Building a Liuto Attiorbato
There is a rich repertory for this instrument, with printed books by Piccinini (1623 and 1639), Melii (five books, 1614 to 1620), and Gianoncelli (1650), even continuing into the 18th century with Zamboni (1718), also many manuscript sources. Kapsberger (1611), since at one point he uses 11 courses, might also be added to the list.
There are also plenty of historical examples of this instrument which survive in museums. The best known were made by Matteo Sellas in Venice in the 1630s. The classic form of this lute has 14 courses, all double, with seven courses on the fingerboard and a further seven bass courses (played open) which go to a second pegbox as on a theorbo, hence the name. The top seven courses use the normal lute tuning, and the basses are tuned to a scale, so it is essentially a seven-course lute with a further octave of basses.
Not all the surviving instruments have 14 courses, and sources of music suggest that earlier versions of this design often had fewer courses. Piccinini’s requires 13 courses, with the 13th course tuned to a high note (typically f#) rather than the expected low G. In order the play the entire surviving repertory, however, one needs all 14 courses, with the 14th course tuned to low F, making a range of three octaves across the open strings. These instruments were made in various sizes, the most common size having string lengths of about 58 and 85cm and another (presumably tuned a tone lower) about 64 and 93cm. I decided to go for the larger size in order to minimize any problems in trying to accommodate so many strings, and a correspondingly wide bridge, on such a small instrument. The actual pitch used historically is of course unknown, but modern experience with gut strings suggests an upper limit of about a’=415 for the 64cm lute (with the top string at nominal g’), with the double first course suggesting a still lower pitch, perhaps about a’=392. If this seems low, it is worth remembering that the breaking pitch of a string is not the only factor – one also has to consider the tension which is acceptable to the player and the instrument.
Originally the strings would have been all gut, with no metal-wound strings (which had not yet been invented and in any case were never, I believe, used on lutes before the 20th century). I used gut for all the thin strings, and Savarez KF (a kind of harp string) for the basses.
Many of the surviving instruments of this type have ribs of alternate materials, often in ebony and ivory, and elaborate marquetry neck and pegbox veneers in the same materials. A predominantly black and white colour scheme is difficult to implement in modern times, when ivory is unavailable, and any wood substitute is prone to being accidentally stained by absorbing black dust from the ebony. Using alternate ribs of different colours is perhaps most effective when there is a big contrast between the dark colour and the light colour, so the dark colour could be ebony, but what to use for the light colour? Holly is extremely difficult to find in pieces which are large enough and white enough to make ribs out of, and ebony is a very unsatisfactory material for ribs because it needs to be very thin in order not to be too heavy and is very prone to splitting. I therefore abandoned the idea of having ribs of alternate colours and made the body out of rosewood with white (holly) lines between the ribs:
The bodies of the Sellas instruments are all very similar, with 15 ribs and a rather shallow section:
Several of the Sellas lutes have fingerboards with three decorated ivory panels, which on one of his instruments shows a battle between the Venetians and the Turks, with the latter being vanquished by the superior firepower of the former! Leaving retrospective political correctness aside, I felt that keeping a mostly white fingerboard clean was too difficult a prospect but also felt that a plain ebony fingerboard was a rather large expanse of black to leave completely undecorated, so I made a panel which represents a nod in the direction of the original instruments and helps to relieve the blackness:
Even having decided to go for the larger model, I still had considerable worries about the width of the bridge – would the ends of the bridge come too close to the edge of the soundboard and inhibit its acoustic response? Another consideration is that the string spacing should be narrow enough that it is theoretically possible to play the top string and the bottom string at the same time, so my customer’s handspan was another factor. Perhaps more important, though, is that the spacing of strings within a course must not be too narrow, or there is a danger of the strings clashing together. This issue is further complicated by the fact that the 8th to 14th courses are considerably longer than the others and might therefore be expected to have a bigger amplitude of vibration. Also the spacing of courses must not be so narrow that it places impossible demands on the right-hand technique of the player.
All these conflicting considerations gave me much pause for thought. In the end I settled on the widest spacing I felt I could get away with, making the 8th to 14th courses slightly wider than the others. The total width across the strings at the bridge was 155mm, perhaps pushing the limits as far as the “handspan” issue was concerned, but with any luck satisfying the other requirements. I made the bridge out of pear, and made the decorative veneer on the top quite thin, hoping to keep the total mass down and retain some flexibility:
Another worry was the spacing of the courses at the top nut. Again I went for a relatively wide spacing, but also spaced the strings within a course wider than the spacing between courses, on the grounds that you never play two adjacent bass courses at the same time, so clashes between the top string of one course and the bottom string of the next course are less likely than clashes of the two strings within a course:
There were further complications with the upper pegbox. Unlike a theorbo with single basses, where there are only eight pegs in the upper pegbox, here there were 14. What was the best layout of the pegs? In a lute pegbox (as in the lower pegbox here) the first peg – the one nearest the nut – is on the bass side. In the upper pegbox, the surviving liuti attiorbati seem to vary, some having the first peg on the treble side. I couldn’t understand this at first, but of course one reason could be that the last peg on the bass side is therefore slightly further from the nut, and the string (the low octave of the 14th course) therefore traverses a slightly shallower angle as it goes from the nut to the peg, making it less likely that the string tension will tend to pull the peg out of its hole. The string at the other end of the nut (the upper octave of the 8th course) is routed in a more or less straight line to its peg anyway.
Another thought which occurred to me was that which peg was nearest the nut would also determine whether the last peg (at the far end of the pegbox and disconcertingly “round the corner”) was on the bass side or the treble side. At that stage I was uncertain whether this string would be routed to its peg inside or outside the pegbox, so this seemed like a live issue, and again this seemed to favour the layout with the first peg on the treble side. As it turned out, both strings of the 11th course could be routed to their pegs inside the pegbox.
In a “normal” lute (of say seven courses or less) the bridge is placed more or less symmetrically on the body and both the highest string and the lowest string must line up with the edges of the neck, which means the centreline of the neck must be more or less the same as the centreline of the body; but on lutes with many bass strings the neck is offset to the bass side. In the liuto attiorbato (as on most types of baroque lute) it is still desirable to have the bridge more or less on the centreline of the body and the treble edge of the neck is therefore determined by the placement of the first string. On the bass side, however, there is no requirement for any of the courses to line up with the edge of the neck and the neck width is therefore somewhat up for grabs. At the body end, presumably the only requirement is strength. At the level of the lower nut, the same consideration applies, but the width of the neck here also has an effect on the aesthetics of the instrument. At first I was surprised by how wide Sellas made his necks, then I realised that having a relatively wide neck at the level of the lower nut meant that the upper neck was only slightly offset from the lower – whereas if you make the lower neck narrower, you need more offset to get the upper nut in the right place. In other words, Sellas seems to have preferred to have most of the required offset in the relationship between the body and the lower neck, rather than having a more pronounced “double bend” in the instrument.
Soundboard and barring
The surviving instruments have essentially the same pattern of barring as other late 16th-century/early 17th-century Italian lutes: two large bars either side of the rose, two further bars between the rose and the neck block, and two between the rose and the bridge; behind the bridge, a J bar on the bass side and two treble bars, plus the usual complement of thin bars under the rose.
This looks like an awful lot of bars on an instrument with such a short body (shorter than a normal seven-course lute). But it is not so strange when one reflects that with 28 strings instead of 14, the whole instrument has perhaps twice as much total string tension as a seven-course lute. I wanted to use a relatively low tension of about 25 Newtons per string, but even that gives us a total tension of about 700 Newtons! I decided to make the soundboard just a little thicker than I would for a seven-course lute.
The finished instrument
A double top string takes some getting used to, but works well. Overall, the sound is bright, full and sustained, and very clear. You can see video of the construction of this instrument below: