An essay by Clive Titmuss
The Vihuela-Playing Angel gives valuable clues to the size, shape and decorative features of the vihuelas of the mid-1500’s. Notice the presence of only five frets, and a bent-back, lute-style pegbox. The hand-positions would give me, as a teacher, nightmares.
A further point of interest is the fact that in the Iberian culture, the vihuela had effectively superceded the lute as an icon of angelic celebration, as it is widely observed in Italian and other traditions.
The Jaquemart-Andre Vihuela, was probably made in the middle years of the 16th c. It’s likely that it was built as an examination instrument by a luthier who wished to join the guild at the completion of apprenticeship.
The level and complexity of the decoration are probably the factors responsible for its preservation. It was converted to a five-course guitar, probably during the 17th c. The bridge location was clearly changed.
Its overall size and string length (around 72 cm.) means that it is a tenor or baritone pitched instrument, suitable for accompaniment, but of limited use in the difficult solo repertoire. The stretches required in the music of Milan or Mudarra would be all but impossible on this instrument.
But it does provide a model for the construction of the body shape, the decorative scheme, the structure and joinery of the head. The placement of the tiles is static, and the five-rose pattern is probably not a regular feature of contemporary vihuela design. The width of the neck, as with many extant instruments often copied or studied, is not suited to the modern hand.
The ribs of the back, made from nearly 200 pieces, and the sides, are both stabilized by a substrate, which would make the instrument rather heavy and probably not very responsive. This is often a difficulty when assessing historical models; their preservation owes more to appearance than musical utility. The good instruments were simply played out, rarely preserved.
The Quito vihuela, although it is probably an early 17th century example, is a valuable resource for the design of vihuelas, especially their overall size, proportions, bridge position, and layout.
This particular body contour is also observed in a number of Iberian baroque guitars.
The Chambure vihuela, recently re-discovered, has added an important resource to the knowledge of vihuela design. The barring and overall layout, length of the neck, bridge position and the vaulted back are notable details. The curves of the body are easily de-constructed with a straightedge and compass. Note basic nature of the barring, only two main bars, which are slightly skewed. This seeming simplicity produces an instrument with a wealth of overtones and allows the top to be thicker.
Paper drawing (which has the shape, bridge position, rose position, neck width and length), then a mylar tracing taken from it, a plywood half template, a full template which serves as a router guide (using a flush-trimming bit) and two moulds, an internal mould which serves to hold the body shape when the top and back are fitted and the shaped and block-ready main mould, to which the neck, bottom block and sides are attached. The white plywood marking template (on the main mould) locates the screw holes which are used to attach the neck to the mould.
Using the internal mould as a bending template, the sides are bent on a heated copper pipe, allowed to springback for a few hours, then bent again. I don’t wet the sides during bending to avoid distortion of the figured maple. The sides bend easily if their thickness is correct, about 1.5 mm. Final thicknessing before bending is done with a cabinet scraper and hand scraper.
Clamping sides to end-block using a fitted caul screwed to the waste portion of the block. When the neck shank is sawn through, the endblock is also sawn. In this picture the neck-block mounting block, which is mounted to the mould, serves to attach the end-block with a souple of screws in its waste portion. Care is taken to ensure that the tiles line up with the top of the mould line on the end-block and the centre-line of the instrument.
Before the sides are glued to the endblock, they are marked and sawn on a marked block which substitutes for the neck block (with sawn kerfs cut into it). They are then trimmed and fitted when the neck is screwed to the mould.
The sides must fit into the kerfs tightly (fitted with a file), but loose enough to allow glue to be applied. The neck block has a partly sawn-through shank which fixes it to the mould where the marking block is now seen, then it is removed and the neck is attached to the mould.
Once the neck is attached to the mold, the sides are fitted to the slots, trimmed to the mold top and bottom, and glued with thinned hide glue.The alignment of the neck to the centre-line is adjusted by having oversized screw holes and using washers, then tightening the neck in its correct centered position. When the carcass is complete–endblock, sides and neck all joined, the end-block shank and neck shank visible here is sawn through, freeing the whole from the mould.
Before that, the whole carcass is planed flush to the mould to ensure consistent shape. Slight dishing of the body is done to a mark on the inside at the waist with a finely set spokeshave before the soundboard is fitted.
The 3 mm panels of bookmatched figured maple are jointed, a piece of ebony is inserted between them and they are glued together in a jig made for the purpose which uses folding wedges and reject pegs. The wedges in the middle ensure that the underside is flush. The jig is waxed to make release easy. The same jig is used to join the halves of the soundboard.
The glued-up panels are thinned with a jointer plane and smoothing plane in a jig which has rails the same thickness as the eventual back panel thickness. Tools marks are removed with a scraper.
Generally speaking the panels are slightly thicker in the middle, thinner at the edges, a natural tendency of the procedure. Wooden planes with exquisitely sharp irons make the work pleasant and easy. Figured maple must often be planed across, not along the grain, to finish.
A similar procedure is used for the tops, but the wood may be planed more thinly. A variety of planes is used to get the best results.
Bars (in this case for the back) are made from yew, a very light and strong wood which has good acoustical properties. The bars for the soundboard are split from billets, rather than sawn. This simple jig holds the wood while it is dimensioned.
I taper the bars vertically by having the rails planed to differing dimensions. Lining the jig with sandpaper and using wax on the sole of the the plane helps to hold the work without clamps or adhesive. I remove the tool marks with a scraper. The bars need to be straight and clear, with the quartersawn face on the flat dimension.
Clamping the bars to the back with go-bars (made from fir). The cauls are the same shape as the bars. They are made from soft spruce and sandpaper is glued to them, then they are used to do the final shaping of the bars to the exact contour.
The bar adjacent to the heel is slightly taller and thicker, to make the instrument stiffer to resist the pull of the strings.
All the bars are dimensioned after glueing, tapping to hear changes in pitch. When the back is resonating with a high-pitched knock, and a low-pitched “thud” the back is about right.
Paring the bars of the back ready to fit the linings. The bars have been cut to length with the internal mould still in the carcass. A block is used to make each bar-end the same height.
You can clearly see the curvature and height of each bar. They progress from highest at the upper bout, to lowest in the lower bout. It’s important to avoid spacing the bars regularly to ensure that there is no periodicity in the back.
Tapping on the back and listening for a low and high pitched resonance ensures that the top and back do not have the same fundamental frequency. This helps to avoid some notes booming while others are woody.
The inlays are carefully tried in several positions to achieve a pleasing and authentic appearance. I used several basic ideas: the trinity, the four points of the compass around the rose, reflecting the age of exploration, the alpha and omega at the top and bottom, and the focus of the bridge in the centre of the lower bout.
Many vihuelas both modern and historical have too many elements, making them appear busy and confused. Five kinds of tile are used in all.
Once an aesthetically and historically satisfactory layout is arrived at, the various tiles are located by making a mylar template, which plots their exact positions.
A pencil traces the locations onto each of the soundboards.
I record the exact location of the nut, frets, body/neck joint, rose and bridge on a plywood batten, which serves to locate the elements and check their consistency throughout construction.
Each tile is carefully traced with the scalpel, then micro-carving tools remove most of the waste. A tiny gouge, a tiny mortising chisel, and a tiny scythe-knife do the delicate work. The tiles are bevelled slighty with a file to ensure that the tracing is undersized for a tight fit.
The bottoms of the mortises are levelled with a shop-made base and small router with a carbide bit. One may equally use hand tools for irregular inlays, such as those on baroque guitars. Thinned hide glue is put on the mortise and the tile is clamped with a cork-lined plywood caul and weights (planes).
The absence of bars on vihuelas and early guitars means that their soundboards are somewhat thicker than those of lutes, making inlays possible; It’s not really feasible to inlay tiles into a 1.5 mm soundboard. The soundboard here is about 2.5 mm in the centre, thinner at the edges, the reverse of usual lute procedure.