For all you curious Whistlers and Whistlesmiths out there, here are some video mysteries of sound generation and the shapes that generate them. Introducing…
Note the response times, the oscillation sizes, the stability of the soundwaves. There are many more shapes to study if you are curious enough to research them. What are your thoughts on design modifications? Ideas on improvements? Can we discuss older inefficient designs that give poor results? Musicians please reply with complaints about some bad habits your whistles display too.
I don’t see that we can apply these results to whistles, at least not directly. The obstructions in these tests don’t look like a whistle labium because they have empty space behind them. A whistle labium has the whole long body of the whistle behind it, so the air flow above the ramp can’t interact with the air flow below the ramp. This effectively prevents the formation of a Von Karman vortex street.
I have seen images that show a vortex street forming above the ramp, but this would not be a Von Karman vortex street. I have also read opinions that this vortex street does not contribute to the whistle sound, because the frequency is far too high.
The real action in a whistle, I gather, happens before the obstruction, between the fipple and the lip. The oscillation starts as soon as the air emerges from the windway. There’s maybe half a wavelength of oscillation before the air strikes the lip. Somehow, this oscillation interacts with the vibrating air in the air chamber to produce the sustained sound, but I haven’t seen a solid explanation of how yet.
Yes, the studies i showed do not have “Labium Ramps” and you are correct in that the vortex street is created as the airstream exits the windway “lips”, but then the labium “splits” the vortex into upper and lower halves. The angle of the labium and leading edge of the labium can either support the vortex oscillations or hinder them. I was attempting to show these relationships.
Sharp,angled or rounded lip edges on the windway exit can also greatly effect the vortex. This is why I included the “Teapot” example. Modern Recorders have rounded upper and lower lips at the windway exit to eliminate turbulence but many cheaply made tinwhistles have sharp/square lips that do create airstream turbulence.
Here is an example of a flat surface, like the underside of many cheap whistles, that disruptes the vortex flow due to “surface tension”… http://www.youtube.com/watch?v=SyT0qrg57pA
Here is an example of a Vortex Street created in the same way as a windway exit in a duct flute. After creation the Labium splits this into upper and lower halves. If the labium is the right shape it will support the soundwaves and not cause turbulence. http://www.it.cas.cz/files/2-KVS_cooled_cylinder_sm.jpg
It is true that a standard vortex street has a specific frequency, but soundwaves react in a similar fashion. Aerodynamics and Acoustics are closely related.
Handmade recorders almost without exception - and I believe this is also true of baroque originals - have symmetrically chamfered (not rounded) windway exits.
I looked at your “surface tension” example but wasn’t sure how what I was seeing was relevant to windway design. It shows a street forming behind a cylindrical obstruction which causes a flap to oscillate. Are you saying that surface tension is causing the flap to move rather than a vertical pressure differential? Anyway, fluid dynamics was never my strong point.
But fascinating examples. The first one you posted seems to show the street forming behind the wedge-like obstruction and then backing up incoherently into the incident airflow, which is amazing to watch.
It would be interesting to know what the timescales are - how long does it take for a stable oscillation to form behind the windway exit at the kinds of speeds, pressures and other physical parameters typical of whistle or recorder design? This would explain the phenomenon of “chiff”, I think.
The 45 degree chamfered lip edge is a good compromise for a rounded lip and easier to “scrape” with a sharp tool, but modern plasitc Recorders all have rounded lips (easy with injection molding). The difference in efficiency is minimal, but the round lips produce no turbulence. Here is an example of a voicing that uses a rounded lower lip as well as cut “grooves” to reduce turbulent airflow at the window. http://www.organstops.org/o/LabialOboe.gif
There is “backpressure airflow” created in a duct flute caused by a pressure wave traveling up from the bore end hole/closest open tonehole to help create/sustain oscillations. A conical bore (Recorders), a throat restriction or slight reduction in bore diameter below the voicing will also provide faster response time in production of the oscillations, just like the wedge video
Chiff: This is an harmonic tone color created by the undercut chamber everyone seems to want to fill up with wax lately. It acts just like a high pitched panpipe tube. It is very similar to the cavity in a Flute. The size, diameter and depth of this cavity determines the frequency of the “Chiff” in all registers. It must be tuned and “in phase” with the keynote scale or sour harmonics will result. This is not the only source of Chiff though, some of it is produced by the harmonics of the bore shape as well (Benade “Horns,Strings and Harmony”).
In regards to Aerodynamics and Acoustics, I am an acolyte of Benoit Mandelbrot, the father of self-similar fractal patterns that permeate nature. The same patterns found in Aerodynamics also coincide with Acoustics. Study of both is enlightening. http://en.wikipedia.org/wiki/Self-similar
Well, not to be a smartass, but I think that if a rounded windway exit gave a better acoustic result, then makers charging a couple of thousand euros for their instruments would be using it, surely? Or is some turbulence good for a pure, focused tone in some counterintuitive way?
As for chiff, I always understood it to be a transient at the start of the note, whereas “breathiness” sounds more like what you describe.
But: if the street forms between the windway exit and the labium edge, then what function does the edge have? Does it simply set up a length (and hence presumably a fundamental frequency) for the “vibrating string” of the vortex street? Or does it excite the initial vibration, which then causes vortex formation in the gap between the windway exit and the edge itself? There’s no doubt that the shape of the labium is critical to the tone and response of the instrument.
The blade divides the air stream coming out of the windway. The air going below the blade will push against the air in the tube, increasing its pressure. That causes the air in the tube to move. i.e. some of it is pushed out, only to be sucked in immediately due to the fall in pressure inside the tube, etc. establishing a resonance frequency according to the length of the tube. If the speed of the air jet increases sufficiently, more energy is generated, which excites and maintains a higher resonance frequency (first octave note, and higher notes of the harmonic series with even higher jet speeds).
I think the turbulence in the air initially exiting the windway is causing the attack “chiff”, but is unimportant in the stable resonant tone when the whistle “speaks”. So this “chiff” may be the inherent resonance frequency of the vortex street. Blow a focused air stream with your mouth against the edge of a ruler or thin card. You’ll hear a fuzzy tone. That is basically the same as the “chiff” of a whistle before the note speaks, i.e. before resonance involving the whole air column in the tube is established.