AHEAD OF THE CURVE

Recording the archings of instruments is one of the most difficult areas of violin making and restoration. Charline Dequincey describes a method using dental compound which is accessible to anyone, and gives high-quality results

ALL PHOTOS CHARLINE DEQUINCEY

Computerised Tomography (CT) scans are very useful for violin makers, both in the process of making new instruments and in helping to understand the work and intentions of the classic makers. The technology was developed mainly for medical use but has also been used widely in many other fields. However, the equipment for this valuable process is expensive, and difficult to access for many individuals.

One of the most useful functions of CT scanning is to make cross-sectional ‘slices’ of a violin. These allow makers to see the arching curves in various places, as the profile changes constantly across the front and back plates. Makers often use these profiles to compare their progress while carving and refining the arching during construction.

This article presents an ‘analogue’ method of recording arching profiles, using moulding putty mounted on a thin backing. This method is readily accessible and the materials and technology affordable, unlike CT scans or accurate 3D-scanning technologies. Used appropriately, it is safe for the instrument and can provide more precise curves than a plastic profile gauge can record.

The material I use for this process is polyvinyl siloxane (PVS), also known as vinyl polysiloxane (VPS). This is a high-viscosity moulding putty which, like epoxy, is mixed from two parts: a base and a catalyst. It was originally developed to take dental impressions and is therefore a relatively non-toxic material to work with. Another advantage is that it does not warm when curing, which removes concerns about the reaction of varnish under the separating layer.

With a texture similar to modelling clay, PVS is easy to control and sets quite fast. Following 45 seconds of mixing and 2 minutes of working time, it requires a minimum rest period of 4 minutes before removal. Temperature affects the setting time: cold can extend it, warmth can accelerate it. Hence, storing PVS in the fridge will extend its working time by about 50 to 75 per cent.

PVS has a limited shelf life: as it ages, the setting time will increase. I have recorded a sample that was more than a year old taking up to two hours to cure. As I am not sure of the performance of aged PVS in terms of dimensional stability, I always use a product that is as fresh as possible.

Despite its viscosity, PVS can capture very fine details. To avoid trapping air bubbles, however, it must be applied very carefully, pressed firmly on to the selected surface, and completely worked into the corners. Although the product is designed to release easily from the surfaces on which it is setting, I strongly advise using a separating layer (i.e. tin foil), as it is unclear what might happen during the chemical reaction between the catalyst and the base, and how this could interact with instrument varnish. It is also important to note that latex and some modelling clay (perhaps containing sulphur or glycerin) can inhibit the setting. If in doubt, it might be advisable to make a test beforehand.

FIGURE 1A Balls of PVS are weighed out ready for mixing

FIGURE 1B About the same amount of each part is combined

FIGURE 1C The two-part PVS compound is applied

Once cured, the solid PVS has a consistency similar to rubber. Its resilience and tear strength make it possible to mould the arch around the edge and overhang on a violin top, and also to remove it easily. PVS is designed to remain dimensionally stable for seven days: the time frame in dental work during which a positive must be cast. For long-term archival storage of arching templates, one can make a positive cast or simply trace the arching outlines on to good paper.

The first step is to make a backing frame that will keep the PVS in place while it’s curing. For the frame I use a straight, flat piece of MDF, 3mm thick. This supports the soft putty and remains attached to the PVS after curing.

However, it is susceptible to humidity and can distort over time; epoxy can be used to seal and stabilise it. The frames can be reused, after you have recorded a positive or a tracing of the casting.

To control the positioning of the frame during the impression, I clamp frame-holding guides on to each side of the profile. The slot in the middle of these guides matches the thickness of the frame.

Before recording the negative profile, I choose one side of the frame to follow the abstract line of the section to be recorded. I mark the position of this side on a paper grid placed on the clamping surface for future reference. I also recommend noting which are the bass and treble sides.

As the mixing and working times are short, it is important to get everything ready before mixing the base and the catalyst together. For a typical arch impression, I prepare and weigh several balls of material (figure 1). This allows me to keep an accurate ratio of 1:1 and retain some flexibility with the final amounts by adding more while mixing. Then I roll each part, the catalyst and the base, separately, into ‘sausage shapes’ about the length of the surface to cover, and mix the two together. On a violin the total mix for a lower-bout width weighs about 35 grams and for the middle-bout width about 15 grams.

After mixing the putty I apply one sausage, made of about half the mix, on to the surface of the arch, and work the rest into the hollowed parts of the frame. Then I press the frame putty gently into the ‘surface sausage’ and shape the two together. At this point it should look like a satisfyingly attached mass. I leave this to cure while I work on the next one.

It can be difficult, when working in warm conditions, to cover the surface before the putty sets, especially underneath the corners. PVS sticks to itself when curing, so freshly mixed parts can be added to the main part. For an arching template, I start with the arch shape and work as much as I can on to the frame. Then I fill in the other parts, such as around the edges, as they are not as easily accessible. I mix small batches to fill in any remaining spaces. As far as possible I try to cover the full surface at once, because the margin can be visible between the almost-cured putty and any later additions.

The moulding putty can wrap around the overhang of the plate to take the contour of the edgework (figure 2). When removing the cured mould, it can be slid to a narrower part of the instrument to lift it off safely. Where the instrument is narrow, or on a tight recurve, it is best to wrap only on the best-preserved side.

FIGURE 2 How the profile fits on the instrument

FIGURE 3A Making the reference side of the profile

After the putty has been removed from the instrument, the rubbery material can be easily trimmed with sharp tools. I trim the PVS to make the reference side of the mould flat, with the frame material as a guide (figure 3).

FIGURE 3B Front and back of the reference guide

I always note whether the impression was taken from a free plate pressed flat on a surface or on the body of the instrument, with or without string tension. All of this information impacts on the interpretation of the records. Also, if you plan to prepare the frames using a laser cutter, make sure you use a material and machine settings that won’t burn the sides too much. PVS doesn’t hold on to ‘charcoal’! A CNC approach could work well too.

There are situations where the strings, bridge and tailpiece can’t be removed from the instrument while gathering the data. But it is still possible to record the accessible parts of the arching by making purpose-built frames that go under the strings and over the fingerboard or tailpiece. For this operation, the strings and accessories are protected with plastic wrap and paper (figure 4). In some places the missing arches have to be estimated, but since the frames connect both sides, the information is levelled and almost complete. For this purpose I use thicker 6mm plywood for my frames, so they do not flex or distort so much.

Now the arches can be traced on to paper, to be stored as a reference (figure 5). Positive versions can also be made on a similar principle, but using a harder compound such as a casting material like polyurethane (figure 6). I personally prefer to record the full curve above and under the f-holes and in front of the bridge, so I can record the ‘bridge island’ area. Lastly,

FIGURE 4 An example of how to record partial arching

FIGURE 5 Tracing the moulding on to paper

I note whether the arching curves were taken with the strings on, as the curves may vary with string tension. For the same reason, As well as keeping the paper reference, I keep the profiles in a folder for future use, as in figure 6.

It is important to be as precise as possible in centring the instrument’s upper and lower bouts, and securing them in place. For this I use a cradle with gentle pads holding the violin on the edges (figure 7). The cradle fits within a jig that supports a flat table made from thick plywood. It is hollowed to accommodate the outline of a violin body and a reference grid printed on paper is taped on it. This reference surface is supported at the four corners by threaded knobs, allowing height adjustments. It can be used to clamp the positioning guides and also as a reference to note the frame position. It is common for the body of older instruments to be distorted with noticeable bends or twists. So when placing the instrument inside the jig, I take measurements from the edgework to the reference surface. For this reason, it is important to use a surface that is perfectly flat. Writing the measurements down on a drawn outline gives a sense of the distortions that have occurred over time, which helps interpret the shapes of the arching.

FIGURE 6 The 3mm-thick positives are stored flat in a folder

To orient the arching information that I’m recording, I affix a piece of graph paper with a grid on to the table surrounding the violin (figure 8). I have already drawn a rectangle bigger than the instrument outline. I establish the centre by measuring to the purfling line on either side of the upper and lower bouts. This theoretical centre line of the instrument, which may or may not align with the centre joint of the plates, is continued on the graph paper. Once everything is aligned I secure the violin in the cradle. Then I can position the frames to make the arch mouldings. The grid helps to ensure the profiles are perpendicular to the centre line, and I measure their position from the line aligned with the top edge.

On my prepared sheet of graph paper are several horizontal lines in different colours. This helps me to see the correspondence between the bass and treble sides of the diagram.

IT IS COMMON FOR THE BODY OF OLDER INSTRUMENTS TO BE DISTORTED WITH BENDS OR TWISTS

FIGURE 7A A violin is secured in the cradle

FIGURE 7B Both the cradle and violin placed in the jig system

FIGURE 7C How the system looks once the reference plane has been placed and adjusted

FIGURE 8A violin is secured in the cradle

FIGURE 9A Marking the sides to find the centre of the template 112

FIGURE 9B The positions of the frame’s profiles on a violin back

FIGURE 9C The same grid sheet can be used to trace the profiles

I measure the distance of the mark from the top line (aligned with the upper edge) and written on the arching profiles’ reference side. Then, to work out the position of the centre line on the frame, I trace the bass and treble line on the arching profile at the same distance from the outside of the grid, so that I can deduce the centre later. Unless the exact position of the arching profiles is carefully measured and recorded, the profiles will not be useful as a future guide.

The number of cross-templates may vary according to your needs (figure 9). I would generally record:

• the upper and lower bouts (UB and LB) at the purfling’s widest point

• the middle bout (MB), and upper and lower corners (UC and LC), all at the purfling’s narrowest point

• on fronts, maybe the f-hole positions by measuring the arching under and above them (instead of MB and LC)

• at the stop, or f-hole nicks on the front, and the middle of the length on the back

• maybe halfway between the top and upper bout (‘#1’) and halfway between the lower bout and the bottom (‘#8’)

This procedure is not just limited to making arching templates. In instances when I’ve had very limited time with an instrument, and have wanted to keep a record of the toolwork in the fluting with the edgework, I’ve taken a part of this section with PVS. I align the impression perpendicular to the purfling, as much as possible, to record a part of the arching recurve, the blending with the fluting, and the overhang of the edge (figure 10).

I use the same method to study cello neck profiles (figure 11) and scroll fluting profiles, on various places along the scroll outline. It allows me to preserve a physical note of the gouges’ curves used to shape it. Those profiles have also been helpful with neck shaping, in situations where the customers wanted a shape similar to a neck they particularly liked.

This method of recording arching outlines has served me well at times when I wanted to document a violin, as it gives the most reliable details possible within time constraints. While the preparation of the jigs and frames seems like a significant time investment, their reusability saves hours later when opportunities for measuring arise. This method is faster than manually fitting the templates to the arching, and very useful for recording complicated shapes and locating the cross-arching positions precisely. Moreover, this way of archiving arching profiles does not take as much space as storing full plaster casts of plates, while still giving considerable information at a ratio of 1:1.

FIGURE 10 Studying an edgework fluting profile, having previously prepared two partial positive casts

FIGURE 11 Example of cello neck profiles.

THIS METHOD IS VERY USEFUL FOR RECORDING COMPLICATED SHAPES

The size discrepancies when tracing from a hard positive impression are minimal, and the impression is a step up from distorted photocopies of photocopies, or wavy profile gauge tracings. From the viewpoint of preserving instruments, having access to the hard positive impressions protects the original instruments from unnecessary manipulation without sacrificing the integrity of the records. Access to these measurements is not widely available from museums, and this information is valuable to people in our trade. I wish more of these were available in public archives when full access to industrial micro-CT scanning is not possible.

The moulding with a frame system is a concept developed by Stéphane Vaiedelich and described in Damien Sainton’s internship report: Rapport de stage: Participation à la documentation du violon Stradivarius 1716 ‘Provigny’ (August 2003). It is publicly available at the Cité de la musique in Paris under the archive number E1730. I was introduced to the use of PVS through Janet Starck-Toon’s article ‘Building better neck-graft counterforms’, published in Volume 2 of The Conservation, Restoration and Repair of Stringed Instruments and Their Bows, ed. Tom Wilder