Sound analysis is a computer-aided acoustic measurement technique which is used as a diagnostic tool in the MARTIN SCHLESKE MASTER STUDIO FOR VIOLINMAKING. Sound analysis is used to show the sound radiated by the violin's eigenmodes of vibration in a spatial form – an individual "resonance profile". The resonance profile provides the acoustic data needed for subsequent analysis of the harmonic structure (levels of the fundamentals and harmonics) and for psychoacoustic evaluation (excitation of the inner ear). Based on the physical analysis, "musical information" is gained. This information is presented in the form of "colored maps". Comparing the results for different instruments helps provide very clear visual insight.
The spatial level differences in the radiated sound have their acoustic origin in the mode shapes of the resonances. Modal analysis is the only way to study the individual mode shapes. On the other hand, sound analysis is a tool which makes it possible to measure the acoustic efficiency of these resonances. When combined, modal analysis and sound analysis become a comprehensive acoustic diagnostic tool for violin research by allowing investigation of the acoustic origin as well as the acoustic efficiency of the sound radiated by an instrument.
Comparing Resonance Profiles
The resonance profile is determined by measuring transfer functions being obtained by the ratio of
sound pressure p divided by bridge force F as function of frequency f (x-axis) and radiation angle around the instrument. As a result the resonances of the instrument show up as peaks in the spectrum.
Zur Messung: Messung von 6 Übertragungsfunktionen (FRF) mit Abstrahlwinkel (Längengrade) j=0...300°; Dj=60°. Anregung des Instrumentes über Impuls an Stegoberkante-IV: 442x mit Impulshammerpendel. Hammer PCB 086C80. Aufnahme über SPM B&K 2237, FFT-Analysator: Quattro Data Physics; Software SignalCalc ACE.
Je FRF: Setting FRF_NP: Number of Samples = 16384; Bandwidth = 12.8 kHz; Time = 640 ms. FRF Measure: Schalldruck / Kraft. FFT-Window: Response: No; Reference: Impact 7%. Abstand Instrument-Mikrofon = 0,50m. AVG: 36 Mittelungen je FRF, wobei gesamte Vorrichtung (Instrument - Mikrofon) im Raum um jeweils a=10° gedreht wird (Herausmitteln der Raummoden); relative Lage Instrument - Mikrofon dabei für jede FRF unverändert.
L avg. (1/12) = energetisch gemittelter Gesamtpegel (65 Hz...10 kHz) sämtlicher in 1/12 Oktaven (chromatisch) gemittelter Pegel.
Averaging over a certain frequency band allows to realize the overall distribution of energy. The diagram below shows an averaging over a bandwidth of a musical tierce.
During a two-year research project (1999-2000) sponsored by the European Commission, the MARTIN SCHLESKE MASTER STUDIO FOR VIOLINMAKING had the opportunity to develop a method for use in analyzing the sound radiated by violins and presenting it in a manner that corresponds to our hearing. The aim was to obtain significant insights into the relationship between criteria for subjective rating of the sound of a violin and quantities that can be objectively measured from the radiated sound. We will consider some of these research results below.
In the practical world of violinmaking, psychoacoustic sound analysis has the following applications:
- Characterization of an instrument in terms of its strengths and weaknesses.
- Diagnostic tool for use in tonal adjustment (e.g. based on the voice and bridge)
- Comparison of two instruments (e.g. a new instrument with an existing reference instrument)
- Expanded understanding of differences in tonal color by making it possible to view excitation patterns of the inner ear
Overview: "Sound analysis" of violins, violas and celli. The coloured contour diagram is an example of an analysis technique developed by Martin Schleske:
During the first step, the instrument's spatial sound radiation is measured. The measurement process is based on impulse excitation. A computer is then used to compute the "transfer function".
Then, we can see the resonance profile for the violin (here, the "Schreiber" Stradivarius from the year 1712). Starting with this transfer function, characteristic resonance differences between different violins begin to be revealed.
This measurement begins to gain musical relevance once the harmonic spectrum is computed for all of the playable notes and presented in the form of the colored contour diagram shown here. Colored contour lines are used to represent the sound of the violin like the peaks of a mountain range: The "peaks and valleys" in the sound become visible along with strong and weak notes. This also allows us to obtain an objective visual translation of attributes such as "brighter sound", "darker sounding", "even", "uneven" and so on.
For more details on the topic of sound analysis of violins, see:
Martin Schleske: "Empirical Tools in Contemporary Violin Making: Part II: Psychoacoustic Analysis and Use of Acoustical Tools". CAS Journal Vol. 4, No.5 (Series II), Nov 2002
The MARTIN SCHLESKE MASTER STUDIO FOR VIOLINMAKING would like to acknowledge the EC funding (CRAFT programme in DG Research) for our project entitled "Material investigations and advanced methods of production and quality control for bowed instruments". Details can be found in the EC project database at www.cordis.lu under project reference number BRST985465.