ODESSA IV
ODESSA IV
New Orchestra Recordings -Measurements - Systematic Analysis
by Malte Kob (PI) (Detmold University of Music), Martha de Francisco (McGill University),
Caroline Traube (Université de Montréal) and Kit Soden (Université de Montréal)
with Ying-Ying Zhang (McGill University) and Jithin Thilakan (Detmold University of Music)
Research Creation Project Report
1. Introduction
The perception of ensemble sound and the blending of musical instruments play a crucial role in music composition, orchestration, and room acoustics adaptation [1, 2, 3]. Understanding these aspects is essential for both artistic and scientific advancements in music performance and acoustics research. The ODESSA project has previously conducted three orchestral recording sessions to investigate ensemble blending, but due to the complexity of influencing variables, further dedicated recordings and acoustic measurements are necessary for a comprehensive scientific analysis [4, 5, 6].
This report presents the fourth ODESSA recording session, which involves a string ensemble performing selected musical excerpts under controlled acoustic conditions. The study aims to analyze the artistic and acoustic factors contributing to musical blending, incorporating room acoustic variations and ensemble performance attributes.
Figure 1. Performance of the violin ensemble, conducted by Pablo Andoni Olabarría
2. Materials and Methods
The investigation consists of three primary components: the composition of a dedicated musical piece for blending research, an ensemble performance and its recording, and the acoustic measurements and analysis.
2.1 Composition of Musical Work
Previous research indicates that different musical phrases and playing styles can significantly influence the perceived blending of instruments. To explore this phenomenon, a dedicated musical piece was composed specifically for the ODESSA IV recordings. A competition in composition was held among ACTOR student members to ensure the selected piece aligned with the study’s objectives. Several compositions were submitted, and following a thorough evaluation, the prize went to the piece that best allowed the display of instrumental blending. The selected composition, Discanto II, Theme and variations for violin ensemble, by Pablo Andoni Olabarría for an ensemble of nine violins, features carefully crafted variations in tempo, pitch range, and dynamics, providing a structured framework for analyzing ensemble blending across different musical contexts.
2.2 Ensemble Performance and its Recording
Acoustic Environment:
Figure 2 and 3. The MMR with passive acoustic system (with and without curtains) and active Meyer constellation system. While a Sennheiser KU100 and mh acoustics’ em32 Eigenmike are shown hanging from the ceiling, they are permanent room installations and were not used in the study.
The Multimedia Room (MMR) at McGill University’s Schulich School of Music was selected as the venue for the ensemble performance and recording. Designed as an acoustically treated space, the MMR offers controllable acoustic conditions, enabling precise adjustments to the recording environment. Its passive acoustic treatment includes deployable acoustic banners and curtains, while an active virtual acoustic loudspeaker Constellation system further enhances flexibility. This setup allows for systematic variation of acoustic properties, such as reverberation time and early reflections, to study their impact on ensemble blending. Further details about the Constellation system can be found at [https://www.cirmmt.org/en/facilities].
The Recording
The Odessa IV recording, featuring Discanto II, aimed at a capture of the performance of the nine violins from different distances and heights, in addition to providing contact microphones for each instrument. The immersive recording will allow researchers to monitor the sonic characteristics of the nine instruments that concurrently play together to study instrumental blending. For that purpose, various takes of musical phrases were recorded in which an increasing number of violins gradually joined in.
Figure 4. Discanto II, full score, initial page
Figure 5. Discanto II, full score, final page
Composer Pablo Andoni Olabarría, the winner of the ACTOR competition award for the timbre and instrumental blending program, conducted the recording of his work, Discanto II, on October 24-28, 2022.
Recording Setup:
Figure 6. A diagram showing the main, surround, height, and soundfield capture systems in the space of the MMR. Circles represent microphones with omnidirectional polar patterns, while the doubled circles represent the dual cardioid microphones used for the back heights.
Figure 7. Microphone input list
For this study, a string ensemble consisting of nine violins was recorded while performing the selected musical excerpts. These excerpts were carefully chosen by the Tonmeister and artist members of the ACTOR project to encompass a diverse range of blending impressions. To precisely capture the contribution of each instrument to the overall ensemble sound, a close-miking technique was employed for every violinist. DPA 4099VIO microphones were attached to the body of each violin, ensuring the isolation of individual instrument signals, yielding partially anechoic recordings that are essential for later analysis.
Alongside close-miking, conventional channel-based recording techniques were utilized to capture the overall ensemble sound. Schoeps MK4 microphones were placed above each violinist as a spot capture system, while a main image capture was accomplished using DPA 4006 main (LCR) and outrigger (far Left and far Right) microphones placed on a Grace microphone positioning bar in the configuration of a Decca Tree.
Further into the room, another pair of DPA 4006 microphones served as second main microphone pair with more distance, which served for reverberation or surround layer capture. For a height layer capture, two Schoeps MK2H were hung from the ceiling above the front Left and Right positions, and two Sennheiser MKH 800 Twins were placed beyond the surround layer to capture the back height position.
Additionally, binaural capture was conducted using a Sennheiser KU100 dummy head. Two Sennheiser Ambeo microphones were placed between the center and horizontal boom of the Grace system at two height levels: a high capture near coincident with the main LCR system, and a low capture below the conductor stand.
Beyond the musical aspects, the study also examined how variations in the acoustic environment influence the perceived blending of sound. Controlled modifications were introduced to the room acoustics and seating arrangements to systematically explore their effects. Specifically, four distinct acoustic conditions were implemented, each designed to alter key parameters such as reverberation time, early reflection patterns, and spatial diffusion.
In addition to the recording of the full work, experiments were realized by recording unison passages with increasing or decreasing number of violins to gain insight into the blending behaviour of the instruments. The multitrack recording allows an analysis of the blended sound of the ensemble as heard at characteristic points in the space in front or above the instruments where microphones are located, including in the closest range.
2.2.1 Vocal and Instrumental Study, Vibrato Study
Figure 8. Photo of the vibrato study, showing musicians, conductor, and soprano soloist. Two additional Zylia Higher Order Ambisonics microphones can be seen in the foreground of the shot.
A short study was conducted featuring a singer alongside the violin ensemble. Soprano Theodora Nestorova joined the violins performing in unison the soprano solo melody of WA Mozart’s Laudate Dominum K.339 bars 11-24.
The experiments were aimed at studying the blending characteristics of the ensemble sound with or without the vocal soloist. Takes were performed by the violins in unison with increasing number of instruments. The multitrack recording allows insight in the specific sound found at crucial points above and in front of the ensemble. Specific performance directives were given to the musicians for every take.
1. only violins (with or without vibrato, fluid sound, blended sound)
2. soprano and various combinations of violins, following specific performance directives
3. experiments were recorded in different acoustic settings of the MMR Constellation system. The electronically enhanced reverberation system was set to a variety of acoustic conditions, such as “percussion acoustics”, “cathedral acoustics” or “early reflections only”.
The capture of separate layers of sound was realized with the full array of 33 close and far studio microphones. The soprano soloist was captured using a Schoeps MK4 spot microphone. Additionally, two Zylia Higher Order Ambisonics microphones were placed in front of the total ensemble in AB positions spaced one meter apart. This study was part of an ongoing inquiry into dual-soundfield recording systems.
Theodora Nestorova’s evaluations of the recordings were centred on the study of vibrato and its impact on the blend of sounds. See reference [8] from 38:38. Other evaluations of strings and vocals blending are still ongoing.
2.3 Acoustic Measurement and Analysis
Along with the ensemble recordings, a series of room acoustic measurements were taken in the four environments to characterize their individual acoustic properties. The measurements included impulse response recordings (RIR) to capture the acoustic signature of the performance space, and speech transmission index (STIPA) to evaluate speech intelligibility within the room. The RIRs were measured at multiple source-receiver locations, and the room acoustic parameters representing aspects such as perceived reverberation and clarity were extracted in accordance with the ISO standards [7]. Additionally, STI values, representing speech intelligibility, were estimated in accordance with the IEC 60268-16 standard.
Various microphone configurations, including binaural and ambisonic recording setups, were employed to capture different spatial perspectives of the ensemble’s sound field. The binaural recordings, in particular, are expected to provide valuable insights into the perception of blending by simulating a realistic listening experience. The collected measurement data will serve as the basis for subsequent signal evaluation and perceptual listening tests, which will focus on assessing the influence of musical and acoustical factors on the overall blending perception.
Along with the acoustic measurement techniques mentioned above, an acoustic camera was used to record the singer's and string players’ performance, enabling a visual analysis of sound emission from the sources and the room reflections. This device allows for the investigation of the intensity and localization of the sound sources and room reflections for different frequency bands (see Figure 9).
Figure 9. The acoustic camera recording of the performance (on the left), and the screenshot of the screen capture of the camera recording of a performance (on the right)
3. Expected Analysis, Future Works, and Benefits
While the recordings and acoustic measurements have been successfully completed, a comprehensive analysis is yet to be undertaken. The next phase of this project will involve systematically analyzing the recorded signals both objectively and subjectively. This process will include perceptual testing, and signal processing techniques to identify spectral and temporal characteristics associated with blending perception, as well as computational modeling approaches to quantify the relationship between musical features and perceived blending.
A comparative assessment of the four ODESSA recording sessions will be conducted to identify trends and variations in blending perception across different ensemble settings and acoustic environments. The perceptual listening tests will be designed to evaluate the influence of musical and acoustic factors on ensemble sound, with expert listeners providing subjective ratings of blending impressions. The findings from these tests will be correlated with objective acoustic parameters to gain a deeper understanding of the underlying mechanisms of ensemble blending.
The data collected from this project offers several benefits to the broader research community. The availability of systematically recorded and analyzed musical phrases enables further studies on musical blending. The string orchestra recordings and acoustic measurements provide a well-documented dataset that can be used for comparative research and perceptual experiments. Additionally, detailed insights into instrument recording techniques also enhance best practices in music production and performance research.
Beyond the scientific and artistic aspects, this project also presents valuable opportunities for musicians and Tonmeister students. It provides them with hands-on experience in professional acoustic recording techniques used in research, music production, and signal analysis. Through direct engagement with the recording process, participants develop a deeper understanding of how different methods impact sound capture and subsequent analysis. Comparing results obtained from multiple recording techniques also provides critical insights into the scope and limitations of these methods. Furthermore, this research aims to improve the understanding of how blending occurs between instruments in an ensemble setting, offering implications for both artistic and scientific discussions on ensemble sound formation.
4. Conclusion
This investigation marks a step in understanding ensemble sound and musical blending by combining high-quality ensemble recordings with detailed acoustic measurements. The collected data provides a solid foundation for future analysis, contributing to the broader understanding of ensemble performance. In addition to its scientific relevance, the project serves as an educational resource for musicians and audio engineers, offering insights into recording techniques, acoustic analysis, and ensemble performance dynamics.
The findings from this study will have direct implications for music performance, orchestration, and room acoustic design. By bridging the gap between acoustics, recording arts, and musical practice, this research seeks to advance our knowledge of how instrumental blending is shaped by both performance and environmental factors. Through further analysis and perceptual evaluation, this project aims to offer valuable insights into the nature of ensemble sound, paving the way for innovative applications in music technology and acoustic research.
5. References
[1] Thilakan, J., Balamurali, B. T., Chen, J. M., & Kob, M. (2023). Classification of the perceptual impression of source-level blending between violins in a joint performance. Acta Acustica, 7, 62.
[2] Thilakan, J., BT, B., Colella Gomes, O., Chen, J. M., & Kob, M. (2025). Exploring the role of room acoustic environments in the perception of musical blending. The Journal of the Acoustical Society of America, 157(2), 738-754
[3] Thilakan, J. (2025). Acoustic and perceptual aspects of sound sources in the formation of ensemble sound [Doctoral dissertation, Detmold University of Music, Germany] (To be published).
[4] De Francisco, M., Kob, M., Rivest, J. F., & Traube, C. (2019, November). ODESSA–orchestral distribution effects in sound, space and acoustics: an interdisciplinary symphonic recording for the study of orchestral sound blending. In Proceedings of International Symposium on Musical Acoustics (ISMA) (pp. 13-17)
[5] Ioannou, S., & Kob, M. (2019, November). Investigation of the blending of sound in a string ensemble. In Proceedings of International Symposium on Musical Acoustics (ISMA) (pp. 13-17).
[6] Thilakan, J., & Kob, M. (2021). Evaluation of subjective impression of instrument blending in a string ensemble. Fortschritte der Akustik–DAGA, 15-18.
[7] ISO 3382-1. Acoustics - Measurement of room acoustic parameters - Part 1: Performance spaces, Geneva: International Organization for Standardization, 2009.
[8] Nestorova, T. (2023). Finding Consilience in The Vibrato Wars: Hearing, Seeing, & Analyzing the Spectrum of Variability Across Genres. Series Research Alive, Schulich School of Music, McGill University. Timbre and Orchestration Resource (TOR). https://youtu.be/3cxDe4OyIt0?si=DZ-QF2SgRndPSBXU