I could be wrong but as far as I’m aware the fipple-style NAFs are post-Contact, in other words can be no older than around 500 years.
The old flutes I’ve seen in museums are “rim blown” like the Bulgarian kaval. I play kaval and I can play a California Miwok flute, the technique is the same.
The Miwoks and some other tribes continue to make and play their ancient style of flute.
Here are some in museums Native America Flute Catalog
I think the arguments that fipple-style NAFs are post contact are very weak. And in fact, I think an equally strong/weak argument could be made that fipple style
flutes in Europe came from the Americas! Why, well because there are many examples of fipple flutes from the Americas that predate European contact. Maybe
not wooden NAFs, but there are lots of other fipple flutes from the Americas, some of which date back thousands of years and survived because they were constructed
from ceramics. But I think the reality is that humans all over the world discovered this way of producing sound thousands of years ago!
I’ve been interested in North American rim blown flutes for a long time. I have made some replicas from museum artifacts, and have even written up some
flutopedia entries for a few of them. I spent a lot of time learning to play these replica rim blown flutes in different styles, for example, blown obliquely (like a kaval),
upright (like a shakuhachi or quena), and using an interdental style, in which the rim is placed inside the mouth (kind of like a Persian ney). Interestingly, and
of particular relevance to this thread, I think, is that the interdental style which encloses the embouchure hole, produces a much more powerful sound. It is rougher
and somewhat more shocking to a western ear, but also much more emotionally evocative. If anyone is interested in this I have some sound samples of the same
ancestral Puebloan flutes played in different styles (outside the scope of this thread of course, so PM me).
I did some extensive research with Barry (White Crow) Higgins, who incidentally is also the guy who did the spectogram analysis of the warbling flutes, where we
studied some of the earliest photographs of Hoppi and other Puebloan Indians playing their rim blown flutes, and made close examinations of wear marks on 1000+ year old
ancestral puebloan flutes preserved in museums. We came to the conclusion that those rim blown flutes were certainly played using an interdental style.
But the point, as far as this thread is concerned, is that when you enclose the embouchure (or the window) on flutes or whistles, even partially, you can potentially
make such a profound change to the sound that it becomes almost a different instrument in terms of the musical context it can be used in.
Wow, Paddler, those Kwela players are totally on top of it, aren’t they. Now, with just a bit of messing around on the D whistle innocently snoozing beside my keyboard, poking it deeper into the gob (technical whistle term) and playing what they’re playing, I can immediately start to feel that increased power and control.
One of the immediate observations is pitch. In the bottom octave, with the beak just inside your mouth, if you blow harder, you jump to the second octave. In the second octave, you hear quite a rise in pitch before it breaks to the third partial. But with it deeper in the gob, you can prevent the rise in pitch, and are rewarded by an increase in power and edginess. And remembering that Kwela is street music, you’d need to come up with something to give the whistles enough presence outdoors. Imagine a New Orleans wedding parade played with baroque recorders. Nah!
Another observation was the dreaded warble which surfaced particularly on the low notes. But maybe this old narrow bore whistle isn’t the best starting place.
With power comes responsibility, and Kwela players would certainly need to be on the ball if they are all going to remain in tune. I wonder if in Irish whistle playing we’re ready for that challenge. The whistle is such a point & shoot instrument. But if we could harness some Kwela magic and tame it by locking it into hardware (the whistle) rather than limpware (the player), we might be on to something!
I think it remains to be seen whether it is possible to build the hardware to entirely harness some of these effects. I think we would need
to understand what is going on a bit better. I suspect that it is more than simply shading or enclosing the window. I think with some of these
techniques the player’s mouth/throat/etc actually becomes a kind of resonating chamber itself and it is the interplay between the resonance
here and that inside the instrument (or perhaps we should say … the rest of the instrument) that influences the sound. The player’s ability to
micro-adjust the shape of their mouth seems critical here.
You really feel this very obviously when you play a rim-blown flute using the interdental blowing approach. The following video demonstrates a
wide range of different blowing styles and their associated sounds. I think you can both hear the power of the interdental approaches, and see
the player dynamically adjusting the shape of his mouth to achieve the effect he wants. When you play in these interdental styles it is not only
louder, but you can actually feel the vibrations within your vocal tract.
I agree - we couldn’t expect the hardware to be able to replicate any of the dynamic effects for example. But if it can be shown that there is an ideal, or a minimum or maximum desirable window depth for a particular pitch of flute, perhaps influenced by other factors such as bore diameter, we should know that! And looking around at the enormous range of approaches, clearly we don’t, unless there’s something else going on at the same time! Now, possibly Tunborough’s model could cast some light on it, or we may have to try a few experiments. It wouldn’t be too hard to knock up a series of covers of increasing depth for example and see what we learn.
Woah, that kaval player certainly works the instrument hard. But he’s a lot closer to the jet than we are with the whistle. And I don’t think some of those effects would go down well in the session!
Now, here’s another matter that puzzles me, and I’ll snitch an image Richard Cook posted that illustrates it nicely. I’m guessing the upper whistle is an older Generation, and the lower whistle a later version. If you look at the ramp on the upper one, you can see it has two distinct slopes, whereas the lower one doesn’t appear to, or if it has, it’s less obvious.
And in other whistles we see a very short ramp, or a very long ramp. What do we think is going on there? How much ramp do we need? How does that relate to windway height?
You are correct, paddler; WIDesigner addresses tuning, and says little or nothing about volume or tone colour. I can tell you that a higher window wall tends to flatten most of the notes, the second octave more than the first, which is not the direction we want to go with for cylindrical whistles.
I recall reading that the wall around the window somehow helped direct the air stream and stabilize the oscillation, and was particularly important on low whistles, less so on high whistles. Can anybody correct or clarify this?
I can attest that we don’t need any ramp at all. I have an alto A whistle in PVC that has no ramp: the air stream hits a “vertical” wall at the far end of the windway. It plays just fine. I have a sense that the break between octaves is cleaner than usual for most of the notes, with less tendency to warble, although I have no hard evidence that this is true, or that it’s due to the 90 degree ramp angle. This experiment was inspired when I heard about the flauto dolce organ stop: http://organstops.org/f/FlautoDolce.html.
Presumably, a whistle with a tapered head bore, or inserts, could be constructed to correct this extra second octave flattening if there were desirable properties to be obtained from window walls. For example, strengthening the bell note, say.
This might be a direction worth exploring on low whistles.
Arghhh, Tunborough. Whistles with no ramp? What blasphemy is this? And on the holiest day of the year? You realise I will have to report you to the Morality police…
After all, we all know, don’t we, that in a whistle, you direct air down a windway or duct, until it is then “divided by a blade”? That’s common knowledge!
I have an alto A whistle in PVC that has no ramp: the air stream hits a “vertical” wall at the far end of the windway. It plays just fine. I have a sense that the break between octaves is cleaner than usual for most of the notes, with less tendency to warble, although I have no hard evidence that this is true, or that it’s due to the 90 degree ramp angle. This experiment was inspired when I heard about the flauto dolce organ stop: > http://organstops.org/f/FlautoDolce.html> .
Hmmm, that does seem to argue a bit against the common knowledge! And, something you could mention to your defense attorney, flutes have no blade. The sharpest thing a flute has is the angle at the far edge of the embouchure point, and it’s somewhere not much less than 90 degrees. And we Irish don’t aim at it anyway, we aim at the bottom of the hole. Don’t we know nuthing?
It certainly explains why we see such a variety of window and (dare I say it) ramp arrangements out there. This again seems to remind us we are afloat on a sea of unfounded assumptions, supported only on “everybody knows” life rafts. Which are full of holes! We need to make it part of our business to plug then when we find them!
The sharpest thing a flute has is the angle at the far edge of the embouchure point, and it’s somewhere not much less than 90 degrees. And we Irish don’t aim at it anyway, we aim at the bottom of the hole.
Aiming at the bottom of the hole reasonably changes where in the ribbon of air the splitting occurs, when compared with top aim. It does not follow from this that the splitting is done by anything other than the far edge of the embouchure in either case.
FWIW, as I have it out to measure, the ‘blade’ on the c.2007 Silkstone soprano D is a 45 degree (approx) cut across the tube, followed by 10mm of tube surface without a slope and then a steeper step of a mm or so. So geometry is something like the brass Setanta. That 10mm or so is acoustically significant as a ball of blu-tac in there quietens the whistle - not sure what it does to tuning as I haven’t done it since I have had RTTA.
Hi stringbed.
It’s hard to know for sure if the trigger for the switching action of the jet when blowing downwards is the top edge, or perhaps even hitting the wall is enough? And hard to know what proportion of the jet airflow switches over from “inside” to “outside”. We know it’s only a small proportion of the wave cycle - that’s what makes the cycle distinctly asymmetric, rather than sinusoidal, and that in turn leads to the edgy tone, the low proportion of fundamental, the high proportion of the partials, and the (subjective? real? who cares?) increase in power.
But the point I was trying to make was that, in the flute, there is no sharp “blade”. The flute “edge” is close to a 90 degree angle. If you blow “at the edge”, it’s probably presenting something like a +/- 45 degrees to the jet. But if you’re blowing down, it’s going to be more like “close to parallel to the wall” to “the blunt edge of the top”. Shall we guess at +30, -60 degrees? “Blunt edge” seems to be what Tunborough is talking about in the ramp-free whistle context.
Now, a few years back, there was a bit of a fad in the flute world about “cutaways”. These were a flattening of the normally curved surface beyond the embouchure hole, so as to present a much sharper edge at the edge. This was a step towards the notion of “dividing the jet”. The same notion that Tunborough seems to have leveled his 40 pounders against. Like I’m sure many flute makers, I tried it, at least twice. I have two heads still in my workshop drawer which attest to the fact it was not a success. At least in my way of blowing.
Now, here’s an interesting conundrum. When I look straight down the windway of the whistles I like best, I see the ramp. I don’t see down through the body of the whistle unless I really tilt the whistle to force that to happen. When I look down the windway of the whistles that are twitchy, I see some windway but I can also see down through the body. I’d be interested if others find the same, or find differently.
So, if I am right, the whistles that work well for me direct initial air up through the window to the outside world. Or, at least don’t send it down the bore. But the whistles that work badly for me split that airflow between bore and outside. Yet, when I “give the auld flute a blast”, I’m shooting the initial airflow way down into the bore, confidently expecting that the returning pressure wave will redirect it to the outside. So, on the face of that, the physics of flutes and whistles seem inverted. This doesn’t seem likely. Where have I come unstuck?
The physics of how an edge tone induces vibration in an air column are well understood. Marvelous photographs were published no later than the 1960s of the system of vortices at the business end of an organ flue pipe with transparent walls, and smoke rather than air blown into it. The same dynamics apply to the production of sound by blowing across the top of a soda bottle, so the splitter clearly does not need to have an acute angle.
As you note, the essential aerodynamic detail is that the stream of air along the outer side of the splitter lowers the pressure of the air inside the resonator until the differential becomes great enough to pull the “air reed” into the resonator. That then reverses the force and the reed is switched back to the outside of the instrument. If that action results in a regular and stable oscillation, by definition, we’ve got ourselves a flute.
I’ve looked down a prodigious number of recorder windways while taming this behavior. There is a fairly predictable correlation between visible structural detail and the resulting sound. However, it’s not hard to skew that mightily without being able to see what’s responsible for the change.
Before positing that the principles of sound production in transverse flutes differ fundamentally from those of duct flutes, it might be worth investigating how the conceptual model of “shooting the initial airflow way down into the bore” maps into the actual physics of the vibrating system. If one changes from blowing across the top of a coke bottle to blowing into the bottle, its timbre and pitch will also change — but the basic acoustic dynamic hasn’t been inverted.
When I point the beak of a whistle to the light, and stare up the bore from the bell end, what I want to see is a paper-thin sliver of light between the underside of the splitting blade and the bottom of the windway. If I don’t see any light at all, I expect the whistle won’t sound well, or at all. If I see a broad flood of light, I expect to hear a harsh tone from the whistle. I understand that’s one of the tweaks that Jerry Freeman does on his whistle heads, lowering the blade so it is just above the floor of the windway.
As far as I can tell, the precise shape and sharpness of the splitting edge doesn’t make much difference to a whistle’s ability to speak.
I’m not sure it even makes much difference to the quality of tone in terms of the mix of partials, but I do think that it contributes to
what is sometimes referred to as the non-musical aspects of the sound, for example, breathiness etc. Whether this is really non-musical
is another one of those questions that is subject to taste. For instruments, such as recorders used in a certain musical genre, it may be
considered a bug, whereas in others is may be a desirable feature.
The sight-based approaches to evaluating windway floor and blade alignment seem to rest on the assumption that the wind way floor
is flat, and that it aligns with the axis of the bore. I doubt that these assumptions are acoustically necessary, but they do generally hold
true for many whistle construction methods. I have wondered if a carefully constructed arc of the wind way floor might enable a quieter
second octave (more balanced volume between octaves), by deliberately introducing some inefficiency as the whistle is blown harder.
I think that might be a fruitful area for research. I hate playing whistles that are too loud in the upper second octave. So much so
that I pretty much avoid any of the wide bore whistles. To my mind, no amount of improved quality of tone in the bottom end can
compensate for an obnoxiously loud top end that is simply unmusical. 
Sorry, end of rant.
I think the reason it is bad to see a big air gap below splitting edge and above it (i.e., the blade located part way up the wind way height)
is because it makes it difficult for the pressure oscillations within the bore to fully push the air stream above or fully pull it below the edge.
Such whistles are inefficient, to the point of not working at all sometimes, because you end up with some air flowing in the wrong
direction during each oscillation phase, and hence weakening the amplitude of the oscillation.
It might be useful to bring chamfer into this discussion. This is one of the details in kkrell’s labeled drawing and is a major determinant of the air reed’s switching behavior. It also directly affects the balance between responsiveness at the lower and upper extremes of a duct flute’s range. Too little makes it difficult for the lowest notes to speak easily and too much has the same consequence for the highest notes. The chamfer on the block has a clear effect at both extremes and the chamfer on the roof of the windway is more relevant to upper range behavior.
I’m making these pronouncements on the basis of familiarity with recorder voicing but note that the whistles in my arsenal behave as would be expected on the same basis. The ones that are really powerful at the lower end and awkwardly screechy at the top, have the largest block chamfers and no upper chamfers at all. The whistles with the best balance between the extremes have more modest chamfers in both positions. Unsurprisingly, the mediocre performers don’t show any signs of their makers having ascribed particular significance to that design detail.
Heh heh, and what about the slight protrusion of the end of the block past the end of the windway? I imagine you’ve come across that in recorders, stringbed. Has it relevance in whistles (and if not, why not?), and does the degree of relevance vary with the pitch of the whistle, or any other factors? And how much protrusion, and why?
Again, comparing flutes and whistles, with flutes, the player has control over so many of the parameters. More, probably, than they could ever realise. In whistles, these parameters have to be coded into the hardware. Once you look down that road, you realise how many decisions have to be taken by the maker. And hopefully not regretted by the player…
what about the slight protrusion of the end of the block past the end of the windway? I imagine you’ve come across that in recorders…
The block determines the end of the floor of the windway so I’m assuming that you’re asking about the effect of it being offset from the end of the roof of the windway. But is the question about what happens when the block is shifted on an instrument that was voiced with the two ends in alignment, or about the utility of the offset as a voicing device?
In the former case, moving the block a smidge further into the instrument makes it harder for the upper notes to speak and renders the overall sound slightly more diffuse. I’ve never tried deliberately offsetting the ends of the windway as a design element and can’t comment on it. However, I’d be quite surprised to learn that tin whistles behave differently in this regard than recorders do — or the English flageolets that bridge the gap between them.
Again, comparing flutes and whistles, with flutes, the player has control over so many of the parameters.
For sure — and it makes the late-19th-century piccolo-flageolets all the more interesting. These give the player the choice of a flute or whistle mouthpiece on the same body. There’s more detail about them, including a demo recording and a general discussion of the proximity between small flutes and whistles at https://loopholes.blog/fife-flageolet/.
Have you come across any of these piccolo-flageolet combinations that worked well in both modes? I have a period one here which is a dismal failure in both. I reckon it’s proof only that charlatan makers existed back then as well as more recently. Probably significantly it has no maker’s mark (it was never going to attract return business!), but looks like the kind of instrument made popular by William Bainbridge.
I have restored and upgraded a period flageolet to a reasonable level of performance. It was by Wallis, from memory, so at least from a recognised maker. I had to graft on an extension as the D notes were annoyingly sharp. I couldn’t just retune the other notes to the Ds as some of those were tending sharp already, but I was able to tweak the others to all come together pretty well. The voicing wasn’t bad, but the overall performance was still not outstanding, probably due to the combination of small bore, small holes and thick body walls.
Because I haven’t found a piccolo / flageolet combination that works, I wonder even is it a practical idea? We know that the taper of the flute is needed to correct the tuning between the octaves. Would that same taper work for the flageolet mouthpiece as well, or would it overcompensate, driving the upper octave top notes sharp with the higher pressures one needs to get up there? Anyone had practical experience of this?