DESIGN AND CONSTRUCTION OF ORACLE, 24-STRING HARP-SYMPITAR

Jeff and I met at an Alex de Grassi concert in Palo Alto, CA, a couple of years before I began work on his instrument. As the time when I would be able to start building Oracle drew nearer, we began to correspond about design, and he visited the shop to choose woods. The basic idea he presented me with was this: to build an instrument dedicated to the late guitarist Michael Hedges, based loosely on the old Dyer harp-guitar that Hedges had used, but with additional sympathetic strings running inside the main neck, as on my 18-string Sympitars. He wanted something more streamlined and ergonomic than the early harp-guitars tended to be, and he wanted to incorporate a Michael Hedges drawing into the visual design somehow. This was where we started; my job was to design something that satisfied both of us as a visual and tactile work of art as well as having superb function as a musical instrument. And then, of course, to build the thing!

Jeff emailed me a couple of quick drawings he'd had his brother make, representing the spirit of what he wanted.

I spent a while brainstorming ideas and made a number of rough scale drawings to try and find out what direction to go in. Taken out of context these ranged from the crude to the ridiculous. Here's a sampling of a few of my more coherent "studies":

I eventually came up with a design for the basic sculpture of the instrument that we both thought would work:

It kind of amazes me to look at this very preliminary drawing, done roughly to scale, and see how closely the finished instrument follows that first rendering.

The path that led from one to the other took most of a year to travel, and involved a lot more than just turning trees into dust, although I did my share of that in the process.

Jeff chose Sitka spruce for the top and Indian rosewood for the back and sides. Since the body is a full 30 inches long, standard guitar sets wouldn't work. Luckily I had a long billet of Sitka that had been given to my partner and I years ago by a violin builder; I suspect it was intended to be a cello top but wasn't thick enough for that. I was able to resaw it into a beautiful top set that was just long enough. The back was trickier. I was planning to use rosewood I had in stock that Jeff had picked out, and had resigned myself to having to do some pretty strange joinery to get the necessary length, when I had an inspiration. The side pieces were plenty long to cover the whole length of the back, and because of the instrument's design, I didn't need any single piece for the sides to be longer than the 20 inches or so that is standard for a guitar back blank. How about if I switched the use of the pieces? The obvious problem was that I couldn't get nearly enough width from the 2 side blanks to cover the entire back area. As luck would have it, I had one more rosewood back and side set in stock that came from the same tree as the set Jeff had chosen; the grain matched almost perfectly. I was able to join the 4 side blanks together and arrange the grain to match so that it's very hard to tell there are 4 separate pieces.

Once joined, the top and back plates were cut out, inlays were done and then the bracing was glued on:

Next the walnut neck was fabricated, the hollow graphite composite reinforcing tube made and put in place, and the pegheads laminated, shaped and drilled:

The neck and bass head were attached to the braced top before the sides and back were attached. This method of building up neck and body as one unit is the old Spanish style, still common in classical and flamenco guitars but unusual in the world of contemporary steel-string guitar building.

Next the sides were bent using the traditional hot pipe method, trimmed to fit, and attached to neck, corner blocks and end block. Little triangular blocks of wood called tentallones or dientes were glued in around the perimeter to affix the top to the sides. A kerfed or slotted lining strip was glued to the edge of the sides that the back will attach to.

The universal worktable was fitted with sliding L-posts that were formed around the sides to hold them stable during gluing.

The posts have little spring-clamps that attach to the top and hold the back down tight to the sides while the glue sets, assuring a good joint all around the perimeter. A few added go-bars reach areas inaccessible to the post-clamps

When the glue set, the assembled body was removed from the adjustable form, the back and top plates trimmed flush with the sides, and the channels routed for the edge binding and purfling. The binding and purfling strips were glued in, held in place with tape.

With the binding on and scraped down flush with the body, the fingerboard was fretted and fit to neck and sound hole, then glued to the neck. Up until this point the neck was left partly square, for ease in holding in a vise.

The final carving of the neck was done and the instrument was ready for sanding and finishing.

While the waterborne lacquer on the top and the natural drying oil finish on the rest of the instrument were curing, I worked on the complex shaping of the ebony bridge and the routing of the channels for the 12 individual RMC pickups that comprise the 2 hexaphonic sets. Once the top finish was rubbed out and polished, the bridge was attached and the long job of setting up the instrument for playing began.

For a more in-depth look at Oracle, check out the website Jeff Titus has devoted to all aspects of his Harp-Sympitar.