As opposed to the exact sciences, new insight in violin making cannot always be expressed in terms of formal logic. Even if a great number of measurement methods are now at our disposal, still, the findings of one individual violin maker cannot always be applied across the board.
This type of knowledge is best transmitted within small groups of makers who are likewise friends and colleagues. Having worked together for several years and spent some time clarifying main concepts and goals, we can successfully embark on joint research projects and share our knowledge with one another. Although we only seldom find formal, generally binding answers to our questions, we are thrilled with any new discovery that reduces the amount of areas marked as “unexplored” on our violin making globe.
The viola is one such terra incognita. There is no consensus among musicians or luthiers regarding what sound a viola is supposed to have. For most parameters such as corpus length, string length, f-hole measurements, and bridge dimensions, we still don’t have universally accepted benchmarks when it comes to the viola.
The 2016-2019 Obialto project was one of the first major expeditions ever undertaken into unmapped viola territory:
In a group of viola makers at the VSA Oberlin Meeting in 2016,1 we started out by sharing our experience and insight into the connection between viola model and viola sound. Our goal was to draw up a new viola model as a point of departure for all Oberlin violin makers to make a series of concrete instruments and subsequently compare their characteristics.
After a period of 10 days diskussing and drawing more than 50 model variants, we finally agreed on one form: the Obialto. Sixty identical inner forms were then prepared for the Oberlin violin makers to take home with them.
The drawing of the Obialto model was propelled by a team of the violin makers: Alkis Rappas (Kingwood, TX), Eugene Holtier (Cleveland), Marilyn Wallin (Lincoln, NE), Peter Goodfellow (AUS), Raymond Schryer (CAN), Sam Zygmuntovicz (NY), William Scott (Minneapolis), and myself.
I would also want to express my special thanks to Prof. Harry Mairson (Boston), who provided us with his proportional drawing software and helped us in practice to draw lots of the variants.
In early 2017, on the basis of that inner form, each participating luthier planned and made his or her own viola, varnished it, and brought it to the next Oberlin meeting:
In the end 26 Obialto violas were brought to the 2017 Oberlin Meeting!1
All participating luthiers had been asked to measure and note their instrument’s 80 most important making parameters: thus, we were not only able to compare the violas by sight, but we had detailed knowledge of their inside measurements, their weight distribution, their eigenmodes.
Below: the large round table at Oberlin College on which all Obialto instruments were presented together!
To evaluate the viola – the former terra incognita – we implemented the following comparative tests:
• Perceived timbre test in a double-blind concert, with evaluation questionnaires for the listeners and the performer. All 26 violas were played in the Warner concert hall by a professional violist (thanks to Chris Sandvoss!)
• In a soundproofed room, we measured the sound radiation profiles of all the Obialto violas on an Impulse Measurement Rig by means of an impact hammer, recorded with twelve microphone positions. The curves, featured below, were averaged and analyzed with George Stoppani’s software. (Thanks to M. J. Kwan and Bill Scott for their patience during the 14-hour measurement session!)
• At Clonick Recording Studio at Oberlin College, a professional violist made sound recordings of all violas, playing the same piece every time with a steady microphone position and distance. These recordings were anonymized and placed as graphic access symbols on a computer desktop screen (thanks go to Yi-Ping Yang, Claudia Fritz, and Felix Krafft!).
Test persons from the Oberlin meeting, as well as musicians interested in the subject, were thus able to freely categorize the recordings and evaluate them according to their auditory impression.3 In 2018/19, a research team led by Claudia Fritz (Sorbonne, Paris) analyzed those categorizations in order to relate them to the instruments’ construction parameters. They presented their research results at the ISMA in 20194.
• Furthermore we had personal playing and listening sessions: Yi-Ping Yang played the violas for small groups and commented in detail on their playing features and their sound.
The Obialto project was thus the first systematic investigation of the signature resonances and frequency bands of the viola. For the violin we already knew a series of factors that play a decisive role in its sound: the wood’s sound velocity and density, the tap tones, the plate archings and thicknesses, among others. Those factors have now likewise been investigated for the viola.
Moreover, those luthiers who are also violists had the time to play the 26 Obialto violas and compare each instrument’s construction parameters with their impressions from playing and listening to it – that was an incomparable experience!
For me this was truly something like drawing up an initial map of unexplored territory.
What a chance it is that the Oberlin meeting has this tradition of working enthusiastically together toward the same goal – that is what empowered this groundbreaking research project!
1 VSA Summer Workshops in Oberlin/Ohio - www.vsaweb.org/Oberlinworkshops
2 Certain colleagues use the spelling "Obiealto"
3 In recent years, psycho-acoustical test such as the Free Sorting Task or Free Evaluation Task have been applied to ensure that evaluations and assessments als replicable. See fpr instance the Bilbao test or Joe Curtin's Free Sorting Task.
4 To consult the results of Claudia Fritz's acoustical research team in greater detail, follow this link: ("The ObieAlto Project: looking for relationships between construction parameters & sound quality" - abstract for ISMA 2019)