Buenos Aires – 5 to 9 September, 2016 Acoustics for the 21st Century… PROCEEDINGS of the 22nd International Congress on Acoustics Music Perception: Paper ICA2016-227 Subjective preference of classical guitar strokes “apoyando” and “tirando” related to its harmonic components and autocorrelation function Joaquin Garcia(a), Shin-ichi Sato(b), Florent Masson(c) (a)Universidad Nacional de Tres de Febrero, Argentina, [email protected] (b)Universidad Nacional de Tres de Febrero, Argentina, [email protected] (c)Universidad Nacional de Tres de Febrero, Argentina, [email protected] Abstract Tone production of classical guitar performance is an essential part for musicians to transmit their sentimental and interpretative intentions. This work investigates subjective preferences of two common plucking techniques used by guitar players, apoyando (rest) and tirando (free) strokes. Six excerpts of classical guitar music with different tempos and range of frequency were performed using the two techniques and were recorded for the subjective tests. Two groups of subjects, guitar players and people who do not play guitar, were investigated to see if both groups evaluate the guitar timbre in different way or not. AB test was conducted with 50 persons for each group asking which technique is preferred and have more sound quality. Then the harmonic components and autocorrelation function (ACF) of each stroke were analysed to relate with the characteristics of the music program (tempo and frequency range) and subjective preferences. The effective duration of ACF is defined with the taue (τe) parameter. Results of the subjective test showed that harmonic content did not define preferences, but higher taue values of the ACF were correlated with a higher sound guitar quality. Keywords: Plucking technique, Apoyando, Tirando. 22nd International Congress on Acoustics, ICA 2016 Buenos Aires – 5 to 9 September, 2016 Acoustics for the 21st Century… Subjective preference of classical guitar strokes “apoyando” and “tirando” related to its harmonic components and autocorrelation function 1. Introduction Classical guitar has become one of the most popular instruments of the twentieth century since Andres Segovia introduced it in the stages in 1928. The modern six string guitar is a descendant of the sixteenth century Spanish vihuela [1]. The Spanish luthier Antonio de Torres (1817–1892) contributed much to the development of the modern classical guitar when he enlarged the body and introduced a fan-shaped pattern of braces to the top plate. Francisco Tarrega (1852–1909), one of the most important guitar players and composers in the 19th century, added a new fingering technique (apoyando) and increased the variety of the guitar expression, feature so important to the composers in the romantic period. Apoyando stroke (rest) is a technique used by the right hand of guitar performers. The fingertip or nail pushes through the string and, after releasing it, comes to rest on the next one. Another commonly used technique is the tirando stroke (free). The nail rise clear of the next string and yet releasing the string from a point below its original plane. Figure 1 shows the two types of strokes and its motions. The essential difference is in the angle at which the string is released. Apoyando tends to set the string vibrating at a steeper angle, with a stronger component perpendicular to the soundboard, and therefore gives a richer sound than tirando. The latter is generally more suitable than apoyando when playing near the bridge and for harmonics, at least on the treble strings [2]. Source: (Taylor, 1985) Figure 1: (a) Tirando and (b) apoyando stroke [2] When the string is released at a perpendicular angle from the top plate the sound has strong amplitude and a high value of decay slope. This results in a hefty sound but not very prolonged in time. The case of horizontal release provides not very high amplitude but a more sustained note. Finally the two plucking techniques are a combination of both components with different proportions. Several researches have been done around tone production on classical guitar. Meyer studied the quality of fifteen different guitars by using a subjective test applied to 40 persons [3]. Using six excerpts of music he evaluated the subjective preferences and found a relationship with the frequency response of the guitar. The third resonance and the third octaves from 80 to 125 Hz, 250 to 400 Hz, and 315 to 500 Hz had the greatest importance on guitar quality, followed by the Q-value of the first resonance and the peak level of the second resonance. Dayan and Behar 2 22nd International Congress on Acoustics, ICA 2016 Buenos Aires – 5 to 9 September, 2016 Acoustics for the 21st Century… studied the quality of strings for classical guitars [4]. They reported a correlation between acoustical measurements (mainly sound levels and frequencies of the fundamental and overtones) with the subjective quality of “goodness” of guitar strings. However, there has been few discussion about preference between plucking techniques. Therefore this work aims to combine the objective analysis of the two common strokes apoyando and tirando, with objective parameters as harmonic content and temporal information extracted from the autocorrelation function of each technique. Six excerpts of music performed with the two strokes, were used in a subjective test. Preferences of each technique were evaluated in two groups of subjects, guitar players and people who do not play guitar, in order to see if both groups evaluate the guitar timbre in different way or not. An AB test was conducted to evaluate subjective preference of each plucking technique. Then, spectrum and psychoacoustical parameters as sharpness Z, Tonal dissonance (S), Roughness and ACF parameters were calculated. The effective duration (τe) of the running autocorrelation function (r- ACF) was also calculated [5]. Sharpness is a measure of the high frequency content of a sound, the greater the proportion of high frequencies the “sharper” the sound. Zwicker and Fastl define a sound of sharpness 1 (acum) as “a narrow band noise one critical band wide at a centre frequency of 1kHz having a level of 60dB” [6]. Roughness is defined by the perception of amplitude modulated tones in a broadband. Rapid modulations (between 20 and 250 Hz) result in sounds perceived as rough [7]. For slower modulations, (of about 4 Hz) listeners perceive fluctuations of loudness. The sensation of fluctuating loudness is called fluctuation strength. Tonal Dissonance (S) implies a combination of notes that sound harsh or unpleasant to people when they are played at the same time. Dissonance measures the harshness or roughness of the acoustic spectrum and many musical pieces involve a balanced combination of consonance and dissonance sounds as a resource, because the releases of harmonic tension create a pleasure sensation [8]. The ACF describes temporal information of signals. It indicates how similar a signal with its own shifted version is. It is recognized as a method for estimating the fundamental frequency of a sound signal, as determined by the delay time of the first major peak τ1. The effective duration of the envelope of the ACF, τe, which is defined by the 10 percentile delay and which represents a repetitive feature contained in the sound source itself [9]. τe is a parameter strongly related with music scores and musical expression. Ando et al. introduced a running ACF (r-ACF) analysis for music signals as a more accuracy method than the long-time ACF calculation [10]. He found that the preferred delay time of a single reflection considering a music motif can be calculated from the effective duration of the long-time ACF and the amplitude of reflection [11]. When music signals contains fluctuation in tempo the preferred delay time is represented more accurately by minimum value of the (τe)min of the r-ACF. Kato et al. studied singing voices related to room acoustics in terms of the (τe)min of the r-ACF [12]. They showed that “vibrato extent” and “singing volume” play an important role in decreasing the (τe)min value. 2. Procedure 2.1. Sound signals 3 22nd International Congress on Acoustics, ICA 2016 Buenos Aires – 5 to 9 September, 2016 Acoustics for the 21st Century… Six excerpts of guitar music were recorded for the subjective test considering different tempos and frequency ranges. The purpose is to cover a varied repertoire to evaluate the two plucking techniques. Table 1 shows the features of the six music excerpts. Music scores are shown in Figure 2. The recording was done in a room with a volume of 18 m3 with carpet floor. Porous absorptive materials were present on the walls and on the roof. An Earthworks M50 microphone was used with a Focusrite Scarllett audio interface. Also a professional hand-made classical guitar, Spanish model with red cedar top plate and rosewood back plate and ribs, was used for the recordings. Table 1: Description, time and tempo of the six music excerpts used in the test Description Duration Tempo Music [s] [bpm] Sarabande with a slow tempo. The melody line in is the upper 36 40 1 part accompanied in chords. Rapid series of semiquavers with fast tempo and melody line in 35 80 2 the down part. The center of tone is at A3 (220Hz). Excerpt of Adelita (Tarrega). Wide tonal range and melody more 23 70 3 strongly pronounced. Monophonic piece in the low and middle register with 2 octaves 27 55 4 of tonal range and andante tempo. Monophonic piece with fast tempo. Demand a clear separation of 18 150 5 the notes and use similar figures on various strings. 6 Calm quaver figures to be played exclusively on the highest 10 120 string. Figure 2: Music scores The microphone was located pointing the bridge of the guitar from the side and 20 cm away in order to capture its complete spectrum.