Augmented reality

Luthier Paul Noulet and design engineer Benoit Dupeux combine their specialisms to find out how 21st-century technology can add to the luthier’s arsenal of violin making tools

CNC machining of a Stradivari mould

. ALL OTHER IMAGES COURTESY BENOIT DUPEUX

It is fascinating to note the amount of time taken by luthiers on preparing their tools. We can spend longer making templates and jigs than on the actual carving of the wood that makes up the instrument. It might not be the most rewarding part of the job, but the quality of the finished product can often depend on it. For this reason, we believe that utilising modern technology in the violin making process can relieve luthiers of these tasks while bringing higher precision to the end result.

This was our starting point when we decided to demonstrate how traditional and modern methods could be combined in a clever and constructive way, without ignoring the technical requirements of conventional lutherie. At first we simply intended to manufacture an accurate set of custom templates and moulds using modern tools such as computer-assisted design (CAD), computer numerical control (CNC) machining, and laser cutting. But as work progressed, we both realised that we could take our experiments much further – in fact, an immense new field of possibilities was opening up to us.

FOR A BOT TOM-UP COPY, YOU NEED TO RECREATE THE MOULD AND TEMPLATES THAT THE MASTER MOST LIKELY USED

When you decide to make a copy, there are two basic strategies available to you. First, there is the ‘top-down’ approach, where you try to reproduce the instrument as you see it with the greatest possible accuracy. Here the focus is on the result, although the luthier may use different processes or tools from the original ones to achieve it. Secondly, the ‘bottom-up’ approach, where you attempt to stick to the original intention of the master whose instrument is being copied. Here, what matters is the path and the logic of the method you follow, to reach a similar goal even if the end result has visible differences from the model (in particular if you take into account the differences in wood properties).

Both approaches have their pros and cons, and they can both benefit from modern digital tools. On the next page we list a few examples.

FIGURE 1 Computer-assisted design of a 3D mould

FIGURE 2 Afour-corner gauge, showing its perfect match with the back template

FIGURE 3 Flexible f-hole template made by laser engraving and cutting

For the top-down approach, you may want to reproduce all the instrument’s distortions and imperfections, as are visible on most old violins. For this purpose, we have experimented with making a ‘3D mould’ (figure 1), which reproduces the inner shape of the rib garland with extreme precision. Using a multi-part mould is nothing new, but using CAD and CNC to reproduce the shape of a known violin easily is a small revolution. It would probably have taken us weeks to do it by hand, and the result would not have been as precise. Having made the mould, it is easy to use: you lay the ribs around the edge as normal, then you trim their height following the edges of the counterforms, and then you remove them to fit the linings.

Even simple flat templates, such as corner gauges, take a new dimension when designed in such a way (figure 2). We had the opportunity of testing these on a copy of the 1732 ‘Willemotte’ Stradivari, working to the outlines on The Strad ’s poster (see December 2020). It takes a little digital magic to un-distort and scale the pictures accurately on computer, but we can then sketch the templates by following the picture outline, and send the program to the CNC machine to cut them out.

Another impressive example is the flexible template that we made for copying the f-holes of the 1702 ‘Lukens, Edler, Voicu’ Stradivari (figure 3). This template is made of a 0.4mm transparent plastic film, which is too thin to be milled. So we used laser cutting and got an amazing result. The template was computer-designed; we made a 3D projection of the f-holes on the arched surface of the top plate, in such a way that the final template would reproduce exactly the original f-hole outline when wrapped around it.

For a bottom-up copy, you need to recreate as accurately as possible the mould and templates that the master most likely used in their time. Then it is up to the luthier to reproduce (or not) their imperfections in a similar way to the originals. Here again, modern tools bring a lot of advantages.

A good example is the reconstruction we made of Stradivari’s ‘P(B)’ mould (MS6). This form is notoriously worn in its lower bout area; it even looks as though it has been deliberately re-cut to reduce its width, then roughly rounded again (figure 4). CAD allowed us to recover the original outline. First we used François Denis’s findings to investigate the sketching parameters of this mould. It was soon clear that it was geometrically impossible to draw circular arcs around the 196mm-wide worn lower bout. So we had to increase that width incrementally until we reached a realistic 198mm, at which point the arcs properly connected to the rest of the outline. But then we still had to distort this sketch in the same way as the original ‘P(B)’ mould. If you closely look at the lines drawn on Stradivari’s moulds, you can see how the subtle deviations impact on the positions of the corners and the slight asymmetry of the bouts. We decided it was worth reproducing these imperfections because they make the mould look more ‘alive’.

FIGURE 4 Stradivari’s ‘P(B)’ mould, showing how much wider the lower bout used to be

Having now experimented on a few copies, we believe that this is just the start. Traditional violin making is already a very comprehensive and highly technical field, but luthiers obviously cannot master all the ancient and all the modern techniques at the same time. It would be too much to cope with. So collaborating with engineers is a highly valuable opportunity for luthiers, which will ultimately help us take the next great leap forward.