Sketching technics
All the layouts are established on the basis of reliable data, such as for example the original reproduction plates provided by the museums, scans, posters and dimensional statements compiled by experts. However, this data is never directly usable. Tracing a violin mould does not consist of stupidly tracing the outlines of a photo or reproducing a 3D scan, it is a complex operation which requires weeks of work and which sometimes resembles a real police investigation.
First of all, it often starts by a big effort of retouching the images by digital means to eliminate the very frequent optical distortions. It is true that digital photography and processing algorithms have enabled significant progress: recently published posters are often much more usable than those of twenty years ago. But despite this, there are still more or less marked inconsistencies between the images or scans and the dimensional readings. It is then necessary to do research, cross check data, compare and understand what is going wrong to finally decide on the corrections to be made, either to the images or to the final dimensions. In general, “flat” areas such as the outline of the body are quite easy to correct because optical distortions are quite predictable there, but for raised areas such as the volute, it can quickly turn into a nightmare.
At this stage, it is very useful to stick to the original sketching methods of the moulds and the different parts of the instrument in order to interpret the visible deviations. For this, I rely a lot on the work of François Denis, but not only. Sometimes I rediscover sketching methods that had never been described, especially for the area of Cs and corners.
Then, once the data is made reliable, you have to spend time interpreting it to understand what should be kept and what should be corrected.
The mechanical distortions of the instrument, in particular, must be eliminated. It is known that the top plate bulges in its center under the thrust of the sound post and in its upper and lower bouts under the compression exerted by the strings. But these phenomena are also reflected in 3D with more or less strong consequences on the outline of the body, on the neck angle, on the length of the instrument, etc. And this is not always predictable and can vary from one instrument to another depending on the wood, the thicknesses, the string length, the string gauge...
In any case, it is a bad idea to copy a distorted instrument because the distortions would continue to grow up once it has been strung up and played for a while, turning it into an infamous Quasimodo. It is therefore important that the mould and templates are based on the shapes of the instrument before deformation.
In the same way, the instrument may have undergone repairs or modifications that must be detected and understood in order to decide whether or not to correct them. On ancient instruments, the top plate may have been removed and glued back in place a considerable amount of times, which may end up reducing the ribs height significantly. The problem may have been corrected by raising them again...or not...or too much...It is important to find out what was the original ribs height because it directly impacts the inner air volume, and therefore its eigenfrequency. The geometry of the neck is also a very thorny issue: almost all the instruments prior to the 18th century had their neck replaced, to switch them from the baroque setup to the classical setup. This operation may have been carried out together with a modifications to the string length and the neck angle, in order to preserve (or not) the mechanical balance, depending on the new set of strings used. The distance from the nut to the edge of the body can then deviate from the standard 130 mm which is yet a major ergonomic criteria for the musician. You then have to understand the logic that was followed and decide whether to stick to it or not. On violas, these changes are even more difficult to interpret. In any case, it is necessary to go through a fine calculation of the mechanical balance of the instrument to establish the final size of the neck because minor variations in angle and length can radically change the behavior.
On the contrary, some irregularities or singularities of the instrument fully belong to the master’s work. They are here since the beginning and it may be desirable to keep them because they contribute entirely to the personality of the instrument, both aesthetically and functionally. A good example is the asymmetry of the corners or soundholes: their position and their curvature. But it is the whole outline of the rib garland that must be considered because certain masters, like Guarneri Del Gesu, seemed to use and abuse the technique which consists in deforming it before sketching the plates outline, in order to produce instruments more or less wide, long or even distorted along a diagonal. It would be naive to believe that these singularities were the result of chance or a poor mastery of construction techniques, it is likely that they were totally deliberate. However, only you can judge whether or not to reproduce them on the instrument you are building. And only you can also decide how to reproduce them if so.
The masters worked with geometrically perfect tools and created the asymmetries during construction, but you may prefer to use a mould created specifically to precisely replicate those asymmetries. This is the reason why I always strive to offer two different types of layouts when building moulds and templates:
«Geometric» outline
This is the preferred choice for creating geometrically “perfect” instruments or when you want to apply your own techniques to create the asymmetries during construction. It corresponds to a perfect reproduction, without wear or deformation or approximation of the original mould used by the master. No doubt Stradivari and his colleagues would have used CAD drawing and CNC machining to build their tools if they could, judging by the care and high precision they put into them.
The axes are perfectly centered and perpendicular and the layout is a succession of circular arcs which follow the proportions and sketching rules known from that period or rediscovered for the occasion.
Of course, the Geometric tools are easier to use throughout the construction because everything is perfectly aligned and symmetrical, the process in straightforward.
«Asymmetric» outline
It is the relevant choice when you want to thouroughly reproduce a famous instrument including all the asymmetries and singularities that form its own character. Here, it is no longer the luthier who creates these singularities, but the mould and the templates. You just have to accurately follow the rules of traditional constructions and the final result will be amazingly compliant to the model’s shape.
The outline is made of very smooth spline curves based on the shape of the model, after removing all the unwanted mechanical distortions.
You should be aware that using such tools whose axes are not centered with respect to the outline can be confusing. It is preferable to have a good knowledge of lutherie techniques in order not to lose track of the global instrument geometry, for example when sketching the plates outline, positioning the soundholes or aligning the neck.
F-holes template
Drawing an F-holes template is a fairly complex operation if you want to do it right. I have seen many people just tracing over the f-holes from a picture and directly using it as a template. Obviously, the top plate arching makes this method very inaccurate. Yet, the shape of the soundholes and their positioning has a significant impact on the sound and on the mechanical properties of this central island which is the most mobile area of the violin. It is therefore crucial to devote the utmost care to it.
Here again, CAD tools are of great help. My method consists in tracing a 2D sketch of the soundholes and plate edges, from a carefully straightened and corrected photo. Then I build the 3D shape of the plate arching and projection my 2D sketch on it. Finally I unfold the plot obtained on the plate arching by flatening it. I can then rework this flat sketch to reconstitute the exact holes diameters and connect them to the curves.
The final template is laser cut in a flexible and transparent film. It is very easy to use and very precise since you just have to press it over the top plate arching while positioning it using the edges.
Of course, I make "Geometric" or "Asymmetric" F-holes templates depending on the type of kit in which they will be used.
Scroll
While the shape of the scroll has no direct impact on the sound or the ergonomics of the instrument, it contributes greatly to its aesthetics and its personality. It would therefore be a pitty to rush it.
More than for any other part of the instrument, the notion of «Geometric» or «Asymmetric» layout takes on its full meaning here because most of the scrolls mades by the ancient masters were very asymmetrical.
Still following the same logic, a «Geometric» head template is therefore made in a theoretically perfect way and the luthier will have to shift it when he sketches each side of the neck in order to create the asymmetry. Conversely, a pair of «Asymmetric» head templates is made so that the offset is already created between the left and right templates and you just have to position them in a strictly identical manner on each side.
Designing scrolls is a science in its own right. One could devote a lifetime to it! Fortunately, there are quite a few publications on the subject. I like to use the sketching method developed by Kevin Kelly as a base, but with (sometimes very strong) variations of the basic pattern in order to accurately reproduce the particular movements intended by the master. Each case is different and it often takes brain twisting to reconstitute the original sketching method by adjusting the basic pattern and the position of arc centers.
But the game is worth the effort because this method has the immense advantage of creating a perfectly regular and progressive spiral, which is almost impossible to achieve by tinkering with spline curves.
