Early in 2020 I built an amplified cristal baschet style instrument which I have composed a number of works for (listen to these works here). A large part of the building process was dedicated to tuning the instrument, and once everything was cut, filed and polished the tuning process took a few days, a few deconstructions/reconstructions, and a few headaches. The two instruments I built are tuned to a 30-pitch-per-octave 31-limit just intonation system I devised, and only one octave is available. It is a limiting system. The lattice is shown below.
There are a number of factors affecting pitch, these are:
- position of tuning weights on the threaded bolt;
- Dimensions of tuning weights;
- position of cable lugs on threaded bolt;
- length and thickness of threaded bolts;
- glass rods.
Additionally, there are some factors that affect the pitch stability (i.e. noise and interesting timbres versus pure sine-wave-like pitches), these are:
- tightness of nuts;
- angle of cable lugs;
- thickness of base plate and positioning of threaded bolts.
Position of tuning weights
Imagine that the tuning weights are like the nut on a stringed instrument (that is the little piece of ebony wood where the scroll meets the fingerboard), altering the position of the tuning weight effectively gives you a different length 'string' (threaded bolt), longer strings make deeper pitches, shorter strings make higher pitches. Tension is not really a factor with cristal baschets.
Dimensions of tuning weights
Because I was building an instrument that I wanted to be light and easily transportable I experimented using hollow tubing for the tuning weights. This works to a degree but I wouldn't recommend it, pitch stability is more reliable with solid weights. I ended up using two different thicknesses of tuning weight, and cut all my weights to identical lengths although I am aware that many cristal baschets use tuning weights of varying lengths. I found that thicker weights worked more effectively for the longer threaded bolts, but I have noticed other cristal baschets use the thicker weights for very high pitches also; my instruments do not have any very high pitches. I would be interested to hear your thoughts should you have experimented with different length tuning weights, please contact me if you have done this.
Position of the cable lugs
Initially this was not on my radar at all, and my placement of cable lugs was admittedly somewhat haphazard. Eventually I placed all of the cable lugs at 1/3 of the length of the threaded bolt to begin with, which should bring out the third harmonic 3:2 of the bolt. I have not measured the fundamental frequency of the bolts, nor have I done a spectral analysis of the instrument so I cannot say for certain if this placement of the lugs produces a tone that is predominantly 1:1 with strong 3:2 qualities, or if it simply produces the perfect fifth above the fundamental in the same way that a string player would produce the third harmonic on a string, again please contact me if you have knowledge of this curious situation. Once the cable lugs were placed at 1/3 of the bolt lengths, the final position of the cable lugs was used as a fine tuning tool, shifting them slightly up and down to achieve the desired pitch.
Length and thickness of threaded bolts
The threaded bolts I used for my instruments were M6 size (6mm diameter), but other cristals I'm sure would use different size bolts. Just as the length and width of a violin or guitar string affects pitch, so too does length and pitch of cristal baschet metal rods. While this seems simple and obvious, when you actually come to make an instrument you're faced with the question of 'How long do my rods need to be to achieve the pitches I want?' and this is a difficult question to answer precisely because there are so many other factors affecting pitch and as there appears to be no standard way to produce a cristal baschet style instrument, every instrument is going to be a bit different. The way I worked around this problem was to cut my rods to lengths that differ by approximately 2-3cm. So for one instrument I have three rods that are 36.5cm, three rods that are 33.5cm, three rods that are 30.5cm, two rods that are 28cm, and two rods that are 26cm, and one rod that is 24.5cm. Each rod produces a slightly different pitch which is achieved through altering the other variable elements of the mechanism.
Glass rods
Based on my experimentation, the glass rods themselves are not producing pitched sounds (when bowed using the traditional method for bowing the instrument, for extended performance techniques see my recent post here). The glass rods themselves aren't actually the resonating body, they are simply the medium for transferring energy from your fingers into the threaded bolts, and it is the threaded bolts that are singing. This seems to me to be the reason why the term bowing is often used in association with the instrument, as a bow itself does not produce sound (in this case we have a glass bow), but instead creates the correct level of friction and transfers energy into a string which then produces sound. Keep the idea of bowing a violin in your mind for now. There are physical limitations any string player faces that are determined by the bow. As a performer bows their instrument, eventually the player will run out of bow to use and needs to change direction (ending one note and starting another), louder notes require more bow and cannot be held for as long as very soft notes. All of these concepts apply to the cristal baschet and while it seems obvious in hindsight, I wish someone told me this before I built my instrument, I probably would have opted for longer glass rods (longer bows) which would have allowed me to create longer notes.
Tightness of nuts
The tightness of all twenty nuts per mechanism (yes, twenty!) is critical for pitch stability. If you have loose nuts somewhere, you'll know about it because you will be driven slightly crazy by buzzing sounds that you can't locate. You will get good at tightening nuts if you make a cristal baschet style instrument. If one of the ten nuts that are attached to the M6 bolt that resonates are not correctly tightened, the pitch of the bolt will have a tendency to fluctuate wildly (in some cases this may be desirable) but tightening the nuts will address this so that the pitch is relatively stable. Imagine how a guitar string sounds when it is not correctly tightened, this is the same problem you are dealing with when the nuts are slightly loose on a cristal baschet, but the problem is far less obvious on a cristal because you can't see or feel that there is no tension like you can with a guitar.
Angle of cable lugs
This is a discovery I made in my pursuit to make a cristal baschet style instrument that packs up into a suitcase. I initially ordered 50ea 90 degree cable lugs and when I began attaching the glass rods I discovered the instrument wasn't quite making the sounds I had expected. This was because the glass rods and the threaded bolts were running parallel to one another (there is some physics at work here and when I come across the name for this phenomenon, I will update this). To produce pure tones the glass rod needs be perpendicular to the threaded bolt, not parallel. This is an interesting phenomenon, as it may not be your goal to produce pure tones, and the further away you get from a perpendicular angle (and closer to parallel), the more complex and noisy the sounds produced by the instrument become.
Thickness of base plate and positioning of threaded bolts
The base plates I used for my instruments were very thin, probably too thin, again this was the case due to my mission to create a light and transportable instrument. Observing other cristal baschet style instruments, you'll notice that often the base plate is quite thick, and often thicker in some sections than others. I have not tested this theory but I believe the thickness and flexibility of the base plate affects pitch stability. When I say pitch stability in this context, I am referring to the tendency for some threaded bolts to produce more than one pitch (not simultaneously, but in succession). I have spoken with other instrument makers who believe that the design of the resonators also affects this, as I have not built resonators and my instruments are amplified instead, I am unable to comment on this. Positioning of the threaded bolts in the case of my instruments was rather problematic. This may or may not be an issue for a cristal instrument tuned in equal temperament, but I suspect it was more of a problem for my instrument because for every one pitch there is in equal temperament, my instruments have two and a half. I found that two bolts positioned next to one another and tuned very closely would often want to change their fundamental frequency to whatever frequency the other bolt was resonating at. I overcame this obstacle by rearranging the physical positioning of the bolts so that the layout of pitches was not in ascending order like that of a piano. Instead, the lowest pitches are in the middle of the instruments, and the pitches ascend in both left and right directions in the same way that the pitches are arranged on an Mbira.
I hope I have been able to provide you, dear reader, with some useful insights about how a cristal baschet style instrument can be tuned and may you use my failures to avoid your own failures.