Hidden in plain sight

The Stradivari moulds preserved at Cremona’s Museo del Violino still have secrets to give up. David Beard re-examines all 17 survivors to find how they reveal details from rib heights to the maker’s working method

Points of interest to violin and bow makers

Antonio Stradivari’s making process has been the subject of study and myth-making for the past 300 years. Luthiers and researchers alike have studied everything from the varnish to the bee-stings to learn about the evolution of his personal method – but are there clues hidden in plain sight? In this article I will be looking at the moulds preserved in Cremona’s Museo del Violino (MdV), exploring their features and their use, to see what secrets they might hold.

The MdV collection holds fourteen moulds for violins, three for violas and one for a piccolo violin, while other artefacts suggest his workshop also used nine or more cello moulds. Excluding perhaps two in Paris, none of the cello moulds have survived. There are common features to all the moulds, including pairs of lateral ‘cross-lines’ running across the mould for both the lower and upper corners, as well as marked centre lines, rib guides, and patterns of drilled holes. These were all part of the Old Cremona system of posts and counterforms used to bind sides against the mould while bending and gluing the ribs. Of course, they all have cut out-recesses to receive the corner-blocks and end-blocks as well. While not overtly displayed, the moulds share the same sort of geometry seen in the instruments (figure 1). Lastly, most of them show one or two identifying letters. Some are also dated.

Details of the moulds vary along several dimensions. The progression of such details helps reveal the order in which the moulds were made. In the earliest, for example, the calculation and presentation of the rib guides hasn’t settled into its later pattern. Furthermore, visible ‘pry cuts’, made by the luthier to free the finished sides from the mould, are absent from the first moulds. Likewise, Stradivari carved or marked dates on to his moulds after 1689, but not before.

Count Cozio di Salabue, who owned and studied the moulds for a significant period of time, and others complicated the evidence by retouching original markings and adding text, but we can still surmise that the identifying letters originated from Stradivari. As well as appearing on the moulds, these and similar identifiers appear on many artefacts. They give the appearance of being inked with many pens, in many sittings, over a long period of the master’s career.

The mould groupings shown in figure 2 mostly agree with conventional wisdom. However, I have suggested the ‘G’ mould might belong before 1689. I’ve also suggested grouping the ‘PM’ (MS1062), as a companion to ‘MB’ (MS1). Historically, the ‘PM’ was missed in most earlier lists of the moulds. However, the features seen on the ‘PM’ mould are entirely consistent within the rest of the group. Hopefully, inclusion here will invite greater public discussion and examination of the ‘PM’ mould.

The rib guides are the pairs of arcs marked near the centre of most of the moulds. They are highlighted in red in figure 2. The most typical of these markings show a centre mark and a pair of arcs. But the earlier moulds show variations leading to those typical markings. The ‘MB’ shows a circle to give the main rib height. And the first viola mould shows rib heights with reversed pairs of arcs.

THE PROGRESSION OF DETAILS IN THE MOULDS HELPS REVEAL THE ORDER IN WHICH THEY WERE MADE

These rib height markings also help us by giving heights for blocks. The top-block takes the reduced guide height, while all the other blocks use the full guide height. Excepting the ‘MB’ mould, all the moulds display pairs of longer and shorter arcs, denoting the full and reduced block/rib heights.

Leaving aside the smaller ‘Q’ and piccolo moulds, the rib guides on the violin moulds run from 31.2mm to 32.6mm (full), averaging 31.7mm, and reduced from 29.5mm to 30.6mm, with 29.9mm as the average. The tenor viola shows 42.8mm and 39.9mm. The contraltos average 38.8mm and 36.6mm.

To shape the blocks we can use templates. Then we need to ensure that the blocks and counterforms fit well together. The traditional method is for the ribs to be bent to the mould and then glued to the blocks (figure 3, page 72). In modern making, the ribs are mostly bent by heat. However, there are further possible methods. For example, these moulds are also conducive to ‘wet bending’: you can soak the ribs, then bind them to the mould. Once dry, they will retain the bound shape. In a couple stages of wet binding and drying, you can bend straight ribs into a good fit on the mould.

ALL IMAGES COURTESY DAVID BEARD

FIGURE 3 The traditional Cremonese way of bending the ribs to the blocks

FIGURE 4 The bottom-block of the ‘T’ mould shows a radius of 4/5 the mould’s full length

What balance of heat and water bending the old masters used isn’t known, but we do have clues that water bending played a role. The general skew and irregularity of classical sides is something strongly promoted by water bending. We also have specific examples, like the ‘Messiah’, where in one place, the rib surface in the flank of the lower bout has dimpled out from a wet/dry cycle of bending, with the rib bound under too much pressure against the mould. After the ribs, the linings are completed, with the centre bout linings mortised into the blocks.

But what if we lack templates for the block shapes? Not to worry. The moulds show all we need, if we know their secrets. The end-block shapes are just sweeps of a long arc. As with the instruments, the long arcs for these Cremona moulds take their radii as thirds or fifths of the body length. Testing against the mould makes clear which radius was used. Figure 4 shows how the lower block of the ‘T’ mould shows a 4/5 radius.

To recover the geometry for a cornerblock shape, we first need to know the main centre bout circle. With the Stradivari moulds, the diameter of this circle will match the width of either the upper bout, or most commonly the lower bout. Testing against the mould will clarify which. The corner recesses themselves will show where the corner tip is meant to be. Just extend a vertical line from where the bout curve reaches the corner recess. Then extend the cross-line to intersect it.

THE RANGE OF SIZE AND VARIATION IN THE STRADIVARI INSTRUMENTS IS GREATER THAN MIGHT BE EXPECTED FROM THE DIFFERENCES IN THE MOULDS

The outer corner circle’s radius is given by the distance between the pair of cross-lines. The primary inner corner circle is either a reuse of this, or a simple ‘part’ of the main circle. A third or a fourth are most common. In most cases, the primary inner circle will take our curve shape into the corner recess, but not to the corner tip. We then use a ‘helper’ circle to complete the shape to the corner. This can be anything from 3/4 to 2/3, 2/5, or 1/3 of the primary circle, but 1/2 is most common. Testing will again show which to use.

These principles reveal how the moulds display their original geometry (and that of the instruments) in plain sight. But they also show us how we can create variations.

For a modern violin manufacturer, 17 moulds would translate simply into 17 production models. But we don’t see such a direct relationship between Stradivari’s moulds and the resulting range of instruments. The range of sizes and variations of details in the geometry seen in the Stradivari instruments is greater than might be expected from the differences between the moulds.

Asymmetries are another big part of this. The instruments show strong asymmetry and irregularity in their corner placements and shaping, in the square and curve of the ribs, in the overhangs, in their centre bouts, from the treble to the bass side, and even in dimensions from back plate to top plate. In contrast, excepting wear, the moulds are notably symmetrical.

Perhaps it’s easy to understand how the Old Cremona system of using thin wooden moulds, with its methods of posts and binding, might not truly control the squareness of the ribs. But let’s also look at how the sides lead to instruments.

Once off the mould, the sides are very flexible. In the working methods of the Old Cremona making families, as described more than 20 years ago by Roger Hargrave, the finished sides with neck attached are pinned to a board for the back. These pins fix their length and their position on the board. But this also allows the whole side and neck structure to pivot and twist at the top-and bottom-blocks, around those pins. These flexible sides can also be pushed in, in order to decrease the bout width, or pulled out to increase it. Also the corners can be pushed around considerably. But each such adjustment will flex and change the curved shape of the sides through the flanks and bouts.

FIGURE 5 Four examples of asymmetry in Cremonese violins – even though they were made on symmetrical moulds

ALL IMAGES AND DIAGRAMS COURTESY DAVID BEARD

The sides are then arranged as desired and clamped in place, while the profile of their resulting shape is etched into the board for the back. It is this etched shape of the sides that guides the creation of an instrument’s actual back outline geometry.

We can see how this whole process puts an indirectness between the mould design and a resulting instrument design. A single mould can yield not only many instruments, but also many variations of design and even length. We can also see in this the origins of the characteristic irregularities and asymmetries of classical Old Cremonese making. We can see why actually asymmetrical design choices were made in some classical instruments, though the moulds are symmetrical.

The 1669 ‘Tullaye’ violin gives a striking example. Built on a symmetrical mould, the ‘Tullaye’ has the centre bout on the bass side, sized off the lower bout width. Yet Stradivari, needing to follow the actual disposition of the sides, ends up asymmetrically choosing to size the treble side off the upper bout.

We can observe similar asymmetric choices of the centre bout sizing in other classical examples, including the 1664 Andrea Guarneri tenor viola, Stradivari’s 1684 ‘Lord Elphinstone’ violin, and the 1742 ‘Lord Wilton’ by Guarneri ‘del Gesù’ (figure 5).

So we see the Old Cremona masters using a building system that starts from cleanly designed moulds, but which inherently pushes them into irregularities and asymmetries. Perhaps the next secret to unravel will be whether this aspect of their making was a weakness, or a virtue.