Modeling the Violin, Computers the new Apprentices’

In Stradivarius’ day he and other violin makers used apprentices who often paid to work with the master for several years before being able to make their own instruments. These apprentices were initially given the most rudimentary work like sweeping and emptying the volumes of wood shavings created in a violin shop. Eventually they would progress to doing the labor intensive aspects of violin making, the purely mechanical task which must be done before the delicate craftsmanship of a master would be required.  Apprentices learned tool skills and honed their woodworking abilities, and also saved the masters innumerable hours of labor in the early task of creating a violin.

Today, there are few apprenticeships available. I actually completed one in ’85 as an electrician for UAW/GM. Much of my apprenticeship involved computer automation and robotics maintenance, a far step from making violins. But finally, that early work in computer automation, and many more years as a programmer etc in the IT industry bring me back to the work of the apprentice in a violin shop. As I’ve been making my violins, now on numbers 8 & 9, manually doing those mechanical woodworking steps; the algorithms and processes to automate some of the mundane work swirls around my head.

Make no mistake, a computer or CNC machine can not create a high quality violin. There are too many variables. Unlike something like an iPhone, a violin is made of wood. Each tree grows in a different climate, different moisture, on a hill, on a flat (hill sides affect the compression of wood grain). So, the material for two violins will vary in the grain density and this will affect the resonate frequencies of the completed instrument, which determine the tone of the instrument. An iPhone is made from steel, glass, and silicon; manufactured materials precisely controlled, and hence can be precisely manufactured. Violins are made from materials that mother nature creates, the violin maker must blend with nature’s results to craft a high quality violin.

In like manner computer automation can blend with the violin maker as an apprentice; doing the purely mechanical task, doing the repetitive manual tasks. With this objective in mind I’ve been doing some research and collecting of data sets to help violin makers use computer automation. Most violin makers don’t have 30 years experience in computers and automation, they have years and years of violin making experience; something I’m slowly working at. The last four years I’ve been learning how to make a violin using the traditional Cremona methods, as we understand them today. My mind has often drifted to the automation angle, but I’ve said no. No for two reasons. One, I’ve spent the last 3 decades working with computers, I want to create something old world, something more tangible than software. Two, without building the manual skills, the tool skills, the domain knowledge of how to create a hand made violin- using automation would be a waste of time. Working on violins 8 & 9 I know a little bit, mostly what the apprentices did, the mechanical aspects of the craft. It will take me many more years to garner the intuition to master the violin plate tuning, varnishing, etc etc required for a master piece.

In September I attended the Violin Society of America (VSA) convention in Indianapolis Indiana. My  head spun with information on all aspects of the stringed instrument world. I had the chance to see and play violins from the modern best makers; Sam Zygmuntowicz and Greg Alf. I held in my hands a Pietro Guarneri violin made in 1710. I was in nirvana. Most significant for this post, I met Dr Steve Sirr. Dr Sirr is a radiologist from the University of Minnesota (I had just been to the UofM on a computer issue, small world). For the past several years Dr Sirr has been making CT scans of some of the world’s greatest instruments. He currently has the scans of 40 instruments archived at the Smithsonian Institute. Dr. Sirr and I spoke and he made the raw CT DICOM data files available to me. From these CT scans I’m making the solid models required to prepare the machining instructions to enable a computer numerical machine (CNC) to be my apprentice. Part of my hope is to use the CNC as an apprentice to speed the mechanical work so I can focus more on understanding the more intuitive craftsmanship and tuning aspects of creating fine violins.

Dr. Steve Sirr and Sam Zygmuntowicz with the Strad 3D Project as well as many others have graciously shared their work in capturing these CT scans. In like manner I’ll be posting the model files on this site as I continue this fascinating project.



Stradivari Vn ‘Sunrise’ 1677:  (STL 3D file exported from CT DICOM)  (Blender solid model file created from STL)

Stradivari Vn ‘Betts’ 1704:  (STL 3D file exported from CT DICOM)  (Blender solid model file created from STL)

My journey continues…..