frequency One further notes that as the base frequency becomes higher, the range of

CASJ Vol. 3, No. 7 (Series II), May 1999 15

G

O

-&"-"--

O Wang and Burroughs - Directivity Patterns

Figure 12 ■ Comparisons of directivity patterns at high frequencies which are less ACKNOWLEDGMENTS than a semitone apart: a) 4645 Hz from bowing D string and 4667 Hz from bowing E string in horizontal plane; and b) 3469 Hz from bowing A string and 3531 Hz The authors gratefully recognize from bowing D string in longitudinal plane. The rms difference, Arms for each case Robert Kubli, Rich Hochron, Gory Elko and is provided. others at Bell Laboratories of Lucent Technologies for the use of their facilities a) D string4645 Hz vs. E string 4667 Hz (Arms=6.3 dB) b) A string3469 Hz vs. D string 3531 Hz (Arms=6.B dB) 90 90 and for their help in making the directivity measurements. Support for this research has been provided by a National Science Foundation Graduate Research Fellowship and a Bell Laboratories Graduate Research Program for Women 130 Grant.

REFERENCES ■D string, 4645 Hz ■A string, 3468 Hz -- E string, 4667 Hz D string, 3531 Hz Askenfelt, A. 1989. "Measurement of the bowing parameters in violin playing. II: Bow-bridge distance, frequency difference percentage in which all violins. dynamic range, and limits of bow Am., the rms difference remains under 4 dB Comparisons between the patterns force,"/. Acoust. Soc. vol. 86, no. becomes smaller. A small frequency obtained in this study show similarities 2: 503-516. change is more likely to result in signifi- between the directivity patterns at har- Bailey, M., and G. Bissinger. 1997. cantly different patterns at higher fre- monics of different strings which are "Measurement of direct radiation from quencies; the patterns are much more located at nearly the same frequency, violin excited by force hammer impact complicated at these higher frequencies demonstrating that the location of the at bridge," J. Acoust. Soc. Am., vol. so on that small shifts in frequency have bowed excitation has small effect the 101, no. 5, part 2: 3144. more drastic effects on the radiation pat- far-field radiation. Computed rms differ- terns. Even a change in frequency less encesbetweenradiation patterns further- Bies, D. A., and C. H. Hansen. 1988. than a semitone apart, which corresponds more illustrate that the changes in pat- Engineering Noise Control: Theory to approximately a 6% change of fre- terns are small with frequencies below andPractice (London: Unwin Hyman). quency, leads to significant differences Hz. However, for above 800 frequencies Bissinger, G. 1995. "Some mechanical between radiation patterns above 1000 800 Hz, radiation patterns alter signifi- and acoustical consequences of the Hz. Examples of such comparisons are cantly for changes in frequency as small violin soundpost," Acoust. Soc. Am., shown in Figures 12(a-b). as a semitone, corroborating Weinreich's J. vol. 97, no. 5, part 1: 3154-3164. "Directional Tone Color" phenomenon. Conclusions It has also been noted that the sensitivity Bork, I. 1993. "Impulse measurements Power spectra and directivity pat- of the radiation patterns to percent on violins," Proceedings of the terns have been acquired in two planes changes in frequencies increases with fre- Stockholm Music Acoustics Conference around a violin, while using a recently quency. ■ CASJ 1993, 355-360. constructed open-frame mechanical bow- Cremer, ing machine. The far-field directivity L. 1984. The Physics of the Violin, translated by S. Allen patterns have omnidirectional character- John (Cambridge, MA: MIT Press). istics at frequencies less than 600 Hz, then show distinct trends up to 1600Hz Hutchins, C. M. 1973. in the horizontal plane and up to 1400 "Instrumentation and methods for Hz in the longitudinal plane, becoming violin testing," J. Aud. Eng. Soc, vol. more highly varied and directional above 21, no. 7: 563-570. that. Directivity patterns published by others follow these trends, indicating that they may be generally applicable to

16 CASJ Vol. 3, No. 7 (Series II), May 1999 Wang and Burroughs - Directivity Patterns

Hutchins, C. M. 1983. "A history of Meinel, H. 1937. Saunders, F. A. 1937. "The mechanical violin research,"/. Acoust. Soc. Am., vol. Frequenzkurven von Geigen," Akust. action of violins," J. Acoust. Soc. Am., 73, no. 5: 1421-1440. Z., vol. 2: 22-33. vol. 9, no. 2: 91-98. Jansson, E. V 1976. "Long-time-aver- Moral, J. A., and E. V Jansson. 1982. Schumacher, R. T. 1994. age-spectra applied to analysis of "Eigenmodes, input admittance, and "Measurements of some parameters of music. Part III: a simple method for the function of the violin," Acustica, bowing," J. Acoust. Soc. Am., vol. 96, surveyable analysis of complex sound vol. 50, no. 5: 329-337. no. 4: 1985-1998. sources by means of a Mors A Krokstad and G. "Violin chamber," Acustica, vol. 34, no. 5: LH" " > J- Weinreich, 1985a. «' "A °- 275 280 Lokberg. 1998. computer-based radiativity: concepts and method for quality control of violins measurements," Proceedings of the Jansson, E., I. Bork, and J. Meyer. 1986. using impulse excitation," Proceedings Stockholm Music Acoustics Conference "Investigations into the acoustical of the International Symposium on 1983 Vol. 2, 99-109. properties of the violin," Acustica, vol. Musical Acoustics 1998, 23-28. " sum rule 62 no 1- 1-15 ' 1985b. Miiller, G., and W. Lauterborn. 1996. and the dipole moment of the violin," Kondo, M., N. Hirano, M. Hirota, and "The bowed string as a nonlinear J. Acoust. Soc. Am., vol. 77, no. 2: H. Kubota. 1980. "Principal dynamical system," Acustica united 710-718. directions of elasticity of a bridge on with Acta Acustica, vol. 82, no. 4: ■ 1997 "Directional tone belly and holographic maps of the 657-664. " color," sound box excited in those directions," , J' AcomL Soc Am voL 101> no - 4: Proceedings of the 10th International K l^h The Bowed String, 2338-2346. Congress on Acoustics 1980: Vol. 111, J^*'£ Weinreich, G., and R. Causse. 1986. SSJ-j- Rodgers,Rodeers. O. E. 1991. "Efferr"Effect on platenlare "Elerrrnnir"Electronic bows:hnwc rlifrifflldigital andanrl analog,"analncr " Marshall K D 1985 "Modal analysis frequencies of local woodremoval from Proceedings of the 12th International violln lates su of a violin," Acoust. Soc. Am., vol. P PPorted at the edges ConZ"s on Acoustics Vol. 111, K3-7. J. 1, >" 77 no 2- 695 709 Catgut Acoust. Soc. J., vol. no. 8: ' 7-11. Meyer, 1972. "Directivity of the J. 1993 "The limits of violin bowed stringed instruments and its ' " —late an overview effect on orchestral sound in concert P Proceedingstuni^' >" halls," J. Acoust. Soc. Am., vol. 51, no. of the Stockholm Music (" 1994 2009 Acoustics Conference 1993, 411-419.

CASJ Vol. 3, No. 7 (Series II), May 1999 17

"Über

'>

' ' ' INNOVATION IN VIOLINMAKING

Joseph Curtin Joseph Curtin Studios 205 North Main Street Ann Arbor, MI 48104, USA

The violin is a cultural icon as well as a working tool, and departuresfrom its traditional form have been variously regarded as impossible (it would no longer be a violin), unnecessary (the violin is already perfect) , and unacceptable (players would notplay it). Is itpossible to change, or even "improve" the classical violin? Where would onebegin? A number ofhighly qualified vio- linmakers andresearchers are investigating innovative approaches to violinmaking. While thecraft has withstood many similar efforts in the past, makers today are able to learn from acousticians, engineers, and material scientists, as well asfrom thefour centuries ofviolinmaking whichprecede them. In order to be successful, an innovation - whether in the acoustical, aesthetic, or ergonomic domain - must offer the violinist some tangible advantage overa more traditional instrument. In thispaper the author will consider what it is that musicians lookfor in a violin, then examine the possibilities ofchanging the traditional violin in order to give them moreofwhat they want.

violin emerged, in remarkably strings, making new violins look old, and bring to the workbench, Theperfect form, from the workshop of developing the highly sophisticated Fortunately, this is changing. An Andrea Amati in the early fifteen hun- restoration techniques needed to keep ever-increasing number ofviolinmakers - dreds. Since then it has changed only in the Old Italian instruments from collaps- and I think much credit for this must go detail. Stradivari and Guarneri del Gesu ing in exhaustion. to the Catgut Acoustical Society - are lis- developed the models now considered While the twentieth century has tening to some of the answers science is most successful for concert use. Within seen a vast increase in the scientific giving to the fundamental questions of fifty years of their deaths, the guild sys- understanding of the violin, we violin- the craft. At the same time, a number of tern in Italy disintegrated, and with it makers have to a large extent resisted highly qualified craftsmen have turned the legacy of workshop know-how these contamination by such understanding, their attention to innovation, guilds had harbored. For the next two This may be because it too often comes Violinmakers Christophe Landon, Roger centuries the thrust of violinmaking was couched in language opaque to us, or Lanne, and Guy Rabut have experiment- toward recapturing the "lost secrets" of because partial understanding -by cast- ed with new aesthetics for their instru- Cremona: the methods and materials, ing doubt on our day-to-day work habits ments. The synthetic bows of Michael and the understanding of how to use - can be paralyzing, or because our pro- Duff and Benoit Rolland have achieved them, that the old masters were pre- fession has a kind of over-developed wide acceptance by professional players, sumed to possess. The eighteen hun- immune system which rejects any ideas Research scientist and Charles dreds saw the introduction of the "mod- not coming from within its ranks. Or it Besnainou has laid the groundwork for ern setup" - longer necks, , may be that we prefer to believe our building superb stringed instruments string-lengths, and bassbars. And that, beloved instrument operates less by cold using carbon fiber composites. The above more or less, was that. Most of the pro- physics than by an ongoing succession of are people whose work I am familiar fession's ingenuity has since gone into miracles. For whatever reason, under- with. There are undoubtedly others, mechanizing production, creating better standing is too often the last tool we For my part, after building some 150

18 CASJ Vol. 3, No 7 (Series II), May 1999 Curtin - Innovation in Violinmaking

instruments on Guarneri and Stradivari kind of violin, one senses immediately humidity, crack-resistant, and for that models, I feel somewhat like a civil-war many possibilities for aesthetic changes - matter, less expensive. Anyone who has re-enactor - one of those people who changes which do not affect function in spent time in the violin world knows dress up in period costume and replay the least. There also seem to be a num- that, whether or not the violin is in some battles lost or won a long timeago. And ber of things one could change to make platonic sense perfect, real violins are so, to celebrate my twentieth year as a the instrument more comfortable to play, not. maker, I am devoting as much time as I more resistant to damage, and easier to At the same time, makers have been can to exploring different possible futures maintain. In this paper, I will consider trying to improve violins for a very long for violinmaking - or at least my own only innovation related to the sound of time, and so succeeding at it is to some violinmaking. the violin. Here, the issue becomes extent like shaving a few tenths of a sec- If I were to guess in which direction somewhat confusing. Are we trying to ond off the hundred-meter dash. One the violin is trying to evolve, I would say, make a better sounding violin, or aviolin senses some fundamental limit being in the same direction as it always has - with a completely new sound? I believe approached by ever-narrowing incre- toward a bigger sound. The lower arch- there is little chance of doing the latter, ments. On the other hand, if you want to ing introduced by Stradivari and at least without the aid of electronics. travel a hundred meters as quickly as Guarneri, along with the modern bridge, Any acoustic instrument relying on possible, running may not be the quick- bow, and setup, have all tended to maxi- bowed strings and played with normal est way. Are there alternative approach- mize powerand focus. I believe this evo- will sound, I think, like es to the craft that can perhaps change lution will continue, ifit can, because the a violin. I say this after listening to two the limits set by traditional designs and forces which originally fuelled it - larger extreme cases. Thefirst was a solid-body materials? halls, the greater volume of its frequent violin connected to an amplifier and accompanist, the piano, the competitive speaker. One hears the unfiltered string What is a good violin? instincts ofviolinists - are still very much signal, but the sound is still, unquestion- Let us define a good violin as one a with us. Added to this, the recording ably, a violin's - albeit an unpleasant one. good violinist loves to play. There are industry has increased the pressure on At the other extreme, when a CD of a many other possible definitions, but I players, I think, by raising listener expec- good violinist is played through believe this to be the most useful to mak- tation. Recordings allow, in exchange for Weinreich's DTC loudspeaker (1) - which ers and researchers. What then are the a certain loss of "realism," a kind of superimposes a as characteristics that good players respond intensity and presence to the sound that complex as that of the violin being to in a violin? Tone quality, projection, is almost completely lost in large halls. played - it is surprising how little differ- response, evenness, sensitivity to , Recordings allow one to hear the violin ence it makes: the difference between lis- and dynamic range are undoubtedly not as it sounds in a hall, but as it sounds tening to the violin in one room rather important. Projection and response I under the ear of the violinist. Purists than another, perhaps. From these and consider "absolute" qualities. By this I may well say such "close-miking" of the other informal experiments, such as lis- mean the more the better. Of course instrument is unrealistic, even vulgar. tening through an equalizer set every there will always be players who do not Personally, I love it. I love hearing violins possible way, I conclude that we are more want to stand out in an ensemble - ordis- in small rooms where the intensity of or less stuck with the sound of a violin. turb the family cat. Their needs are well sound can be almost painful. I love vio- The best we can hope to do, through met by existing violins, but I think they lins that project something of this inten- innovation, is maximize the things about would be still better met by violins with sity in large halls. And though I am not violins that good players and their audi- more projection. These can be strung an accomplished violinist, I love hearing ences seem to like, and/or get closer to with lighter strings, reducing the overall what a great violin does under my ear, our ownindividual tonal ideals. output and giving the bonus of increased being fully aware that some important There is a stale debate as to whether playability (2). I have neverheard a vio- portion of this experience is inaccessible it is, in principle, possible to improve on linist say, "This instrument responds too to the listener. It is part of a closed loop the best old violins. Can one imagine quickly." connecting player, bow, and instrument. something better than Perlman's Projection might be defined, quasi- But it is within this closed loop, I believe, Stradivari or Paganini's "Cannon"? From scientifically, as how loud an instrument that innovation in violinmaking will the violinist's point of view, I think the sounds to the listener for a given string either succeed or fail. answer is an easy yes. Make them more signal. This would seem to be related to powerful and faster responding. Make how loud it sounds to the player, but in Is innovation possible? them more even, less prone to wolf- fact there is an odd independence. There If one sets about designing a new notes, stable in the face of changing are violins which sound loud under the

CASJ Vol. 3, No. 7 (Series II), May 1999 19 Curtin - Innovation in Violinmaking

ear but do not project well, and those note shows up at the strongestresonance, body modes and the Helmholtz reso- which seem quiet under the ear but carry it can be assumed the bowed string is nance. very well. There are also, of course, giving about as much as is feasible. It is widely known that a vibrating instruments that are quiet or loud to Perhaps more could be taken if one plate radiates sound of a given frequency player and listener alike. In my experi- designed an instrument where the strong most effectively when the length of the ence violinists vary a fair amount on how resonance causing the wolf note is bending wave in the plate is equal to or much sound they want under their ear, replaced by several smaller resonances. larger than the of that fre- but not a lot in their requirement that The instrument could then be "pushed" quency in the surrounding air. If the the instrument project. further - until a wolf note began to speed of sound in air is 340 meters per The above definition of projection appear at the next highest resonance. second, then the wavelength of sound for relies on reports of a listener's experience, Another way to maximizethe energy 1000 hz is 34 cm. This is about the and so is not readily measurable. available for radiation is to minimize the vibrating length of a violin top, so if we However, projection can presumably be energy lost in other ways, such as inter- were able to arrange for the lowest mode related to both the total amount ofsound nal losses in the violin's vibrating compo- of the top be IOOOhz, then it, and all radiated by an instrument, and its spec- nents. The use of materials with lower higher modes, would radiate very well. tral composition. The frequency range will accomplish this, and by As the lowest mode for the supported top between about 2000 hz and 4000 hz has doing so yield an increased of of a traditional violin is in the neighbor- often been singled out as especially vibration for a given force input. But it hood of 600 hz, a considerably stiffer important, lying as it does in the region will also slow the response time. The than normal top would be required. of the ear's greatest sensitivity. In the transients will take longer to die away, An obvious way to achieve this is to remainder of this paper, however, I will and thus each note will take longer to get leave the wood thicker. Unfortunately, not worry too much about theparticular started. Rather than trade off on this stiffer but thicker plate will be more recipe required. Having claimed that response time, abetter strategy might be massive, thus decreasing the amplitude being heard is of primary importance to to reduce the mass of the vibrating com- of vibration. A more fruitful approach is most violinists, I will simply explore the ponents. For a given force input, a to use less dense wood. Even left the possibility of making more sound in all lighter body will reach a larger amplitude same thickness, its resonances would be registers. of vibration, while taking no longer to higher in frequency and its impedance get to this level. lower - a double advantage. Howfar can Building o More Powerful Violin Now, if we decide to build a violin one take this? The average density for Cremer (3) suggests that about 4% with lighter plates, the greateramplitude the European spruce used by violinmak- of the energy applied to the violin via the achieved implies greater bridge motion, ers is about .4 grams per c.c. It is possi- bow is radiated as sound. Before setting and with it the associated tendency ble to find wood much lighter than this - out to build a more powerful violin, one toward wolf notes. There is, however, a I have samples as light as .33, and even should ask whether it is possible to get way out - make a bridge that, for a given lighter with some North American more sound out of the bowed string than movement at the string notch, provides a species. My guess is that the average a good traditional violin does. This is an greatermovement at thefeet than does a density of the top wood of good old important question; if the violin already conventional bridge. In other words, instruments is in the range of .36 to .38. radiates as much energy as the string has change the effective leverage of the As this is easy to measure using a CAT to offer, there is little point in trying to bridge. This can be accomplished by scan, and as such scanning of instru- do better. making the bridge lower and/or wider. ments is now being done, I eagerly await Now, the regular vibration of the Beyond a certain point, both these a survey on the subject. string depends upon its oscillating ener- options necessitate changes in the tradi- Are there undesirable consequences gy not being lost too quickly. A violin tional design, but this, after all, is what to using less dense material? Yes, sever- drains energy from the string mainly via we set out to do. al. The plate becomes more susceptible motion of the bridge. When the bridge Before examining further the possi- to dents and abrasions. More important- motion - and thus the drain in energy - bilities of building lighter violins, I ly, the pressure on the soundpost causes it becomes excessive, a wolf note occurs. would like to touch on the radiation of to dig into and otherwise damage the Wolf notes are, in a sense, indicators to the instrument with regards to frequen- inside surface of the top. I installed a the maker that he is arriving at the limit cy. It is fair to look at high and low fre- veneer of maple about .25 mm thick in for the amount of power that can be quency radiation separately, as frequen- the soundpost area of an experimental extracted. As most traditional violins are cies above about IOOOhz are radiated by top made with very light spruce. It driven at least to the point where a wolf the top alone, and those below by whole worked so well than I now do this on all

20 CASJ Vol. 3, No. 7 (Series II), May 1999 Curtin - Innovation in Violinmaking

my instruments. It reinforces this other- sure, but an equally satisfyingone, I find. quency radiation is achieved by modes wisefragilearea, seems, if anything, ben- It offers the primal satisfaction of com- involving various contortions to the eficial to the sound, and can be easily bining ingredients and putting them into instrument's body. Because the body is removed. the oven, followed by the excitement of small in relation to the wavelength of A final consequence to using less seeing how they turned out. Besnainou sound at these frequencies, the extent dense wood is that, if the plates have reports (personal communication) that that these modes radiate sound is direct- been tuned to normal tap tones, they will this approach offers possibilities for ly proportional to the net changes in the be thinner than normal plates, and thus building instruments significantly lower volume of the body they engender. Such less able to withstand the considerable in weight than is feasible using tradition- volumetric changes dependon the asym- static forces on them. The result is an al methods. Furthermore, such instru- metry of the modes (5). Martin Schleske acceleration of the distortion that tends ments have so far proved stable overtime has developedpractical ways to optimize to occur, overthe years, in all violins. and insensitive to changes in humidity. the shapes of these modes. I think he has To get back to the frequency charac- made an important contribution to vio- Lighter Still... teristics of the violin, we can in principal linmaking methodology. Violinmakers often attempt to stiff- improve the radiation at high frequencies But standing back a little, I believe en wood with special varnishes or by using stiffer and lighter tops. I do not that one of the weaknesses (perhaps grounds. Although it is fairly easy to yet know how this might affectan instru- "characteristics" is a less provocative somewhat stiffen spruce across the grain ment's low frequency response, but word) ofviolin-family instruments is the this way, spruce is such a stiff material for imagine that changes in the total design relative paucity of modes at low frequen- its weight along the grain that there are would be necessitated. Such questions cies. This paucity should come as no sur- almost no natural materials you can aside, what can be done to help the radi- prise. The violin can be looked at as a apply that will increase the stiffness more ation at low frequencies? "flexing shell," and the modes of such than simply leaving theplate thicker. An It is widely known that the structures are on average evenly spaced engineer, looking at the problem, might Helmholz resonance is the lowest radiat- with regard to frequency. Musical pitch, suggestwe find a way to get the stiffness ing resonance on a violin. Weinreich (3) on the other hand, goes up in proportion to migrate toward the surfaces of the has pointed out that its strength in rela- to an exponential increase in frequency. plate, where it can do the most work. tion to the other modes can be changed Thus the low modes of a violin are rela- This can be done by putting a veneer of only with respect to its damping. A tively widely spaced with respect to the wood or other material on a very light decrease in damping, allowing larger violin's low notes. As not all of these core material, thus achieving a much amplitude, might be effected by round- modes radiate, the fundamentals and higher stiffness to weight ratio than is ing the edges of thef-holes. Interestingly, lower partials oflow notes are often poor- possible with a homogenous material. the edges of the f-holes of most old ly supported. The opposite is true at There is the further advantage that the instruments have been rounded with high frequencies, where each note has dense surface will resist damage more time, while violinmakers often pride relatively few partials in the audible effectively. If the surface material resists themselves on the crispness of their cut. range, and a great many closely spaced plastic deformation, then long term dis- Some simple experimentation would modes to support them. I think this tortion will become less of an issue. determine how significant such differ- explains why, at least to my ears, violins Charles Besnainou has developed ences are. More radically, onemight try differ in their highest registers more in workshop techniques for using layers of changing the f-hole's shape. The damp- their volume than in their tone color. At carbon fiber, plastic foam, and wood to ing of the is largely any rate, given the native variability of build string instruments. His system, determined, for a hole of given area, by wood, the complexity of the violin's presented in a workshop at ISMA9B, is the total length of the edges of the hole. structure, and the inevitable differences built around carbon fiber cloth that has An f-hole has rather long edges for its among hand-made instruments, it is dif- been impregnated with epoxy resin. This area. The least-damped hole would be a ficult to predict and control the radiativ- is layered with wood veneers and sheets circular. Given that the length and ity of the lower modes. Because they are offoam, then vacuum-formed into a cast placement of the f-holes are important in scarce, the stakes are high - each mode and cured in a specially built oven. None lending flexibilityto the bridge-carrying becomes crucial to the success of the of the equipment needed is particularly part of the top, circular f-holes don't instrument in a way that most of the expensive. More importantly from the seem feasible. However, a simplified f- higher modes are not. maker's point of view, it is a "friendlier" hole design might reduce damping some- I have for some time believed that process than one might expect - a differ- what. violins would, on average, sound better if ent experiencethan carving wood, to be The rest of the instrument's low fre- there were significantly more radiating

CASJ Vol. 3, No. 7 (Series II), May 1999 21 Curtin - Innovotion in Violinmaking

modes in the low range. This belief was efficient radiation. I suspect that an REFERENCES reinforced when I played the string signal instrument with a greater than normal from a solid body violin through density of radiating modes under 1000 1. Weinreich, G., "Radiativity revisited: Weinreich's aforementioned DTC speak- hz will have desirable musical character- theory and experiment ten years er. The speaker has a great many modes istics. later," Proceedings of the Stockholm in the violin's low range, and though it It must be remembered that the vio- Musical Acoustics Conference, 1993, introduces several not-violin-like quali- lin is a cultural icon as well as a working p. 434. ties to the sound, the richness and eve- tool, and iconoclasts - those who smash erm N- e owea rmg-> ness lent to the low and middle range is of icons - cannot expecta warm welcome ' S> ' Mattituck, N.Y, Amereon, 1991, remarkable. at the temple gates. At the same time, a I don't know how one might get a growing number ofplayers, frustrated by ■ >P"■ significantly greater number ofradiating the sheer amounts of money they are 3 Cremer L. Physics of the Violin modes in the low range of a traditional being asked to pay for often-mediocre Cambridge, MA, MIT Press, 1984, violin. Perhaps additional resonant ele- instruments, are becoming increasingly ch. 9, p. 203. ments could be incorporated into some open to innovation. The success of such Weinreich, novel design. I have several ideas in this innovation should be judgednot by look- 4. G. 1997. The Directional direction, and will report on them after ing back over our shoulders to Cremona, Tone Color Loudspeaker," Journal of further experimentation. but in terms of how well they both meet the Acoustical Society ofAmerica, 101, the needs of musicians and satisfy our p. 3071. Summary and Final Thoughts own evolving sense of tonal beauty. 5. Schleske, M. 1996. "On Making we hope new ■ If to create instruments CASJ 'Tonal Copies' of a Violin," CAS in tra- that some sense work better than Journal, Nov. p. 18-28. ditional ones, it is important to identify just what, in physical terms, we want ACKNOWLEDGMENTS themtnem to ao.do. Wewe otherwiseotnerwise risknsK our workworK Editor's Note' ending on an already up crowded shelf My thanks go to Charles Besnainou, Thispaper waspresentedby Mr. Curtin at the marked "Irrelevant Innovations." I for shoring his work with composites, to ISMA9B meeting June 1998 believe that better projection will always Xovier Boutillon, for our many conversa- be welcomed by musicians, and think tions about violin acoustics, and especially that this, and other benefits, will be to Gabriel Weinreich, for reviewing this gained by reducing the mass of the paper and for his many years of intellec- vibrating components of the instrument, tual generosity. while doing everything possible to ensure

22 CASJ Vol. 3, No. 7 (Series II), May 1999

1C ON THE ACOUSTICS OF THE VIOLIN BRIDGE OR BODY HILL

Erik V Jansson Benedykt K. Niewczyk KTH (Royal Institute of Technology) Violin Workshop Dept. of Speech, Music, and Hearing 6 1-776 POZNAN S-100 44 Stockholm, Sweden ul Wozna 6, Poland E-mail: [email protected] Tel: 48.61.852.5726

In measuredfrequency responses ofviolins ofsoloist quality a broad maximum is usuallyfound at about 2.5 kHz which has been called the "bridge hill" as the bridge has its major resonantfre- quency nearby. Experiments show that this name is misleading, since thefrequency andbandwidth of the hill can be influenced by the properties of the top plate between thef-holes as well as by the bridge.

Introduction "main wood resonance", and (ii) a A typical admittance curve from this Several studies of violin acoustics marked "hill" in the response (in this set of measurements illustrates these fea- have suggested that a feature which has case, the input admittance at the bridge) tures. Figure 1 shows the result, in been called the "bridge hill" is an impor- in the range 2.5-3kHz. However, in con- amplitude and phase, for a tant quality indicator. The first was an trast to the earlier results, these violins violin of 1709- investigation of the acoustical correlates tended to show a broad and rounded The two peaks labelled "PI" and of quality judgements by players (Alonso "hill" rather than a sharp one. "P2" are the major contributors to the Moral and Jansson 1982a). The most sig- nificant correlation was with the height tj- -\- \ l i i ° Figure *1 ■ frequencyresponse ofc a violin orc soloist qualityr with principal peaks PIm of the low-frequency response peaks, and P2 (A . Stradivarius 1709, cfjansson 1997). called the main wood resonance (Saunders 1946, Hutchins 1962) and nowknown to involve a cluster ofmodes of the body and internal air in the fre- quency range 450-600 Hz. The second correlate is the concern of the present paper. It was found that the slope of the amplitude spectrum of radiated sound in the range 1.5-3 kHz was important. Better instruments showed a marked increase of level, rising to a peak around 2.5-3 kHz. A later study (Duennwald 1982) also found high levels in this frequency range to correlate with high qulity of a violin. Finally, a study oftheresponse characteristics of 25 violins of soloist quality (Jansson 1997) found that these instruments shared (i) strong response in the region of the

CASJ Vol. 3, No. 7 (Series II), May 1999 23

■ ■ " ■ ■ Jansson and Niewczyk - On the Acoustics of the Violin

"main wood resonance" cluster. Their Figure 3 ■ Frequency responses (level and phase) of two bridges on violin S nature has been discussed elsewhere Niewczyk. The two responses are drawn in the same plots to simplify comparison. (Alonso Moral and Jansson 1982b, Resonant frequency of the normal bridge is 3.0 kHz (solid line) and of the special Marshall 1985). The "hill" appears as a bridge 7.6 kHz (dashed line). broad hump in theregion 2-4 kHz in the amplitude plot, and as a downward ramp in phase by approximately 180° in the same range. Both amplitude and phase responses have a large number of sharp peaks and dips superimposed on the broad "hill". The fact that the smooth level maximum is accompanied by a smooth ramp in phase is to be expected from fundamental relations between level and phase. The evidence of these various studies strongly suggests that it would be valu- able to understand how and when the "hill" arises, and which constructional features of the violin govern its frequen- cy and bandwidth. A series of experi- ments have been carried out to investi- gate these questions, and the results are somewhat surprising. In the early studies (Jansson 1990), nance strongly, producing a peak in the Is the "bridge hill" governed by the hill was called the "bridge hill". This admittance curve (c.f: Molin et al 1990). the bridge or the body? was because the most obvious interpreta- Forces through the bridge feet then drive For a first and very fundamental tion of the hill involves the lowest in- the body of the instrument, so that experiment a violin ofprofessional quali- plane resonance of the violin bridge, enhanced body response can also be ty (S Niewczyk) was selected. It was which occurs in exactly this frequency expected in this frequency range. The measured with its normal bridge and range for a nonnal bridge. In this reso- rocking motion of this bridge resonance with a special bridge, which was made nance the top part of the bridge rocks would tend to exert an oscillating from a solid wooden blank for a bridge from side to side relative to the lower moment on the body. This qualitative without any cutouts, c.f. Fig 2b. The part, flexing through the narrow waist explanation of the "hill" is thus consis- special bridge was tapered and had the region of the bridge, see Fig. 2 tent with the finding that the hill is not same height and arch as the normal (cf Reinicke in Cremer 1984 Fig. 9.10 seen when the violin is driven at the cen- bridge, but its feet were only 2mm high left). tre of the bridge, normal to the plane of and were shaped as straight columns. The oscillating transverse force from the instrument body. The normal bridge had its first in-plane the vibrating string will excite this reso- resonant frequency at 3.0 kHz (when its feet wererigidly clamped), while the spe- , cltu bridge „. . c , , . c _, .. . . x , , r had no resonances until 7.6 Figure ~I ■ rirst in-plane resonance of the bridge (after Reinicke) and b) sketch of i , r ° kHz, farc above the frequencyr of ther nor- solid tapered bridge without cutouts. i r i The input admittance of the violin was measured with both bridges, using a standard measurement technique (Jansson 1997). This employs impulse excitation with a hardtipped hammer of the violin bridge, while the violin is sup- ported in a horizontal position on two felt-covered supports. The resulting lat- eral bridge vibrations are recorded by a

24 CASJ Vol. 3, No. 7 (Series II), May 1999

,-,. i, ,TT it,, ■ Jansson and Niewczyk - On the Acoustics of the Violin

Figure 4a ■ Experiments with stiffen- erties but by some aspect of the body Figure 4b ■ Experiments with stiffen- ings in thef-hole region Positions of stiff- properties. Although the association with ings in thef-hole region. Step 1 all stiff- ening elements, direction of grains, and the bridge resonance seemed plausible, enings elements in place, step 2 the B marking bridge. "bridge hill" no longer seems an apposite "stamps"removed, step 3 bars cut to half name for this feature. width, step 4 ribs shortened to bridge width, step 5 ribs thinned towards ends, Experiments and step 6 all ribs removed. B marks A natural question after this result bridge. is: has the bridge no influence on the fre- quency response? A conventional bridge 2 was tuned to an unusually high clamped frequency (3.6 kHz) and fitted to the 3 same violin, S Niewczyk. The clamped resonance of this bridge was then tuned 4 down in steps and the input admittance 6 (amplitude and phase) of the violin was measured. The frequencies of the "hill" and the clamped bridge resonance at 2.0 kHz (the sharp peak a few hundred each stage are listed in Table 1. Hz above the phase shift frequency in the Thereafter the influence of the prime diagrams may be an artifact in the meas- suspect was investigated in another series urements). With the bridge tuned still coil excited by a very small magnet of measurements - the region of the top lower the phase step becomes a phase waxed to the corner of the bridge. plate between the f-holes. This region slope around 2 kHz. The phase response, The resulting frequency responses was partially stiffened by 3 mm thick peaks and dips, becomes smoother with a are shown in Fig. 3- The responses of the pieces of wood, which were removed in lower tuned bridge. Note that the two bridges on the same violin are super- steps, cf. Fig. 4. The input admittance change from step to slope occurs when imposed, the two amplitude responses in was again measured after each step. the clamped bridge resonance is tuned the upper diagram and the two phase lower than phase step frequency with a responses in the lower. In the phase Results normal bridge. response, both curves show a steep step The frequency responses for the In the amplitude responses the same passing zerodegrees at 2.3 kHz. The two bridge tuning experiments are given in low frequency filtering effect by the curves are so similar that they look like a Fig. 5. The phase response is the easiest bridge can be found, see Fig. sb. The single curve at a quick glance. The to evaluate and will therefore be dis- response becomes smoother at high fre- amplitude responses are also similar, cussed first. The three highest-tuned quencies with a lower tuned bridge. although the normal bridge gives a 5 dB bridges show broadly similar phase With the bridge tuned in the range 3.6 higher peak at 2.3 kHz and a somewhat responses - a major phase step at 2.4kHz to 2.7 kHz the influence on the hill is lower level just below 2 kHz. We are but a tendencyto a slope towards higher moderate but for the lower tuned bridge forced to conclude that the "hill" is not frequencies, see fig. sa. With the bridge it is considerable. When the bridge is governed in this case by the bridge prop- tuned a step lower, the phase step falls at tuned lower than the bridge hill then the hill becomes a somewhat monotonous slope (cf diagram 3 and 4 from the top). Table 1 ■ Bridge hill peak frequencies and resonant frequencies of the Thus the experiments show that the clamped bridge bridge does influence the frequency response. From the third to the fourth diagram (from the top) there is a large iecialbrid ;e ste: 2.7kHz 3.6 kHz difference, both in hill shape and phase Normalbrid 2.4 3.0 shift. »ecial brid ;e ste: 2.3 2.7 The phase responses from the per- 2.0 2.2 turbation experiments with stiffenings in 1.9 2.0 the f-hole region are given in Fig. 6a. The 1.7 1.8 top three responses show a fairly steep 1.5 1.5 step at closely the same frequency but

CASJ Vol. 3, No. 7 (Series II), May 1999 25 Jansson and Niewczyk - On the Acoustics of the Violin

Figure 5 a ■ Frequency responses Figure 5b ■ Frequencyresponses (level) Figure 6a ■ Frequency responses (phase) (phase) of a bridge with monotonously ofa bridge with monotonously decreased of a bridge on experimental violin KA decreased bridge frequency on violin S bridge frequency on violin S Niewczyk. with stiffenings as defined in Fig. 4b Niewczyk. Resonant frequencies of the Resonant frequencies of the seven (from top to bottom - step 1 to step 6). seven bridges isolated as in table 1 (3.6, bridges isolated as in table 1 (3.6, 3-0, 3.0, 2.7, 2.2, 2.0, 1.8, and 1.5 kHz from 2.7, 2.2, 2.0, 1.8, and 1.5 kHz from top top to bottom). tobottom).

exchanging bridges between two violins that the "bridge hill" was rather insensi- tive to the bridge properties and more sensitive to the body properties. Thus it seemed that the "bridge hill" labeling was rather misleading. More detailed investigation was then carried out. The "bridge hill" was shown to be rather independent on the bridge properties as long as the bridge was tuned to a "nor- the fourth at a considerably lower fre- between the bridge and the f-holes has a mal" or higher frequency. Investigations quency. The lowest two responses show a large influence. It is difficult to compare with stiffening of the top plate in the similar slope on the average. the bridge tuning effects with plate stiff- region between the f-holes proved that In the amplitude responses the ening effects. The influence of both is the influence of the stiffness in thisregion upper three diagrams show a fairly simi- considerable but it is difficult to predict if can be large. The experiments with lar, somewhat symmetrical hill at a fairly the stiffening effects are of "real" magni- bridge tuning showed that a bridge similar frequency, in line with the phase tude (the differences in bridge tunings tuned lower than "normal" may have a responses, see Fig. 6b. The lower three are clearly very large). considerable influence though. Thus we responses tend to show a maximum fol- conclude from our experiments that the lowed by a somewhat ragged but monot- Conclusion hill is mainly a body hilt, which is main- onous slope. The difference between the The main question of this investiga- ly determined by the properties of the third and fourth responses are again tion: Is the "bridge hill" governed by the violin body (presumably the top plate in large, i.e. the stiffness of the top plate bridge or the body? It was found by the f-hole region) excited by the force

26 CASJ Vol. 3, No. 7 (Series II), May 1999 Jansson and Niewczyk - On the Acoustics of the Violin

Figure 6b ■ Frequency responses (level) lins? Further investigations to identify Jansson, E.V. 1997. "Admittance of a bridge on experimental violin KA the structural causes of the "hill" are Measurements of 25 High Quality with stiffenings as defined in Fig. 4b being made, and perhaps it will be possi- Violins," Acustica Acta, vol. 83: 337- (from top to bottom - step 1 to step 6). ble to answer some of these questions in 341. the near future. ■ CASJ Jansson, E.V, Niewczyk, B.K. and Fryden, L. 1997. "Mobility at the ACKNOWLEDGMENTS violin bridge and bridge properties," Proc. Inst. Acoustics 15, May 1997, The paper has been reviewed and Vol. 19:5: 17-21. carefully edited by Dr James Woodhouse, Marshall, K.D. 1985. "Modal analysis of o help hereby gratefully acknowledged a violin," J. Acoust. Soc. Am. vol. 77: by the authors. The project has been sup- 695-709. ported by the Swedish Natural Science Research Council, the Royal Institute of Molin, N.E., Wahlin, A.O. and Jansson, Technology (KTH), the Swedish Institute E.V. 1990.*** "Transient wave and the Wenner-Gren Center Foundation. response of theviolin body,"/. Acoust. Soc. Am. vol. 88: 2479-2481. Moral, J.A., and Jansson, E.V 1982. REFERENCES "Input admittance, eigenmodes, and quality of violins,"(a), STL-QPSR 2- Cremer, L. 1981. Physik der Geige. /1982- 60-72 Stuttgart: Hirzel Vertag. English translation by J. Allen, The Physics of Moral, J.A. and J;Jansson, E.V 1982.*** the Violin, MIT Press, (1984); Part 11, "Eigenmodes, input admittance, and Ch. land 2.1. the function of the violin," (b) Acustica, vol. 50: 329-337. Duennwald, H. "ZurMessungvon , 1982. „ . . W Geigenfrequenzgangen," Acustica vol , 1973.* «"ÜbertragungseigenschaftenTT des couple of the bridge feet. It should be 51: 281-287. noted that it is common violin making Streichinstrumentstegs," Catgut Hutchins, CM. "The Physics Acoust. Soc. Newsletter No 19: practice to make parts of this region 1962.* of Violins," Am., 26-34. thicker than the rest of the topplate. Sci. November: 78-93. The response of the Stradivarius vio- Hutchins, CM. 1975. Benchmark Papers Saunders, FAFA. 1946.* "The Mechanical lin from 100 Hz to 5 kHz is shown in in Acoustics/5 Musical Acoustics, Part 1. action of instruments of the violin Fig. 1. In large the "hill" of the level (Dowden, Hutchinson & Ross, family," J.j Acoust. Soc. Am. vol. 17: response is rather smooth, but ragged edited by C. Hutchins). 169-186. with many minor peaks and dips, and its Saunders, maximum at approximately 2.5 kHz. Hutchins, CM. 1976. Benchmark Papers FA. 1953.** "Recent work on The phase response shows a slowly in Acoustics l6 Musical Acoustics, Part 2. violins,"/. Acoust. Soc. Am. vol. 25: decreasing, ragged slope. This "hill" was (Dowden, Hutchinson & Ross, 491-498. found typical for the average violin of edited by C. Hutchins). soloist quality (Jansson 1997). CM. Hutchins, 1997. Research Papers in Listed original papers can also be found if Assuming that the Stradivarius vio- Violin Acoustics 1975-1993, (J. Acoust. (noth- marked in Hutchins (1975), in lin had a "normally" tuned bridge Soc. Am., edited by C Hutchins). * ** ing exceptional was noted), then its Hutchins (1976), and *** in Hutchins smooth "hill" should be shaped by the Jansson, E.V 1990. "Experiments with (1997). body and presumably by the stiffness in violin string and bridge," Applied the f-hole region (wood, thickness, f- Acoustics vol. 30: 133-146. holes, and arching). Or is the smooth "hill" aproperty that is obtainedwith the aged wood of the good old Italian vio-

CASJ Vol. 3, No. 7 (Series II), May 1999 27 E STRING WHISTLES

Bruce Stough 2913 Princeton Avenue Minneapolis, MN 55416-1956 E-mail: [email protected]

Thispaper argues that the that can be bowed on an open E string is a torsional mode. The attributes of the whistle mode are compared to characteristics oftransverse, torsional and lon- gitudinal string modes, including such characteristics as fundamentalfrequency, excitation mode, volume, and the effects ofchanges in tension andstring diameter. It isfound that the whistle mode closely matches the expected torsional characteristics. The paper also describes an experimental setup which provides a direct observation ofthe whis- tle mode. A mirror on the string reflects a dot oflaser light in an arc about the string when the whistle is bowed.

the Summer of 1997, my violin acquitted when I reproduced the whistle mode. The table is followed by justifica- Inteacher called me with an E string on an E string tensioned and clamped tion for the listed attributes. Given the problem. She was working on a fast over a hardwood beam. characteristics of the string modes, it Klezmer band piece which had a slur should be possible to determine which from the first Don the A string to the String Modes column of attributes matches the charac- open E string. Often, instead of the open Strings tend to vibrate in three teristics of the E string whistle. E she got a high pitched whistle. This modes: from side to side (transverse), happened only on her good violin. Her along the length of the string (longitudi- Tension Attributes backup instrument never had this prob- nal), and around the axis of the string The relationship between tension lem. She left the offending instrument (torsional). The table below shows and frequency can be derived from the 1 with me to try to figure out what was important attributes of each type of frequency equations for the modes. happening. After some experimentation, I found that the whistle could be reproduced by Table l ■ StrinS modes slurring gently from the A string to the Longitudinal open E with afast bow. Once the whis- Transverse Torsional Frequency increases No change No change tle it will continue even when the Effect of a change in starts, tension with the squareroot of bow pressure is increased. The bowing the tension. point is not at all critical. The spectrum Effectof changein Frequency drops as No change No change of the whistle (see Figure 1) shows diameter diameter increases under constant tension. strong, clear peaks at about 4.800 kHz Expectedfundamental Varies with tension 4.796kHz 7.780 kHz kHz. The whistle is not terri- and 9-600 frequency when length is bly loud, but is a surprise when you get 330mm it instead of an expected open E. Couplingto the instrument High Low High There were three likely sources: the Expected excitation mode Across the Across the Along the bow, the string, and the instrument. The was acquitted when I found that I , bow formulaswere derivedfrom information in chapter BofMechanical Vibrations (Rao). I assumed that same whistle by bow- could produce the the stfing was firmly fixed at each end) which was confirmedby Gillan and Elliott (1989). The string ing the instrument with a rosined birch was treated as a thin uniform shaft to derive the torsional andlongitudinal formulas. Theformulas were dowel. Similarly, the instrument was framed to express theresult in , rather than in radians per second.

28 CASJ Vol. 3, No. 7 (Series II), May 1999

strin, strin; strin; Stough - E String Whistles

(1) Transverse Figure 1 ■ The spectrum of the whistle bowed on my violin number 37. n P co n 21 \p

where n is the harmonic number, / is the string length, P is the tension of the string, and p is the mass of the string per unit length.

(2) Torsional:

n G CO n 21 Vp

where n is the harmonic number, / is the string length, G is the shear modulus of the string, and p is the mass density of the string.

Frequency (Hz) (3) Longitudinal:

fl E Figure 2 ■ The CO spectrum of the whistle bowed on a 330mm string clamped to the n hardwood beam. 21 \p where n is the harmonic number, / is the string length, E is the Young's modulus of the string, and p is the mass density of the string. Since equations (2) and (3) have no term for tension, we should expect that only transverse modes are affected by changes in tension. String Diometer The mass term in equation (1) is given as the mass per unit length. Therefore, as the diameter of the string increases at constant tension, the trans- verse frequency goes down by the square root of the increase of mass per unit length. The density term in equations (2) and (3) are in units of mass density. Since mass density does not change as the string's diameter is changed, the torsion- al and longitudinal modes are not affect- ed by a change in string diameter.

3, CASJ Vol. No. 7 (Series II), May 1999 29 Stough - E String Whistles

Expected Fundamental Frequency Table 2 ■ Tension results As shown by equation (1), the funda- mental frequency of the transverse mode Bowedfundamental Rubbedfundamental is determined by the density per unit E strin tuned to 660 Hz 4820.7 Hz 7709.3 Hz length, the length and the tension. E strin tuned to 440 Hz 4863.8 Hz 7764.7 Hz The expected torsional and longitu- were cal- dinal fundamental frequencies Table 3 ■ String diameter results culated from equations (2) and (3) using the following nominal values: Bowedfundamental Rubbedfundamental Li ;ht E strin (0.246mm) 4874.6 Hz 7673.9 Hz llNewto 4855.7 Hz £ = 2.068-10 /m/ 2 Heav E strin (0.275mm) 7590.5 Hz

10 kink in the " 10 NewtoV bridge, as well as push-pull against the return of the transverse string G = 7.68 m 2 and bridge. The longitudinal mode reflected by the string termination. In , provides the same push-pull against the the case of a bowed torsional mode, the P~ />o4z /m3 nut and bridge. The torsional mode pro- release in each cycle would be caused by vides only a small twisting moment to the tendency of the string to twist back These values were taken from the the nut and bridge, as the mass of the after being twisted forward. The longi- 25th edition ofMachinery's Handbook, and string rotates around its axis. tudinal modes would be excited by have been converted to SI units. motion along the length of the string. Expected Excitation Mode Expected Coupling tothe Instrument The transverse and torsional modes Whistle Mode Attributes Both the transverse and longitudinal are excited by motion perpendicular to I examined the torsional and longi- modes should couple well to the instru- the string. In the case of the transverse, tudinal modes by mounting a plain steel ments. The transverse mode will provide this is the stick-slip mechanism, where medium weight E string 6mm above a a perpendicular rocking impulse to the the release of each cycle is caused by the stout hardwood beam. The string was terminated by 6mm thick aluminum plates firmly clamped to the beam. The Figure 3 ■ The spectrum of the rubbed longitudinal mode of a 330mm string clamping points could be moved to clamped to the hardwood beam. adjust the string length. When bowed lightly with a fast bow, the whistle could be reliably produced, and it occurred at the same frequency as with a similar string on an instrument. The longitudi- nal modes could be excited by rubbing rosin along the length of the string, and stroking it lightly with acloth. Figure 2 shows a typical spectrum of the bowed whistle, and figure 3 shows a typical spectrum of stroked longitudinal modes. In both cases the stopped string length was 330mm. (The data was collected using the FREQS.I spectrum analysis program, with a sample size of 16k and a sample rate of 44,100 samples per sec- ond.) Results are givenfor the longitudinal vibrations as well as the whistle mode we are mainly interested in. These are a cross check. If the longitudinal results Frequency (Hz) match the expectations from the table

30 CASJ Vol. 3, No. 7 (Series II), May 1999 Stough - E String Whistles

above, it should increase our confidence Table 4 ■ Summary of results in the results for the bowed whistle mode. Bowed whistle Tension Results Effect ofa chan ;e in tension Minuscule chan I made trials to measure the bowed Effect of a chan :e in diameter Minuscule chan whistle and rubbed fundamental fre- Fundamental fre luenc when len ;th is 330mm 4.820 kHz quencies with a normal weight plain steel Cou lin to the instrument Prett weak E string at normal tension, and also Ex »ected excitation mode Across the strin tuned down to 440 Hz. The string was clamped to the hardwood beam, with vibrating length of 330mm. The one third reduction in tension ings of the rubbed E string were in the normally needs a one second stroke resulted in a 0.89% change in the fre- mid- 80 dB range. Readings of the of about 420mm of hair. At 4.820 quency of the bowed whistle fundamen- bowed whistle were in the lower 70 dB kHz, this allows .087mm of hair to tal, and a 0.5% change in the rubbed range. This indicates that both the pass per cycle. longitudinal fundamental. See table 2. transverse and longitudinal modes couple Both the bowed and rubbed frequencies well to the instrument, but the bowed 3. One early whistle suspect was a high are essentially unchanged, compared to whistle does not. transverse harmonic of the string. the gross change in the transverse fre- This possibility can be rejected for quency. Excitation Mode Results two reasons. First, the whistle The transverse mode was easily frequency does not fall close to any String Diameter Results bowed, as was the whistle. The longitu- harmonic ofthe E string. Second, if I made trials with light and heavy dinal mode could be excited by rubbing the whistle was based on the seventh gauge E strings. The strings were along the string, as expected. or eighth harmonic of the string, the clamped to the hardwood beam with a bowing position would be critical. vibrating length of 330mm, and tuned to Summary of Results In fact, the bowing position is not at 660 Hz. See table 3. The characteristics of the bowed all critical. The bowed fundamental changed by whistle can be summarized as shown in 0.39%, and the rubbed fundamental table 4. 4. The whistle only occurs on the open changed by 1.09%. Both changes are These results match the attributes of E string. This is because your finger small, especially for the bowed whistle. the torsional mode in the first table tip adds enough torsional damping The somewhat larger change for the lon- above, and allow us to conclude that the on fingered notes to suppress the gitudinal mode can be accounted for if bowed whistle is the torsional mode of mode. the longitudinal impulse continues a the plain steel E string. short distance into the clamps at the In addition to these considerations, 5. The results for the longitudinal string ends, making the string behave as the following evidence adds confirma- modes is consistent with the if it is abit longer. tion: expected characteristics, indicating that the experimental methods are Fundamental Frequency Results 1 . If a wound E string is used, it is next valid, As shown in table 2, the bowed fre- to impossible to bow a whistle. This quency result varies from the expected happens because the winding addsadds Direct Observation torsional frequency by only 0.52%, and additional torsional damping, which The above results provide a good cir- the rubbed longitudinal frequency varies makes it difficult to sustain a cumstantial case that the E string whistle from the expectedvalue by only 0.21%. torsional vibration. is produced by bowing the torsional mode of the plain steel E string. Instrument Coupling Results 2. The fast bow required to produce a However, it has been long assumed that Readings of the bowed open E string whistle is consistent with the while the torsional mode must be taken on a violin, taken about .5 meters from frequency range of the whistle. For a into account in fully describing the nor- the front of the instrument with a Radio significant amount ofhair to pass the mal transverse bowing mode (Cremer p. Shack 33-2055 sound level meter, were string during a cycle, the bow would 112-135), the high damping of torsional in the upper 90 dB range. Similar read- have to movefast. I estimate that it modes in violin strings would prevent

CASJ Vol. 3, No. 7 (Series II), May 1999 31 Stough - E String Whistles

excitation of a pure torsional mode. Pickering (1985) and Gillan and Elliott (1989), meas- Figure 4 ■ The setup used to directly observe the torsional mode using a reflected using different methods, beam. ured the torsional damping of violin strings. Their results indicate that the plain steel E string generally has much lower damping than the other strings. The question is whether the damping is low enough to allow a bowed torsional mode. This can be settled by directly observing the motion of a bowed whistle to verify that it is indeed torsional. My first attempt to directly observe the whistle mode involved gluing a small mirror to the middle of a standard E string. I then planned to bow the whis- tle while shining a laser pointer on the mirror. The path of the reflected spot would then show how the string was moving. This failed, however, because the mirror added too much mass to the string and it no longer was possible to bow the whistle. It also was difficult to firmly mount the mirror to the very thin string. This was compounded by the high frequency of the bowed whistle. I avoided these problems by repeat- Figure 5 ■ The spectrum of the signal from a photo transistor in the path of the thicker laser beam while the whistle is bowed. ing the experiment with a much and longer string. This reduced the rela- tive mass of the mirror and reduced the whistle frequency, making the observa- tionmuch easier. I began by mounting a 0.78mm piano wire string horizontally on a one meter long woodenbeam, tightenedby a small turnbuckle. Once tight, the vibrating portion of the string was formed by clamping its endpoints to the beam between metal blocks, leaving a 785mm length free to vibrate. It was very easy to bow the whistle on this setup, and the fundamental of the whis- tle was measured (via FFT analysis of an audio sample) to be 1999 Hz. This is in close agreement with the torsional for- mula, which gives an expected funda- mental of 2012 Hz. To observe the motion of the bowed whistle, I glued one fleck of silver glitter to the center of the string to act as a mir- ror. I then aimed a laser pointer (Radio Shack LX1000) perpendicularly across

32 CASJ Vol. 3, No. 7 (Series II), May 1999 Stough - E String Whistles the string so that the reflected spot fell sis of this signal gave the same funda- REFERENCES on a piece of paper which was 200mm mental frequency as the audio sample, as from the string and mounted broad side shown by figure 5. Cremer, L. 1984. The Physics of the to the string. See figure 4. Violin (Cambridge Massachusetts: When the transverse mode was Conclusion The MIT Press) translated by J. bowed, the reflected spot appeared as a The E string whistle is caused by Allen. smudge about 10mm across. But when I exciting the torsional modes of the plain Gillan, F. S. and Elliott, S. J. 1989. bowed the whistle mode, the reflected steel string. Experience has shown that "Measurement of the Torsional spotformed a clean straight vertical line, the whistle can be avoided by either Modes of Vibration of Strings on centered on the original stationary loca- using wound E strings that have high Instruments of the Violin Family", J. tion, indicating that the mirror was torsional damping, or by slowing the of Sound and Vibration, Vol. 130: rotating around its axis by up to around bow and attacking the string firmly, so 347-351. 90 degrees. The length of the line var- the torsional mode cannot establish itself Oberg, E. (Editor) and McCauley ied with the aptitude of the whistle note, in place of the desired transverse mode. (Editor). 1996. Machinerys When the bow left the string, the line The damping for wound E strings is gen- Handbook (25nd Edition) (New shrank slowly enough to be readily fol- erally much greater than that of plain E York: Industrial Press). lowed by the eye (rather like the picture strings, as shown by Pickering (1985). Pickering, N. C. 1985. "Physical on an old television shrinking to a dot It turned out my teacher's backup Properties of Violin Strings",/, ofthe when turned off). The slow decay violin already had a Dominant wound E Catgut Acoustical Society, No. 44. implies low damping, and provides a string, and that is why she did not have a Reprinted in Hutchins, C. M. direct way to measure the damping. problem with it. This is a heavily damped (editor): Research Papers in Violin In order to verify that the audio sam- string. I replaced theplain string on her Acoustics 1975-1993" Vol. 1, p 229- -ple corresponds to the observed torsional good violin with a similar wound string, 232, 1997. motion, I inserted a photo transistor into and the problem went away. ■ CASJ Rao, Singiresu S. 1986. Mechanical the path of the beam, and recorded the Vibrations (Reading Massachusetts: output of the transistor. An FFT analy- Addison-Weslev)

CASJ Vol. 3, No. 7 (Series II), May 1999 33 G. Bissinger. 1998. "AO and Al cou- R. Pitteroff, J. Woodhouse. 1998. C. Talmadge, A. Tubis, P. Piskorski, G. pling, arching, rib height, and f-hole "Mechanics ofthe contact areabetween a R. Long. 1997. "Modeling otoacoustic geometry dependence in the 2 degree-of- violin bow and a string. Part II: emission fine structure," Diversity in freedom network model of violin cavity Simulating the bowed string" Acustica Auditory Mechanics, edited by E. Lewis, G. modes," J. Acoust. Soc. Am. Vol. 104, Vol. 84, No. 4 (July/August): 744-757. .Long, R. Lyon, P Narins, C. Steele N0.6 (November): 3608-3615. (World Scientific, Singapore): 462-471. R. Pitteroff, J. Woodhouse. 1998. V Doutant, D. Matignon, A. Chaigne. "Mechanicsof the contact areabetween a C. Talmadge, A. Tubis, G. Long, P. 1998. "Numerical simulation of xylo- violin bow and a string. Part III: Piskorski. 1998. "Modeling otoacoustic phones. 11. Time-domain modeling of Parameter dependence," Acustica Vol. 84 emission and hearing threshold fine the resonator and of the radiated sound No. 5 (September/October):947-956. structures,"/. Acoust. Soc. Amer. Vol. 104, pressure, J. Acoust. Soc. Amer. Vol. 104, No. 3, Pt. 1 (September): 1517-1543. No. 3, Pt. 1 (September): 1633-1647. A. Runnemalm, N-E. Molin, E.V Jansson. 1998. "Operating deflection C. L. Talmadge, G. R. Long, A. Tubis, N. Fletcher, A. Tarnopolsky. 1999. shapes and the function of the violin," S. Dhar. 1999- "Experimental confirma- "Blowing pressure, power, and spectrum Speech, Music and Hearing Quarterly tionof the two-sourceinterference model in playing,"/. Acoust. Soc. Amer. Progress and Status Report. TMH-QPSR for the fine structure of distortion prod- Vol. 105, No. 2, Pt. 1, (February): 874- 3/1998: 5-18. uct otoacoustic emissions,"/. Acoust. Soc. -881. Amer. Vol. 105, No. 1 (January): 275- D. Russell, T. D. Rossing. 1998. -292. N. Fletcher, A. Tarnopolsky. 1999- "Testing the non-linearity of piano ham- -"Acoustics of the avian vocal tract," / mers using residual shock spectra," A. Tubis, C. L. Talmadge. 1988. "Ear Acoust. Soc. Amer. Vol. 105, No. 1 Acustica Vol. 84, No. 5 canal reflectance in the presence of spon- (January): 35-49- (September/October): 967-975. taneous otoacoustic emissions,"/ Acoust. Soc. Amer. Vol. 103 (January):4s4-461. L. Fuks, B. Hammarberg, J. R. Rustad, L. H. Morset. 1998. Sundberg. 1998. "A self-sustained "Investigation of the near field of a loud- L. Tronchin. 1998. Abstract - Doctoral vocal- ventricular phonation mode: speaker using tomographic reconstruc- thesis. "New theoretical and experimen- acoustical, aerodynamic and glotto- tion from TV-holography measure- tal methods for the evaluation of Psycho graphic evidences," Speech, Music and ments,"/ Acoust. Soc. Amer. Vol. 104, No. Acoustic Characteristics of Acoustic Hearing Quarterly Progress and Status 3, Pt. 1 (September): 1503-1508. Quality of Concert Halls," Acustica, Vol. Report, TMH- QPSR 3/1998: 49-60. 84, No. 5 (September/October). P. G. Stelmachowicz, D. E. Lewis, B. J. Gilbert, S. Ponthus, J. Petiot. 1998. Hoover, D. H. Keefe. 1999- E. G. Williams. 1998. "Supersonic "Artificial buzzing lips and brass instru- "Subjective effects of peak clipping and acoustic intensity on planar sources," / ments,"/ Acoust. Soc. Amer. Vol. 104,No. compression limiting in normal and hear- Acoust. Soc. Amer. Vol. 104, No. 5 3, Pt. 1 (September): 1627-1632. ing-impaired children and adults," / (November): 2845-2850. Acoust. Soc. Amer. Vol. 105, No. 1 C. Nederveen. 1998. "Influence of a (January):4l2-422. J. Woodhouse. 1998. "The acoustics of toroidal bend on wind instrument tun- AO-B0 Mode matching in a violin," ing,"/ Acoust. Soc. Amer. Vol. 104, No. 3, Acustica, Vol. .84 No. 5 Pt. 1 (September): 1616-1626. (September/October): 947-956.

34 CASJ Vol. 3, No. 7 (Series II), May 1999 Abstracts corresponding to the Table of Contents of ISMA9B can now be accessed at the CAS Website: www.marymt.edu/~cas

TABLE OF CONTENTS

I. Bowed Strings

GeorgeBissinger WHAT VIOLINMAKERS CANLEARN FROM MODAL ANALYSIS 5

Joseph Curtin INNOVATIONIN VIOLINMAKING 1 1

Paula M. Marston,Philip L. Marston SIMPLE DEMONSTRATION OF HELMHOLTZ MOTION OF ABOWED VIOLIN STRING 17

Lars Henrik Morset,AsbjernKrokstad, Ole JohanLokberg A COMPUTER-BASEDMETHOD FOR QUALITY CONTROL OF VIOLINS USING IMPULSE EXCITATION 23

Gerhard H. Muller,Helmut A Muller CHOICE OF STRINGS FOR INSTRUMENTS OF THE VIOLIN FAMILYUSING SIMPLIFIED THEORETICALASPECTS 29 Hyung GookMyung, Goeng-Mo Sung ANALYSIS OF DIFFERENCEBETWEEN UP-BOWED AND DOWN-BOWED VIOLIN TONE 35

TerenceM.Pamplin THE INFLUENCE OF THE BANDORAON THE ACOUSTICS AND ORIGIN OF THEBAROQUE BARYTONAND ITS CHARACTERISTICTONAL FEATURE OF SYMPATHETIC RESONANCE 41 Oliver E. Rodgers THE PRINCIPAL VIBRATINGMODES OF A VIOLIN WHICH PRODUCE SOUND .47

OliverE. Rodgers A TECHNICAL APPROACH IN 1998 TO VIOLIN PLATE AND CORPUS TUNING AND ADJUSTING 53 Kazuyuki Tomarikawa VIOLIN WOOD TREATMENT .59

Lamberto Tronchin, AlessandroCocchi ON THE TONING OF CELLO: THE EFFECT ON DAMPINGIN THE SOUND BOARD 65 DuaneVoskuil TOWARDS A THEORY OF VIOLIN EIGENMODE TUNING .:,...... 71 Shigeru Yoshikawa BOWING POSITIONS TO PLAY CELLO TONES ON THE VIOLIN 77

CASJ Vol. 3, No. 7 (Series II), May 1999 35 11. guitars and Plucked Strings

Boullosa, Felipe OrdufiaBustamante Ricardo R. GUITAR » ARTHURH. BENADE: ON SOUNDRADIATION INTHE CLASSICAL

91 ATNEW" WIDOW ON TONE, RE-SCALINGFREQUENCY RESPONSE FUNCTIONS

BUILDING 97 SSRHATION OF MUSICAL ACOUSTICSRESEARCH TO GUITARDESIGN AND E "tvT^^ phasein abaritone guitar E. Richardson Maria Pavlidou,Bernard luy0Q THE ACOUSTICS OF THE ANCIENT GREEK GUITAR

, Bernard E. Richardson ] 5 THE CLASSICAL GUITAR: TONEBYDESIGN

Samo §ali, JanezKopa£ TnT7„ OFDIFFERENT MACHINING PROCESSES ON THEACOUSTIC THE INFLUENCE Ul PROPERTIES OFWOODEN RESONANT BOARD

m. Brass Instruments

R. DeanAyers }lzy?« NEWPERSPECTIVES ON THE BRASS INSTRUMENTS

Matthias Bertsch ,^$ INTONATION ONTRUMPETS

Cullen, JoelGilbert, D. Murray Campbell, CliveA. Greated John S. MOUTH: ACOUSTICAL MEASUREMENTS IN RESONATORSDRIVEN BY ANARTIFICAL OSCILLATIONTHRESHOLD BEHAVIOR

D. Murray Campbell, Thomas MacGiffivray RECONSTRUCTION OF THE CARNYX

ChristopheVergez,Xavier Rodet . -, EXPERDvIENTS WITH ANARTIFICIALMOUTH FOR TRUMPET

H. A K. Wright,D. MurrayCampbell THE INFLUENCE OF THE MOUTHPIECE ON THE OF CUP-MOUTHPIECE WIND INSTRUMENTS

36 CASJ Vol. 3 No. 7 (Series II), May 1999 IV. Woodwind Instruments

Jean-Pierre Dalmont, JoelGilbert STANDARDAND INVERTEDHELMHOLTZ MOTION IN CONICAL WOODWINDS 167

F. N. Tanani MODEL OF THE VIBROACOUSTICBEHAVIOUR OFA 173

JeanKergomard, JoelGilbert OF REED-LIKE CYLINDRICAL INSTRUMENTS: TRANSITIONBETWEEN "COOPERATING"AND "NON-COOPERATING"MODES 179

Cornells J. Nederveen, Jean-PierreDalmont EXPERIMENTAL DETERMINATION OF THE INFLUENCES OF BORE PERTURBATIONS ONRESONANCEFREQUENCIES OF WOODWIND INSTRUMENTS 185 Edward Pillinger THE EFFECT OFDESIGN ON THE TONE &RESPONSE OF CLARINET MOUTHPIECES& PC BASED WAVEANALYSIS AS ADESIGNAID 191

GaryP. Scavone,Perry R. Cook REAL-TIME COMPUTERMODELING OF WOODWIND INSTRUMENTS 197

C. Segoufin, B.Fabre, M. P. Verge, E. M. S. Hanssen, A. P. J. Wijnands, A. Hirschberg RECORDER WINDWAYPROFILE: INFLUENCE ON SOUND PRODUCTION 203

D. B. Sharp, T. J. MacGillivray,W. Ring, J. M. Buick,D. Murray Campbell ACOUSTICAL COMPARISON OF CROOKS 209

William J. Strong REED TRIGGERING INCONES AND CYLINDERS 215

Taka'aki Tachibana,Ken'ichiro Man'syo,Ken'ichi Onodera,Kin'ya Takahashi, Tohru Idogawa SOUNDINGMECHANISM OF A SINGLE-REED INSTRUMENT WITH A CYLINDRICAL AIR-COLUMN 221

Rob van Acht THE SOUND QUALITY OFDUTCHWIND INSTRUMENTSFROM THEBAROQUE PERIOD: THERESULTS (2) 227

V. Pianos and Organs Pianos

Joaquim Agullo, XavierPages OPTIMUMPIANO TUNING. AMATRIX GENERALIZED INVERSEAPPROACH 235

AlexanderGalembo,Anders Askenfelt,Lola L. Cuddy INHARMONICITYAND QUALITY OF PIANO TONES 241 L. Rossi, G. Girolami BEATS INPIANO SOUNDS AND SHORT TIME FOURIERTRANSFORMS 247

CASJ Vol. 3, No. 7 (Series II), May 1999 37

Gautier, Olivier Thomas, David Rousseau, Rene Causse, Eric Marandas EFFECT OFHAMMER STRIKING IRREGULARITIES AND COMPARISON OF THE 253 MISTUNING ON THE DOUBLE DECAY OF PIANO TONES

Organs

P. Cottingham,Casey A. Fetzer James ZOi ACOUSTICS OF THEKHAEN

Andras Miklds,JuditAngster _ OF OPENLABIAL ORGAN PIPES; THE EFFECT OF THE SIZE OF THE SOUNDRADIATION 267 OPENINGS ON THE STRUCTURE

Yokota, VastfiSU,M. Scholz,M. Kleiner V Rioux. M. D. 273 ■PREUMDNARYSTUDY OFAN ORGAN BUILDER'S PERCEPTION OF AFLUEPIPE SOUND

Domenico StanziaL Davide Bonsi,Nicola Prodi FOR ANALYZING THERADIATIVE CHARACTERISTICS ANINTENSIMETRIC TECHNIQUE 279 OF SOUND SOURCES: A CASE APPLICATION TO AN

ShigeruYoshikawa IN PRE-SATURATED, POST-SATURATED, AND ORGAN-PIPE JETBEHAVIORS 283 MODE-TRANSITIONAL CONDITIONS

VI. The Analysisand Synthesis of Music by Computers

291 MUSICAL INSTOUMENT IDENTIFICATION USINGAUTOCORRELATION COEFFICIENTS

Chris Chafe OQ7 VICARIOUS SYNTHESIZERS: LISTENING FOR TIMBRE '

RESPONSES, ADA-TABASE OFMEA^RED MUSICAL INSTRUMENTBODY RADIATION IMPULSE APPLICATIONS FOR EXPLORING AND UTILIZINGTHE MEASURED AND COMPUTER 3 FILTER FUNCTIONS

Maarten van Walstijn, Julius O. Smith USE OF TRUNCATED INFINITE IMPULSERESPONSE OUR) FILTERS INIMPLEMENTING WAVEGUIDE MODELS OF FLARED HORNS AND PIECEWISE EFFICIENT DIGITAL 309 CONICAL BORES WITH UNSTABLE ONE-POLE FILTERELEMENTS

38 CASJ Vol. 3, No. 7 (Se ies II), May 1999

"

m VII. Perception and

Helen Pearson Fowler FROMTHE FIFTH SERIES TO SIGN, SERIAL OR "S"RATIOS 317

Ilia Kiufla, Andrzej Rakowski, Piotr Rogowski, Andrzej Miskiewicz PITCHAND PITCH STRENGTH OF CELLO VIBRATO TONES 321

Minora Matsuda,Kouichi Akiyama SPONTANEOUSRANDOMNESS ATTEMPO 327 Brian C. J. Moore PSYCHOACOUSTICASPECTS OF MUSIC PERCEPTIONIN NORMALAND IMPAIRED HEARING 331 Kristin Precoda, Teresa H. Meng LONG-TERM STABILITY OF LISTENING STRATEGIESDETERMINED BYMDS 337

Jan §tepanek, ZdenSk Otcenasek THE INFLUENCE OF SUBHARMONIC COMPONENTS OF STATIONARY VIOLIN TONES ONPERCEPTION 343

VIII. Percussion Instruments and Others Topics

James H. Irwin,Jr. ENERGY TRANSFER INTO DRUMHEADFROM STEEL & NYLON BALL IMPACTS 351 BarryLarkin, RonaldA. Roberts MEASUREMENT OF PERFORMANCERESPONSE DIFFERENCES USINGALUMINUM AND BRASS MARIMBARESONATORS 355 GaryP. Scavone, MaxV. Mathews THE MUSICAL ACOUSTICS RESEARCHLIBRARY 359

Domenico Stanzial,Davide Bonsi TOWARDS A LOCAL THEORY OF REVERBERATION: MONITORING THEBEHAVIOR OF ENERGETIC QUANTITIES DURINGTHE SOUND DECAYWITHIN ADUCT 365

U. Peter Svensson,MayumiNakano, Kimihiro Sakagami, Masayuki Morimoto EFFECTS OF WALL REFLECTIONS ONTHE SOUND RADIATIONFROMA KETTLEDRUM: ANUMERICALSTUDY 371

CyrilTouze, Antoine Chaigne, Thomas D. Rossing, Staffan Schedin ANALYSIS OF CYMBAL VIBRATIONSUSING NONLINEAR SIGNAL PROCESSING METHODS 377 Paul A. Wheeler TEACHING ENGINEERING 383 Zhang -nong A COMPUTERIZED STUDY OF CHINESEFOLK SONGS:AN ANALYSIS OF THE CHARACTERISTICSOF THE MUSIC OF THE FOLK SONGS IN THE CHU AREA 385

CASJ Vol. 3, No. 7 (Series II), May 1999 39 :. ||| £

A Joint Meeting integrating the 137th Meeting of the Acoustical Society of America, the 2nd convention of the European Acoustics Association, and the 25th German Acoustics DAGA conference was held March 15-19, 1999 at the Technical University of Berlin, Germany. Papers presented by CAS Members were listed in the Proceedings as follows:

Session IpAA: Architechural Acoustics: Session 2pMU: Musical Acoustics: Session 3pMU: Musical Acoustics: Concert and Opera Halls: Case Studies of Musical Acoustics and the Musician: T. Musical Instruments and Structural New Halls, Opera Houses: Leo Beranek, D. Rossing, Chair, A. Askenfelt, Cochair: Acoustics II: Piano and Related Cochair, Jurgen Meyer, Cochair: Y K. Guettler, "Musical acoustics, the Instruments: Isao Nakamura, Cochair, I Ando, Y Suzumura, I. Yamamoto, music student, and the music teacher." Bork. Cochair. "Acoustic design of the Tsuyama Music B. E. Richardson, "Acoustical training for Cultural Hall based on the preference stringed instrument makers." A. Friberg, Session 4aMU: Musical Acoustics: theory." L. Tronchin, "Acoustic quality J. Sundberg, "Illustrating music perform- Quality of Musical Instruments and in the Thretre "Palefenice," Venice. ance principles by synthesis.: R. D. Human Voice: Rene Causse, Cochair, Ayers, T. D. Rossing, "Teaching musical Johan Sundberg, Cochair: R. Causse, C. Session IpED: Education in Acoustics: acoustics." X. Boutillon, R. Grijalva, Begnis, N. Misdariis, "Assessment of Acoustics Education 2000: A. J. M. "Control index of the upright and grand musical instrument quality criteria: The Houtsma, "Acoustics education in the piano actions." notion of "openness"for a trumpet." E. Low Countries." T. D. Rossing, V Jansson, "Violin quality and bridge "Acoustics: A route to science literacy in Session 3aMU: Musical Acoustics: mobility." A. Askenfelt, K. Guettler, the 21st Century." Musical Instruments and Structural "Quality aspects of violin bow." M. Acoustics I: Experimental Studies, Castellengo, C. Besnainou, D. Dußois, Session IpMU: Musical Acoustics: Theoretical Models and Numerical "Acoustic quality of musical instruments Sound Production of Wind Instruments, Analysis: Antoine Chaigne, Cochair; and categorization." A. Galembo, A. William J. Strong, Cochair, D. Murray Uwe Hansen, Cochair: B. E. Richardson, Askenfelt, L. L. Cuddy, "On the effect of Campbell, Cochair: T. Idogawa, "Some "Experimental and theoretical studies of phaserelations for theperception ofpitch comments on the artificial blowing of the the modes of stringed-instruments and and timbre of bass tones." reed ." their relevance for quality control of instrument manufacture." A. Chaigne, I. Session 4pMU: Musical Acoustics: Session 2aAAa: Architechural Bork, "Comparison between modal Mapping Multiple Physical and Acoustics: Modeling of Halls; Halls analysis and finite-element modelling of Perceptual Attributes to Musical with Special Features: Jurgen Meyer, a marimba bar." M. Lewney, B. E. Structures: Johan Sundberg, "What Cochair, Leo Beranek, Cochair. Richardson, "Investigating the effect of makes singing expressive?" different strutting arrangements on the Session 2aMU: Musical Acoustics: modes of a guitar soundboard." E. V Session 5aMUa: Musical Acoustics: Modeling Versus Measurements of Wind Jansson and B. K. Niewczyk, "On the General Topics in Musical Acoustics I: L. Instruments: WJ. Strong, "Numerical function of the violin bridge." U. M. Wang, C. B. Burroughs, "A tale of calculations of woodwind instruments Hansen, T.D.Rossing, "Normal modes three violins: Comparison of radiation without adequate experimental data: of vibration in a violin." G. Bissinger, mechanisms measured with near-field Personal experiences. R L. Hoekje. "Flow "Deconstructing the violin - the road acoustic holography." and upstream impedence in wind instru- from B to A." PL. Hoekje, A. Morrison, ments." D. M. Campbell, "Why are his- "Finite-element analysis of vibrating Session 5aMUb: Musical Acoustics: torical brass instruments hard to play in bell." L. Morset, "An investi- General Topics in Musical Acoustics II tune?" J. Gilbert, D. M. Campbell, gation ofvibrational and acoustical prop- (Poster Session): K. Guettler, "Method "Mechanicalresponse ofartificial buzzing erties of the violin using MLS and optical for time-domain localization ofstochastic lips." J. Gilbert, J. P. Dalmont, J. holography." noise in quasiperiodic signals." Kergomard, "Reed instruments, from small to large periodic oscillations." L. Session 3pED: Education in Acoustics: Tronchin, "A virtual reconstruction of the Take Fives - Sharing Ideas for Teaching trumpet." Acoustics: U. J. Hansen, Cochair, A. Kohlrausch, C ochair.

40 CASJ Vol. 3, No. 7 (Series II), May 1999

: FUTURE MEETINGS

Violin Makers Association ofArizona, International (VMAAI) will hold their Annual Convention and Competition October 12 15, 1999 at the Executive Inn, Tucson, Arizona.

Acoustical Society ofAmerica: November 1-5, 1999, Columbus, Ohio December 4-8, 2000, Newport Beach, California June 4-8, 2001, Chicago, Illinois

(For information, contact ASA, 500 Sunnyside Bldv, Woodbury, NY 11797-2999. Tel: 516-576-2360 Fax:sl6-576-2377 E-mail: [email protected]

Australian Acoustical Society, November 26-28, 1999, Melbourne, Australia.

International Congress on Acoustics (ICA), September 2-7, 2001, Rome, Italy.

ISMA 2001, jointly organized by Catgut Acoustical Society and CIARM, September 10-13, 2001, Perugia, Italy.

WORKSHOPS, SEMINARS

55th Composers Conference: Chamber Music Center and Singers Workshop, Wellesley College, July 25 - August 8, 1999. Joan Einhorn, Executive Director, EO.Box 33505, Palm Beach Gardens, FL 33420. Tel/Fax: (561) 625-0365

London International Competition, April 10-16, 2000. Artistic Directors: Yehudi Menuhin and Yfrah Neaman, 62 High Street, FAREHAM, Hampshire, POl6 7BG, England. Tel: 10329 283603 Fax: 10329 281969 E-mail: [email protected] Website: Adresse:www.lsqf.com '

The Henry Mancini Institute, Jack Elliott, Music Director, UCLA, August 1-28, 1999. Visit their Web site: www.amjazzphil.org

TWENTY-FIVE YEAR MEMBERS

The Catgut Acoustical Society is again pleased and proud to add to the previous list of members who have been with us continually for 25 years.

Harold R. Allan Paul B. Ostergaard Norman J. Alstad Thomas D. Rossing Robert and Deena Spear

CASJ Vol. 3, No. 7 (Series II), May 1999 41 Here are some distinguished awards for CAS members from the Acoustical Society of America

Norman C. Pickering, Gold Certificate - (50 years) Thomas D. Rossing, Silver Certificate - (25 years) CONGRATULATIONS!

Elaine Moran: 1999 Distinguished Service Citation "for sustained and dedicated service to the Acoustical Society of America, its officers and members, over many years." CONGRATULATIONS!

Lily Wang: F. V Hunt Postdoctorial Research Fellowship, presented by the Acoustical Society of America - will undertake research program at Technical University of Denmark in Lyngby Subject of research is on the subjective investigations of spatial impressions. CONGRATULATIONS!

Grigori Sedukh: Treble Violinist with the St. Petersburg Hutchins Violiin Octet, violinist with the St. Petersburg Philharmonic Orchestra and concert violin soloist of international note, will receive the title "Knight of the Order of Malta" in recognition of his special services in music. CONGRATULATIONS!

Dimitry Markevitch's Cello Library - A catalog ofhis vast holdings which all cellists and music libraries should possess. "It is an invaluable source of information for those wishing to build programs, increase their knowledge of their instrument and its repertoire. Musicologists will find it a great help in their research work. A reference to keep handy, it contains 320 pages and over 1000 scores. Listed by categories, such as Concertos, Sonatas, Unaccompanied cello, etc., the works are clearly presented with all the necessary information pertaining to each of them, duration, dates, publishers et al. For all significant libraries, it should become a basis for future acquisitions." For more information: Ms. Gitta M. Buschhausen, POUR LA MUSIQUE, ,87 Rue duLac, CH - 1815 Clarens, Switzerland. Tel/Fax: +41 21 981 25 35. E-mail: [email protected]

The Royal Institute of Technology, Stockholm, held a public symposium at KTH on Sunday, September 6, 1998 entitled "Musical Noises" illuminating the importance of noise components in the sounds ofmusical instruments and their physical expla- nations . Also the use of noise in contemporary computer music composition and the processing of noise in digital recording and broadcasting systems were presented. The symposium was a satellite event to the Nordic Acoustic Meeting 1998 (NAM9B) and apart of the activities included in Stockholm - Cultural Capital ofEurope 1998. The programincluded presentations by Vincent Gibiat, "Chaos and Musical Instruments;" B. Fabre, "Noise in Wind Instrument Sound;" J. Sundberg and S. Ternstrom, "Noisein Voice;" Jim Woodhouse, "Noisein Bowed Instrument Sound;" Knut Guettler, Erik Jansson and Anders Askenfelt, "The String Player's Guide to Noise;" H. E. Jarvenhag, "Sound Reproduction and Broadcasting Noise;" Gerald Bennett, "Noise in Composition." The day closed with an electro-acoustic concert organized by Gerald Bennett and Peter Lunden, illustrating the use of noise, in composition during the 20th century.

IN MEMORIAM

Charles S. Johnson Daniel Martin Yehudi Lord Menuhin Leo R. Newfarmer

42 CASJ Vol. 3, No. 7 (Series II), May 1999 YEHUDI LORD MENUHIN 1916 - 1999

Yehudi Menuhin, worldrenowned violinist, was a member of the Catgut Acoustical Society since 1966. He had expressed an interest in our research and development in violin acoustics. When receiving his 2 5-year Certificate of Appreciation, he wrote: "I have always been interested in your publication, and although I must confess to not having read each issue thoroughly, I have always found something interesting in your publication. Many thanks for the Certificate of Appreciation. I am most grateful." In 1995 he wrote "he would be delighted to be a member of the CAS Advisory Council."

Menuhin founded the London Junior Music School in 1963 so that the modern violinist would not be just a player of Vieuxtemps or Tchaikovsky concertos, but a complete musician, at home in all types ofensemble and all styles, a thoroughly edu- cated and versatile musician as well as a skilled fiddler.

Roderick Skeaping, the Research Fellow in charge of the Octet instruments at London's Royal College of Music in the 19705, invited Menuhin to come and try them. Menuhin felt that young players should be made aware of their potential.

The world of string music has lost one of its most distinguished musicians and teachers.

CASJ Vol. 3, No 7 (Series II), May 1999 43 The Octet Development Group of the CAS is working hard to develop ways to present more concerts of the Violin Octet so they can be heard live in performance. Here are some comments:

"Over a lifetime I have heard a thousand concerts by some of the world's greatartists, and the June 5, 1998 concert in Colorado Springs Fine Arts Center, with MaestroSedukh on the treble violin with Madame Dzekster on piano, is one ofthe best three con- certs I ever heard. Sedukh possesses consummate musicality and technique, and in his hands the treble violin ranges from silky legatos and cantabiles to sparkling spiccattos. Dzekster is equally astounding, flawless and sensitive in her accompaniment, with immenserange of tonal colors and expression. The instruments of the Violin Octet could not hope for a better presentation to the world than that provided by these artists." John Randerson

The Denver Post (The Voice of the Rocky Mountain Empire) in their June 11, 1998 column had this to say: "It was an exotical- ly Russian occasion. The pianist, Inga Dzekster, looked like Natasha from "War and Peace" and played like a dream. And the violinist, Grigory Sedukh, chose the kind of heart-on-sleeve repertoire Russian fiddlers have played since the czarist days. The visitors from St. Petersburg made history Wednesday night with a Houston Fine Arts Center concert aimed at introducing Colorado to a new family of violins designed and hand-crafted by Carleen Hutchins, an 87-year-old luthier from Montclair, NJ. She has created eight instruments that extend the range and alter the dimensions and dynamics of the conventional violin,viola, cello and double bass. Sedukh demonstrated three of them, the mezzo violin, an inch larger and considerably duskier in tone than the regular violin, the soprano violin, a half-octave higher and much more brilliant in tone, and the tiny treble violin, an octave above the normal violin, with a sound as sweet and round as a nightingale. The latter instrument, with a high-tension carbon rocket wire E string, was obviously closest to Sedukh's heart. In music by Tchaikovsky, Berlioz, Kreisler, Paganini, Vivaldi, Scarlatti and even an encore of "Silent Night" he filled the hall with soaring lyricism. Like a coloratura soprano, his trills and had clarity and flair, while the small fingerboard enabled him to execute double-stops at intervals impossible on a nor- mal . Hutchins has given music a vibrant new string sonority, which players and composers should embrace." JeffBradley, Critic-at-Large

The Classical recordings column of the Newark Star-Ledger (October 3, 1998), states the following: "In today's evershifting musical landscape, musicians, composers and instrument designers all feel the need to keep changing the rules for repertory, for the music you compose, for the instruments you play. It is a trend and always positive, but rewarding when gotten right. This release gets it right. ( "The New Violin Family" - St. Petersburg-Hutchins Violin Octet (Catgut Acoustical Society) ). The "Hutchins" in this Russian group's name is Dr. Carleen M. Hutchins, the Montclair-based acoustical scientist who has designed and built a series of eight stringed instrumentsranging from the tiny soprano violin to the gigantic contrabass violin. This pha- lanx deploys itself here in repertory from Vivaldi to Tchaikovsky to Yuri Faulk. Best is Vivaldi's Concerto in D Major, Opus 10, No. 3, which gains both textural heft andrhythmic drive through the eight instruments' thicker coverage of the score." Peter Spencer

44 CASJ Vol. 3, No. 7 (Series II), May 1999 We have received many fine comments on the two CDs recorded. Here are a few:

"The CD is simply superb! The quality of the instruments, the artistry of the players (most especially the trebleplayer), thefideli- ty of the recording, the cleverness of the programming (love the Gordon Jacob and the Vivaldi) and, not least, the beautifully written program notes all add up to a brilliant testimony to your remarkable accomplishments. Brava, Brava, Brava!!!" David Walter

"The CD of the music for the Violin Octet is splendid. The Vivaldi that leads off the disc is truly virtuosic. The player has mas- tered the treble violin with extraordinary skill and the result is electrifying. I greatly enjoyed the Tchaikovsky suite. The Russian arranger did a sensitive orchestration for your violins. The suite is a real display piece and, I would expect, can be used by radio stations who want to give a sample of the octet without necessarily playing the whole disc. (We are all aware of the structures of time and taste now being placed on music programming on the radio.) I singled out the Vivaldi soloist, but should not neglect to commend the other players. They have truly mastered these instruments. Combined with their evident musicianship, their playing is most satisfyingin all respects. The technical quality of the recording is very good. There is enough room tone to give the instruments breathing space, yet the microphone pickup is close enough not to diffuse the character ofeach instrument. All in all, this disc should do a great deal to spread word-and sound-of the octet. It will serve as an eloquent introduction to those who are not familiar with theoctet, and a delightfor all ofus who have treasured knowing them over the years. Congratulations!" Frank Lewin

"The exceptional CD's you sent me made the eight instruments sound so full and clear - everything from "soto voce" to "ponti- cello". You put to rest the old cliche that "newinstruments don'trespond". The combination of fine players, excellent arrange- ments, and your wonderfully responsive instruments make a truly impressive CD! Congratulations! Grigori Sedukh's artistry in negotiating the stratosphere on the Treble Violin is amazing - and beautiful." David Schwartz

"I had the great delight oflistening to the CD over the holiday. It really is a superb recording! It seems a long time (early 70's?) since Bernard Robinson brought oneof the original octets to Cardiff. Many, many congratulations! Charles A. Taylor

"I think the disc is splendid and all my reactions to it are strongly positive. The sound quality is beautifully clear and the high professional standard of all the performances have produced a real top class winner!" Maurice Hancock

"Well done! Your Russian musicians have done what I once hoped we'd do here. The CD is splendid. I very much enjoyed lis- tening to it. The Vivaldi arrangement is brilliant!" Michael Mclntyre

"This is just a short note to say how much we have all enjoyed listening to the wonderful music so well chosen and played by the St. Petersburg musicians. The quality of the sound is superb and the musicians really do justice in bringing out the unique tim- bre of Octet instruments." leuan Owen

"Many thanks for the reprint of the excellent article on the Octet about the '30 year experiment'. ("A 30-year experiment in the acoustical and musical development of violin-family instruments" by Carleen M. Hutchins: J. Acoust. Soc. Amer. 92(2), Pt. 1, August 1992 pp 639-650). It contains a wealth ofinformation and is written very clearly - a major contribution to the history of the family, which will be studied for years to come and have an important influence on future bowed-string musical instru- ments." Leo Newfarmer

The CD's are available for $15.00each (plus shipping and handling - $2.00 each per CD for domestic orders and $4.00 each for international orders). Credit card orders (Master Card or Visa) may be placed directly with the Catgut Acoustical Society at 973- -744-4029 or by FAX at 973-744-9197 or by E-mail: [email protected] Checks or money orders may be sent directly to the CAS offices at 112 Essex Avenue, Montclair, NJ 07042.

CASJ Vol. 3, No. 7 (Series II), May 1999 45 " 'i

The New Violin Family The St. Petersburg Hutchins Violin Octet Performing on instruments designed and crafted by Dr. Carleen M. Hutchins Contains works by Vivaldi, Tchaikovsky, Falik, Belov, Rackley and Jacob.

Features selections scored for the entire octet as a separate ensemble, plus selections combining the octet with trumpet, , clarinet and cembalo.

Credit card orders (MC & VISA) may be placed directly with the Catgut Acoustical Society at 973-744-4029 or by FAX at 973-744-9197 or by EMAIL at [email protected] Checks or money orders may be sent directly to the Catgut Acoustical Society offices at 112 Essex Ave. Montclair, NJ 07042 (S&H= $2.00 per CD for domestic orders " $4.00 per CD for international orders)

46 CASJ Vol. 3, No. 7 (Series II), May 1999 CATGUT ACOUSTICAL SOCIETY, INC.

ADVERTISEMENT RATES

Page $350 10" x 7" Half Page 200 4 7/8" x 7" Quarter Page 125 4 7/8" x 3 1/2"

Rates are for black and white camera-ready copy. (No photos)

Deadlines for semi-annual Journal: Material received by February Ist will be published in May Journal. Material received by August Ist will be published in November Journal.

Please fill out and return the blank below

Camera-ready artwork is enclosed.

Special instructions:

Name:

Company name:

Address:

Telephone number: Fax number:

Signature: Date:

Please return to: Catgut Acoustical Society, Inc., 112 Essex Avenue, Montclair, NJ 07042 - FAX: 973-744-9197.

CASJ Vol. 3, No. 7 (Series II), May 1999 47 48 CASJ Vol. 3, No. 7 (Series II), May 1999 Editorial Procedures for Publications not suitable for the CAS Journal, the editor will inform you and Members are sometimesconcerned that papers they submit explain the reasons. The Journal gives priority to short notes of for the CAS Journal are not always published in the upcoming 1000 words or less. issue. It may help to understand the publication process and Please note that submitting a manuscript for editorial consid- schedule: eration is a representationthat thepaper has not been copyright- When your paper is received, the editor sends it to someone ed, published elsewhere, or under consideration for publication conversant with the topic and qualified to review the paper. This elsewhere. can take several months. If changes are needed, the editor sends Items submitted for consideration in the News & the paperback to you with suggestionsfor corrections. When you Correspondence and Meetings, Workshops, Seminars columns return thepaper and theeditor agrees that it is ready for publica- must reach us by February 15 for the May issue and by September tion, it enters the queue for space in the next issue. If thepaper is 5, for the November issue.

Guidelines for Submitting Papers can accept most common word processor formats, e.g. Microsoft Submit all papers to: Word, WordPerfect, ASCII text. Submitting manuscripts in CAS Journal machine-readable form facilitates rapid and accurate publication Catgut Acoustical Society and helps the Society hold down publication costs. We will ask 112 Essex Avenue you to submit accurate drawings of any figures, suitable for pub- Montclair, NJ 07042 lication, with the final manuscript. If the figures, charts, graphs, etc. have been generatedby computer, please submit them on the Provide three copies of a neat and legible paper. Double- floppy disk. We can work with most graphic formats, e.g. TIFF, space, type or print, and number the pages. JPEG, GIF, EPS. Review and follow the style ofpapers published in this issue The text of Journal manuscripts should normally not exceed of the CAS Journal. Note that if your paper is acceptedfor pub- 6000 words (approximately six Journal pages, including illustra- lication, the editor will provide a style sheet with precise instruc- tions). Limit notes to 1000 words. Discussion ofmaterial previ- tions for preparing the final manuscript to meet Journal require- ouslypublished in the Journal may be submitted as a Letter to the ments. References, in particular, must be prepared accurately and Editor and should not exceed 500 words. Closure to discussions formatted as prescribed. We strongly encourage authors to sub- prepared by authors and reviewers should not exceed 250 words mit a copy of the final manuscript on a 3V4 inch floppy disk. We per discussion.

OFFICERS TRUSTEES ADVISORY COUNCIL Violin Octet President Deana R. Campion Donald L. Engle Carleen M. Hutchins J. Maurits Hudig David L. Chrapkiewicz Dennis Flanagan Margaret H. Sachter Executive Vice President Christopher Chafe Frank Lewin Julius VandeKopple Sam R. Compton, Jr. Gabriel Weinreich International Treasurer Joseph Conrad Herman Medwin, Chair Duncan Kidd Joseph Curtin COMMITTEES Vice Presidents Permanent Secretary Uwe Hansen Research Australia: John GodschallJohnson Carleen M. Hutchins Duncan Kidd Oliver E. Rodgers Canada: Warren Reid Permanent Advisor A. Thomas King Journal France: VoichitaBucur Morton A. Hutchins Herman Medwin Gregg T. Associate Editor Germany: Volkmar Tetzner Executive Secretary JoanE. Miller Robert T. Associate Editor Hong Kong: Anton Sie ElizabethMcGilvray Paul Ostergaard Catgut MusicalAcoustics Italy: Domenico Stanzial Edith Munro Research Library Japan: Isao Nakamura John T. Randerson JoanE. Miller Netherlands: Adrian M. Houtsma Oliver E. Rodgers Russia: Marina Markot Margaret H. Sachter Scandinavia: Anders Askenfelt Robert T. Schumacher U.K.: James Woodhouse CASJ Vol. 3, No. 7 (Series II), May 1999

Alf, Schumacher,