I have been playing with the design of my home-made whistle heads, and made some observations. I thought I would share these in the hope of eliciting some more discussion… (oh, whistles are bass-A ~32mm id pipe).
I have widened the windway and window to 20mm from 15mm (both measurements nominal - they usually increase slightly with filing/sanding). I was surprised to find little difference in the ‘air-resistance’ (back-pressure). I did get an increase in volume but smaller than expected. Is there a theoretical maximum volume that can be obtained for a given frequency and size of pipe? Or should I try increasing the window further?
Why increase the window? Well, the volume increase was accompanied by a change in tone. The tone sounded more ‘hollow’, more like the sound obtained by blowing across a wide mouth bottle. I thought this a nice tone, and wondered if I could increase the effect. I will need to move the b1 hole to increase its size, as it is not speaking as clearly as the others. It is a bit on the small size…
I was initially disappointed with the results - the upper notes did not wish to sound at all. Then I thought to I cut away at the top of the windway exit (anyone have a name for this part?) so that it was further back from the face of the fipple. This seemed to increase the ease of playing the higher notes quite markedly, and rescued a whistle head from the failure bin. Has anyone else had any experience of this effect? It would be good to know if this is part of a pattern. It would also be good to get some idea of the limits of this modification. What happens as I cut it back too far? I am loathe to find out the hard way, as a whistle head takes a few hours to make.
Oh, and Feadoggie, I did finally succumb and make a second whistle (to give to another C&F denizen). So you would have won that bet…
Anyway, my low with 22 mm id and 8 to 12 did the same thing. So, I am thinking 15mm. I am not sure where it falls off. For me, I suspect 12 mm. As you alluded to, the hole size and where will be affected.
I have a naf flute that is low D# in whistle language. Well, NAF also. It is a 1 and 1/8 bore, 3/16 walls and the tsh (window) is 3/8 by 7/32. Ok, you will have to use the converter like I do for mm. lol. sorry. It has a hollow, deep, like blowing across a jug sound.
I think you are on the right track. In my humble opinion, you may have hit the limit. I bet feadoggie will have more info than me.
BTW. Your calculator is great. Aside from that, the flutomat is the same as what I use for NAF. Ed Kort does one for NAF flutes that are tuned differently.
(1) TWJCalc can do different tunings if you wish/need.
(2) When you mention NAF do you mean a thing like a whistle, but with an extra chamber before the fipple/window/blade section? Does this make a big difference to the sound? I ask because such should be relatively easy to engineer into my whistle design… oh damn - more whistle making .
I know about the tunings. Your calculator is the best. I love it.
The NAF has a sound chamber, slowing chamber, nest (windway), and bird (block). The slowing chamber should not be needed. The crux of the discussion is still the nest, and the sound chamber and the tsh(window). The length of the nest is usually 3/4 to 1 inch long. It is usually the same width as the tsh (window) and usually only 1/32 deep. This equates to the length of the windway and the window at the end. In this case it is say 3/8 wide by 7/32 long. A plug would take care of the length by moving it closer or farther away.
To recap, on a whistle the windway would be 3/4 to 1 inch long from head of flute to the edge of the window (mouth end). The depth of this would be 1/32 which is usually not an issue in a whistle. The window or tsh would be 3/8 wide by 7/32 long. It is the width that you are messing with.
The slowing chamber does that slows the air down to pass it through and exit to the nest and down to the edge (which is under cut). I have found that the length and the hole placement for diatonic NAF impossible for low D or D#. I am having a frined make a diatonic NAF. I bet its key will be mid g or above. An A is likely.
My understanding is that one of the main issues in voicing a fipple flute is to control the turbulence in the air striking the labium edge. Recorder voicing generally tries to reduce it to a minimum, and one of the ways to do this is to chamfer the block and windway roof symmetrically with a sharp angle at the labium end. Another way is to move the windway exit closer to the labium.
In my experience from fooling around with recorders, moving the block (windway floor) up and forwards helps the top notes speak better, but it can weaken the lowest notes. It also creates a reedier tone which may not be congenial for whistle players. Recorder windways are also usually tighter at the exit and the block is (very slightly) concave.
The only whistle maker I know who voices his whistles a bit like a recorder is Phil Bleazey - and I find his high D whistle to sound more like a recorder than any other I’ve played.
But in any case: you’re voicing a cylindrical instrument in A below middle C - getting two octaves out of it is damn good going.
Interesting - is it known to be harder to voice a lower pitch whistle than a higher pitch one? That would make me feel slightly better about my results. I would not claim two octaves though. The first octave (A - G#) is a doddle. The next half octave (a - d) is reasonable, the next two notes (e, f#) are a fair bit harder, g is just about doable as a slide up from f#, and higher notes are horrid.
But my aim was for the low notes - if I want to play the higher notes I will move up-pitch to a low D. It is a shame not having the whole range, but those low notes are a real joy. And my home made monstrosities make proper low D whistles seem small and easy by comparison .
I did wonder if the removal of the edge of the upper lip was more of an ‘aiming’ device. It may ‘lift’ the average air flow and be similar in effect to lifting the windway floor (wrt the labium).
The first octave (A - G#) is a doddle. The next half octave (a - d) is reasonable, the next two notes (e, f#) are a fair bit harder
… similar symptoms/effects to those I’ve met on a couple of my “creations”.
In my experience a large bore/length ratio seems to limit the high notes available, those that are playable sound very “pure” (no high harmonics)
Having the “sloping bit” (labium?) too close to the windway encourages the whistle to jump into the second octave very readily, a side effect of this being relatively weak (low volume) low notes
Somewhere there’s a “perfect balance”, but, I suspect, only for certain types of music and playing, other styles will require different parameters … even more whistles
32mm seems big - how did you arrive at that id? trials, calculations, wild guess, had the pipe on hand?
too big a bore does make it hard to get the top of the second octave - but sometimes you have to work with the pipe you have, unless you want start with rod and drill/bore/ream.
32mm ID just happened to be the pipe available. My construction technique relies upon the plastic being amenable to ‘solvent-weld’ technology.
The upper notes are actually quite accessible - up to the third D - with the ‘correct’ head design. But with that head design the first octave is not playable . Stretching the playable range and centring it on the whistle is my challenge. A challenge I regularly fail at…
Don’t be too hard on yourself - it took the instrument makers of the past a few hundred years to arrive at the multiply tapered and choked bores of 18th century whistles (recorders) and flutes, and even then there were differences between the various designs to match the characteristics of various types of music, such as the English and German styles.
You could definitely do worse than look into the history of woodwind design. The cylindrical bore leaves you right at the beginning of the whole thing.
Ooooops, that was meant tongue-in-cheek, now you are giving me ideas.
So here is the question: “Does a tapered bore make it easier to achieve a full two octaves, or just to keep the whistle in tune across two octaves?”.
I always suspected that it was the latter, rather than the former, but I have been wrong many more times than I have been right. (Though in my defence being wrong initially is often the right thing to do eventually).
I don’t know the full answer to your question, but my understanding is that a tapered bore shortens the instrument while weakening the lowest notes - due to the smaller amplitude of the waveform at the end of the taper. Also, don’t take anything for granted: even the high baroque recorder, which has a thumbhole to enable the highest notes to speak at lower breath pressures - and which is definitely not in tune when simply overblown - has tuning issues. For instance, adjusting the aperture of the thumbhole allows the player to fine tune the top notes, which tend to be slightly flat. It isn’t simply a matter of opening or closing the hole.
I would guess that getting the full two octaves is more a matter of voicing… but that is a speculation.
Late baroque recorders, by makers like Stanesby, Bressan, Denner and so on, have a stepwise tapering bore with a pronounced choke at the end and sometimes a slight flare at the end of the foot joint (I think). The headjoint is cylindrical, so presumably its length is critical.
How far any of this applies to actual or possible whistles, I don’t know. The tapered bore whistles I’ve seen have a single taper throughout. The recorder was much more complex an instrument in its heyday. And there were no doubt good reasons for such complexity.
Do you mean what I call windway height - one component of the cross-sectional area of the windway?
I had a small play with windway height but my construction methods do not allow fine control. I have tried 2mm, 1.5mm and 1mm. 2mm took far too much air for my tastes, and 1mm far too little, so I have settled on 1.5mm. I may revisit the 1mm windway height as I have run out of 1.5mm plastic sheet.
By windway depth you may mean what I call windway length. If so, then I have only performed some small initial experiments. These seemed to sugest to me that windway length was not a major factor within the limits of my design, but I could always revisit this variable.
You might try the 2mm height again, this time tapering the windway from larger (width-wise) at the mouth to smaller at the exit. Also, lengthening the windway from mouth to exit can have some affect. Angling the windway to allow air to exit at a slightly different angle with respect to how it is aimed at the ramp can also affect some of your playing and tuning characteristics.
There is a tendency with many makers (amature and “professional” alike) to construct windways that are aimed straight at the ramp, and that have rather uniform characteristics. But a few makers out there (Mack Hoover comes immediately to mind) actually shape the windway from all four directions (side to side, and top to bottom) to focus the airstream differently - and this makes a world of difference in the way their whistles play and sound.
Guilty as charged But in my defense, it is a tad tricky to alter some of the characteristics on the windway, due to the tools available to the amateur whistle maker (plus, I am often ecstatic that my whistles work at all ). I am slowly learning though. I have learnt that moving the fipple block changes the characteristics of the whistle without needing to change the window. And I am experimenting with angling the ramps, but am still trying to find an easy way of doing this (all ideas gratefully received). Although I am in the process of making a square low D whistle (I had some square aluminum tubing) and that was easy to play with to get the right angles for the fipple block.
Also guilty as charged but I claim the same defence.
I am interested in the idea of a tapered windway. What does that buy me over a constant width windway of the same (minimum) width?
My construction techniques would be eased by a narrower labium - it would be easier to maintain constant slope/pitch/undercut, so I have several times considered reducing the window/windway/labium width. I have a struggle, though, getting enough air movement to get a good sound, when I increase the labium/window I improve the sound.
I would be very happy if this was because my sound-producing geometry is inefficient, but suspect it has something to do with increasing the area of the window - just like any other tone hole.
Also, surely the angle of the labium is defined by the angle presented to the air flow by its upper and lower faces. The thickness of the plastic by the time I come to carve the labium allows me quite a lot of leeway. Or are you thinking of the angle between the airflow and the axis of the whistle body?
Aaaaaaargh I should be playing… but this is fascinating stuff
Yes - you should be playing… (just kidding!) I’ve personally had better results with tapered windways, opposed to straight windways in whistles that I’ve made by hand. At the risk of being redundant, I’ll offer two theories why this might be so:
Hand-cut windways never achieve perfection - that is, their dimensions (height/width) vary along their length. With straight passages, there’s a significant chance that a choke (constriction/plane of minimum cross section) will be formed somewhere between the entrance and exit of the passage. If so, the highest air speed occurs at this choke and the air speed at the exit end of the windway is not well controlled by inlet pressure. This creates the dreaded “sqeakers/honkers” that are hard to push into the second octave. A tapered windway’s dimensions can vary a bit, but as long as the exit cross-section area is the smallest (from inlet to outlet), exit air speed is much more controllable by inlet pressure. Net result is a more controllable whistle.
By definition, the entrance cross-section area of taper windway > exit cross section. Thus a smaller change in air speed at the inlet would produce larger changes in air speed at the exit. Net result is that the whistle seems more “responsive”. This could easily be overdone, I suspect.
I’d be curious if anyone has experimented with a design that mimics the classic Generation head geometry, but with a bit of side-to-side and top-to-bottom taper. (Disclaimer - I don’t suffer from WhOA, so haven’t sampled a lot of different “designer” whistles - one of you makers may have already implemented this).
.
.
But in my defense, it is a tad tricky to alter some of the characteristics on the windway, due to the tools available to the amateur whistle maker (plus, I am often ecstatic that my whistles work at all 
). I am slowly learning though. I have learnt that moving the fipple block changes the characteristics of the whistle without needing to change the window. And I am experimenting with angling the ramps, but am still trying to find an easy way of doing this (all ideas gratefully received). Although I am in the process of making a square low D whistle (I had some square aluminum tubing) and that was easy to play with to get the right angles for the fipple block.
but I claim the same defence.