ACTA ACUSTICA UNITED WITH ACUSTICA Vol. 96 (2010) 678 – 689 DOI 10.3813/AAA.918322 From Clarinet Control to Timbre Perception Mathieu Barthet, Philippe Guillemain, Richard Kronland-Martinet, Sølvi Ystad CNRS Laboratoire de Mécanique et d’Acoustique, 31 chemin Joseph-Aiguier, 13402 Marseille Cedex 20, France. barthet@lma.cnrs-mrs.fr Summary This study investigates the relationships between the control gestures of the clarinet, the generated timbres and their perceptual representation. The understanding of such relationships can provide great interest in several re- search contexts: synthesis and control (e.g., to improve the quality of current synthesis models), music analysis and perception (e.g., to study music performance), and music information retrieval (e.g., to find relevant acous- tical descriptors for automatic instrument and/or performer identification). A physics-based model was used to generate synthetic clarinet tones by varying the main control parameters of the model (related to the blowing pressure and lip pressure on the reed). 16 participants had to rate the dissimilarities between pairs of different tones and describe the factors on which they based their judgments in a questionnaire. The collected data were subjected to various statistical analyses (multidimensional scaling and hierarchical clustering) in order to obtain a low-dimensional spatial configuration (timbre space) which best represents the dissimilarity ratings. The structure of the clarinet timbre space was interpreted both in terms of control parameters and acoustical descriptors. The analyses revealed a 3-dimensional timbre space, whose dimensions were well correlated to the Attack Time, the Spectral Centroid, and the Odd/Even Ratio. Comparisons of natural and synthetic clarinet tones showed that the Odd/Even Ratio appears to be a good predictor of the beating reed situation, specific to single-reed instruments. PACS no. 43.66.Jh, 43.75.Pq, 43.75.Yy, 43.75.Zz 1. Introduction Since its conception in the 17 th century, the clarinet, a single-reed instrument from the woodwind family, has aroused the interest of composers as a result of its unique timbre [1]. From the acoustical point of view, the clarinet can be described as the association of an exciter (the reed) and a resonator (the air column contained in the bore of the instrument). Under certain control conditions imposed by the performer, the coupling between the exciter and the resonator, can lead to the establishment of self-sustained oscillations which are responsible for the sounds gener- ated by the instrument. By varying their control gestures, performers can produce different timbres (see e.g., [2]). This study aims to increase understanding of the relation- ships between the control of the instrument, the generated sounds, and how they are perceived. The understanding of such relationships can provide great interest in several re- search contexts: synthesis and control (e.g., to improve the quality of current synthesis models), music analysis and perception (e.g., to study music performance), and music information retrieval (e.g., to find relevant acoustical de- scriptors for automatic instrument and/or performer iden- tification). For synthesis purposes, perceptual clues can be used to calibrate the parameters of a synthesis model ac- Received 30 July 2009, accepted 4 March 2010. cording to the behavior of an acoustic instrument, or to propose control devices based on a high-level perceptual description (e.g., “play a bright sound"). For music analy- sis purposes, the study of the links between the control of musical instruments and the perception of generated tones allow to determine the mechanical/acoustical parameters used by performers to alter the timbre of the tones, and find the relevant acoustical descriptors that account for such timbre variations. Such investigations are of importance when studying the role of timbre in musical interpretation (see e.g., [3, 4]) and can give new pedagogical insights. In the music information retrieval context, the results of such studies may be useful to achieve automatic instrument or performer identification based on timbre descriptors. By varying the two main control parameters (related to the blowing pressure and reed aperture) of a physics-based synthesis model of clarinet [5], isolated clarinet tones were generated. Dissimilarity judgments were then collected and subjected to various statistical analyses to uncover the psychomechanical and psychoacoustical factors best pre- dicting the perceptual representation of the various clarinet tones. Timbre is the attribute of the auditory sensation that al- lows tones of equal pitch, loudness and duration to be dis- tinguished [6]. From the cognitive point of view, timbre refers both to the identity of the sound sources (timbre is the attribute that allows two different musical instru- ments playing the same note, with identical sound lev- 678 © S. Hirzel Verlag · EAA