Multi-Whistler Instrument

After thinking up this Idea, wishing I could afford to build it, and finally deciding to post it here, I looked around to see what there might already be, and only saw one thing (first link). This is definitely different in various ways.

From the subtitle one might first think of a steam-whistle calliope, (second link) but steam-whistles don't sound like human whistles. I want this instrument to make musical whistles that are so similar to human whistles that it is difficult to tell the difference. For examples of good musical human whistling, try the third link.

If one had such an instrument, it wouldn't matter how badly you whistle in the normal way, you could become just as musical by playing the instrument!

So, what are the mechanics of human whistling, such that it might be duplicated by an instrument? The Wikipedia article (fourth link) notes that the shape of the mouth and the position of the tongue can affect the tone of the whistle, while the thing that actually vibrates, to make the sound in the first place, is nothing more than the air itself, moving through the puckered lips.

If air flows slowly, it can also flow smoothly and silently. Faster- flowing air can become turbulent and noisy. The interior shape of the empty part of the human mouth, through which the air flows, is so adjustable (via jaw, tongue, and cheeks) that the quality of the noise can vary considerably, mostly from "breathy" to musical.

In terms of existing devices, there is a non-steam calliope that employs flowing air; at the fifth linked page you can hear the noise of the fan as well as the sound from the pipes. It happens that the pipe-sounds are fairly close to musical human whistling, but metallic. This tells us that to make an accurate-sounding whistling machine, we will need softer materials for the pipes...and it would be nice if the fan was a quieter type, too.

I'm sure the steampunk crowd would not want modern plastics used (which seem to be available in any reasonable degree of hardness or softness, but appropriate alternatives might still be findable (soft woods, hard waxes, et cetera). Experimentation is in order!

There is a notable thing about the expert whistling in the third link, and that is all the musical tones are fairly high-pitched. It seems that for the human mouth, the lower the pitch, the less musical is the whistle. This might be something that can be corrected when building a whistling machine; we would like to have as full a range of musical notes as most other instruments. Experimentation is still in order!

Finally, we definitely want this instrument to be able to produce simultaneous whistles of different tones, allowing either/or musical chords and the sound of a whole chorus of whistlers.