Sunday, May 30, 2010

More Shawms - really about tuning, this time

I was going to run the post on staples and tuning together, but the stuff on tuning was taking too long to come to me, so now they are split into separate posts.

Remembering having tuned the first shawm, I was in a mixed state of mind about doing this next pair.  On one hand, I had a working reference for where holes should pretty much be, and I knew that mistakes were almost certainly not terminal.  On the other hand, I was not completely impressed with the playability of the first shawm, and wanted to make the next ones better, as far as tinkering with the tuning would allow.

For practical reasons, I planned to continue to use bassoon reeds, and the staples had been designed on this basis, although I was playing the first shawm enough by then to get a feeling that a bassoon reed was not the ideal form.

I took measurements off the first shawm to where the holes were, using the end of the staple as a datum rather than using the end of the bell as I had done for the first one.  In retrospect this was an obvious thing to do, but it does rather require that you have a working unit to measure, which had not been the case for the first shawm.

These are the numbers.  If you are making a shawm, treat these as starting points only, and be prepared to fiddle.  All dimensions are in millimetres, and the holes number from the top down.

Hole numberDistance from stapleDiameter
11114.1
21657.3
31847.3
42368.0x7.7
52747.5
63037.6
73377.4x7.7
Tuning (2 holes)4168.3x8.7+7.3x7.9
Resonators (2 holes)4857.0+7.0

A pair of dimensions for a hole mean that it is somewhat elliptical, and these are the lengths of the axes.  I drilled the pilot holes in at right angles to the axis, but several holes, no. 4 in particular, have had the inner part of the hole angled in the direction of the reed.  This was a compromise between keeping the outer holes in a place where fingers can readily reach it, while getting an acceptable pitch without a huge hole (which would be more difficult to cover).

The first placement of the holes is best done in a single session using the same reed.  This should not take more than two hours or so.  Don't tinker with the reed too much, other than getting it sounding sufficiently well over the first octave range of the working shawm that you do have. If you don't have a working shawm, this is harder to do...

Since it was never going to get easier, I sealed the inside of the bore by plugging the narrow end and pouring in a measure of thin shellac, and rotating the instrument to thoroughly wet the inside of the bore.  Then a basic coat on the outside to avoid marking the wood with grubby fingers during tuning.

I set up a template in a CAD program of where the finger holes should be, and printed a 1:1 overlay.  With this taped on in the right place, putting the pilot holes in becomes a doddle.

Here is a shot of the second of the pair with the template taped on and tape markers for the resonators and tuning holes near the foot.  Holes are bored straight through the paper template, and it can be rubbed out of the way where not wanted with a gentle application of a rasp.

The hole positions have been rotated around in an bid to improve the fingering of the right hand on the more widely spaced lower holes.  Compared to other factors affecting tuning, this has next to no effect.

The left hand shawm is essentially complete, and you will see that the number 2 hole has been remodelled with wood filler.

Tuning works from the bottom up, so the resonator holes went in first (I have since read that these are more important than I first realised, and there will be much more about this in a later post).  Then the tuning holes, which set the bottom note "all fingers down" pitch.  Some messing around here will be needed to get a reasonable "C".  "Reasonable" means something close to a C for wind pressures that the reed and player can comfortably handle.  This amount of wind pressure becomes the amount of puff that should be used to tune the remaining notes, if the instrument is to be usefully playable.

Here is the first of these two shawms, close to complete in tuning, obviously before I reworked the no. 2 hole.  All the vital tuning tools are to hand: a selection of twist drills from 4.0 to 8.0mm in 0.5mm steps (out of the picture to the right), round and half-round needle files for cleaning up the inside ends of the holes where they enter the bore, water capsule to keep the reed wet while the next hole is bored (near the bell on the left), staple (out of harm's way at the top), and a little electronic keyboard as a pitch reference (out of frame to the left).

It's important that the holes have clean edges, particularly where they enter the bore.  When a hole is first drilled, there are usually whiskers at the bottom end, which are pushed back where they are difficult to reach from the outside.  I dealt to these with a rod of 6mm steel which has the end ground to a 60 degree angle and the sharp edge of this honed and stropped.  By running this rod in from one end or the other, it becomes quite easy to either trim the whiskers off directly with the cutting edge, or to push them across the hole, where they can be cleaned off with a small sharp file from the outside.

When I drill a new finger hole, the rough tuning is done against the keyboard reference, but the final tuning is done against the lower holes already cut, by playing an ascending scale and tuning the new hole so that its note sits comfortably in the scale with a similar intonation to the lower holes.

Small details: when I tuned the first of the second set of shawms, I would spin the twist drills in the forward direction when drilled the holes.  All good, but it would leave an untidy edge on the outer end of the hole.  I treated all these by shaping a small depression across each hole (visible in the photo above).  Visually this was not altogether good. 

For the tuning of the second of the shawms, I would spin the twist drills in the reverse direction when starting the hole.  The trailing edge of the twist drill is still sharp enough to cut the wood well enough, but leaves a cleaner edge to the hole.  The remainder of the hole was cut with the drill spinning forwards.  This is no problem to do with most electric drills.  The result is a cleaner hole, with no need for an extra depression, as you can see here.

I can't say that it makes any real difference in the feel of the instrument, but it looks better.

Finally, here they all are, together.  As you can see, the later shawms have slightly more compact bells than the first one, but have much more sleek lines, and generally look slinkier.

Musically, they are much of a muchness, and playability on any day depends much more on how the reed is feeling than anything else.

From this, I am deducing that I have done as much refining of the shape as is worth doing at the moment.

The next area for improvement is in the reed.  This is not an area that I know a great deal about, and I will need to consult and research.

The new shawms have gone to homes in the Barony of Ildhafn (Auckland, New Zealand), and their new owners will be coming to terms with getting good sounds from them.

To be continued.

1 comment:

  1. Ah, you must be the New Zealand shawm maker that was mentioned to me at Rowany festival.

    I have played around a little with making cylindrical bored reed instruments. My materials and methods are a lot less authentic: For reeds I use drinking straws (carefully thinned at the end, and I'm experimenting with a metal doodad to help shape the end). For bores I make two semicircular grooves with a router, then glue together.

    I have software to do the hole placement and sizing. It can do things like place holes given constraints on the spacing between pairs of holes, or that the spacing is roughly regular between three holes, or so that certain cross-fingerings work, or to keep things in tune between registers.

    I'm currently working on a shawm design, simultaneously optimizing the bore shape and hole positions/sizes. My computer claims to have a design that will work across two octaves, we'll see how it does in real life (and with my fairly limited woodworking skill).

    --
    Paul de la Ville (pfh at logarithmic dot net)

    ReplyDelete