Wood Whistle Makers

[Kedster]

On the note of "Wooden" flutes, I've just heard that my Labu Bamboo Celtic Set is ready to dispatch finally, after about a month or so of building.

The communication was a bit stuttered, but they seem genuinely nice and try to be hopeful, and being able to purchase about 12 whistles for all sorts of weird keys I'd like to try out at a whim, all at $130, was an O-kay deal for me. There's good enough reviews on the YouTube and they seem to be real instruments built to two full octaves, so let's see what they're all about.

The only gripe I've seen across YouTubers is that they're a bit flat - I think they're tuned to the hot climate of bangladesh and so inevitably fall flat at 21*C room temperature. Which is fine for me, since I usually like to play at 432 - but they might be harder to fit in a traditional session. Another thing is, the collection seems to have 2 F# whistles and no D# or E whistles. That's a bit odd, but maybe it'll be the full 12 tones. Let's see what they're like.

[JackL]

PM sent

For those keeping score, I have provided David with details regarding the tuning I hear and measure with my Rover. I am sending the instrument back to him to see if he can adjust it.

[Cyberknight]

And we have to keep in mind that there are tapered bores and TAPERED BORES. When I was trying to get my head around this, I came up with some categorisation of the amount of tapering by dividing them into:

Cylinder (ie no taper)
Mild Taper, and Stepped Taper (eg three sections of telescoping thin-walled tubing)
Bold Taper & Flare
-----
Bold Taper, No Flare
Full length Bold Taper

My feeling was the ones below the dotted line above tended to suffer from weak bottom notes.

Still my favourite among my clutch of tweaked 1970s whistles and the handful I have concocted myself is my pic-whistle, an unholy combination of a Pink Generation Bb head (big bore!), a wooden body from my 12.5mm piccolo design, tapered with flare, and a delrin barrel and brass slide to join it all together. So it falls into the middle category, Bold Taper & Flare. The bottom D is pretty sturdy, and can be pushed into quite a hard sound if desired. The flare tunes the xxx xxx partials better (as it does in the flute), improving stability and tone.

The bottom category above, Full length Bold Taper, includes whistles like the Clarkes. I find them too weak to enjoy. But then I find small bore cylindricals the same. So you have to allow for a pretty wide variation in personal taste here.

Sorry if I misrepresented your position! I appreciate the nuanced clarification.

One thing I'm curious about is flares, because I've never tried a whistle with one. Have you done experiments confirming that flares at the end make the lowest notes louder? If they do, I wonder why they aren't used more commonly.

[Terry McGee]

No, no misrepresentation implied, just reminding us that it's not as simple as say cylindrical vs tapered.

I haven't done anything on flares involving sound level measurement - it rapidly becomes a bit of a nightmare as you'd really need to set up "the blowing machine", find the right flow rates for the whistle concerned, measure all the pitches and levels of all the notes, change or eliminate the flare, measure again, and end up dealing with an awful lot of data, possibly concluding that "it all depends". For example, as I said above in regard to my flared pic-whistle, "The bottom D is pretty sturdy, and can be pushed into quite a hard sound if desired." And whereas, I probably should desire the hard sound, can I be bothered changing my blowing approach to bring it out? Probably not!

The main reason for using terminal flares is tuning rather than volume. If my pic whistle bore continued to taper to the end, the low D would end up distinctly flat compared to the middle D. And probably a bit waffly, as the partials wouldn't be at their ideal frequencies. This usually means that it will overblow easily. The degree of taper required to tune the octaves of all the other fingerings is too great when we come to the perfectly vented end of the tube, so a flare compensates for that. This flare goes back all the way to when the conical tapered flutes were first developed, around Hotteterre's time.

Now, I should add that the flare is essential if you have a cylindrical upper section of bore followed by a bold taper. I don't know if the need applies to mildly tapered whistles or full length bold tapers like the Clarke. And would a mild taper need a mild flare?

I guess any whistle that has the bottom note flatter than its octave is in the market for a flare?

[paddler]

Using a flare in the foot of the bore makes a lot of sense. I'm not sure why it is so rare to see it in whistles. Probably because there are so few conical bore whistles in the first place, combined with constraints imposed by the manufacturing process of those that do exist.

If you look at conical bore flutes that have a short foot (i.e. historically, pretty much any flute that has between 1 and 6 keys) you see that they all have some amount of flare at the foot, produced by back-reaming from the bottom end of the foot. Often this is done using one of the same reamers that were used to create the conical bore. You can tell this by examining the profile of the bore and matching gradients.

Interestingly, flutes with a long foot do not tend to have any flare in the foot, the reason being that the tone holes in the foot (with their open standing keys) seem to provide enough of a frequency filter to help align the harmonics. Although, I have to say that I do notice a difference in the quality of tone between short and long foot flutes. It is subtle, and I won't argue for which is better. With the right flare at the foot, the bell note of a short foot flute can be very powerful. With the wrong flare, or no flare, it does tend to be much less satisfying.

So, I think you are onto something here Terry. I think it makes a lot of sense to experiment with a bit of foot flare for a conical bore whistle. I think I'd start out by using about the same amount of back reaming as you see on a short foot D flute (assuming we're interested in low D whistles here).

[Terry McGee]

Yeah, as a rough guide, flutes ending in the note D tend to have the flare starting about the start of the foot, though I have seen some that extend a little further back into the end of the RH section, requiring a bit of back reaming there too. And, as you say, usually the same taper as the body. But that could be just laziness!

I have seen some back-reaming in flutes ending in C, but it tends to be really minor. I guess you don't want it extending back before the C# key hole, as that hole still needs the corrective effect of the body taper.

For example, on my Prattens Plans, I see that the C# hole is at 91mm from the start of the foot. The bore is still reducing in diameter until it gets to 12.4mm at 100mm, then expands to 12.9mm at the end of the foot, 132.5mm from the start of the foot. So a flare of 0.5mm over the last 32.5mm. Not a big deal, compared to what we see in the Firth & Pond which launched the Grey Larsen Preferred, which expands about 2mm over the last 80 or so mm:



Note that it's flaring to about the same bore diameter as we'd find around the first hole in the RH section! Just checked, and that's where I arrived at with my pic-whistle bore. So it certainly seems to have relevance to bold taper whistles with a cylindrical upper body. Whether it has possible wider application still to be determined!

When the C foot first came out, there was a lot of moaning and groaning in the musical press about how it destroyed the bottom D. My guess is that the unimpressed were missing the boost and focus that the flare gives to the end note. I love it!

[paddler]

Most of the flutes I've profiled have flare that only extends to a point within the foot section. In other words, the minimum bore diameter is located within the foot section at a point that is somewhat closer to the socket than to the bottom end of the foot. I have no idea what is optimum, but I have quite a lot of flutes that perform very well on their bell note with this bore profile.

I'd be a bit suspicious of a flute whose minimum bore diameter was at the bottom end of the right hand section, near the tenon, for reasons of possible tenon compression. But I suppose if the entire foot section has a reverse taper that must indicate that it was reamed from the bottom end entirely. I haven't seen one like this though.

[Kedster]

No, no misrepresentation implied, just reminding us that it's not as simple as say cylindrical vs tapered.

I haven't done anything on flares involving sound level measurement - it rapidly becomes a bit of a nightmare as you'd really need to set up "the blowing machine", find the right flow rates for the whistle concerned, measure all the pitches and levels of all the notes, change or eliminate the flare, measure again, and end up dealing with an awful lot of data, possibly concluding that "it all depends". For example, as I said above in regard to my flared pic-whistle, "The bottom D is pretty sturdy, and can be pushed into quite a hard sound if desired." And whereas, I probably should desire the hard sound, can I be bothered changing my blowing approach to bring it out? Probably not!

The main reason for using terminal flares is tuning rather than volume. If my pic whistle bore continued to taper to the end, the low D would end up distinctly flat compared to the middle D. And probably a bit waffly, as the partials wouldn't be at their ideal frequencies. This usually means that it will overblow easily. The degree of taper required to tune the octaves of all the other fingerings is too great when we come to the perfectly vented end of the tube, so a flare compensates for that. This flare goes back all the way to when the conical tapered flutes were first developed, around Hotteterre's time.

Now, I should add that the flare is essential if you have a cylindrical upper section of bore followed by a bold taper. I don't know if the need applies to mildly tapered whistles or full length bold tapers like the Clarke. And would a mild taper need a mild flare?

I guess any whistle that has the bottom note flatter than its octave is in the market for a flare?

On a tangential note, I kinda of want to build a machine learning ("AI") program that would make it easy to model how various shapes, materials, thicknesses, bores, flares, holes and all effect the pitch.

The coding is the easy part, but I have a very rudimentary knowledge of acoustics, worse yet, I really don't know where to begin teaching the program. But any guidance on that would be greatly appreciated.

[Terry McGee]

Google "woodwind instrument designer". You'll find a Java application and YouTube videos on how to make it work.

WID, AI and 3D printing, just introduce them all to each other and get out of the way....

[Terry McGee]

Most of the flutes I've profiled have flare that only extends to a point within the foot section. In other words, the minimum bore diameter is located within the foot section at a point that is somewhat closer to the socket than to the bottom end of the foot. I have no idea what is optimum, but I have quite a lot of flutes that perform very well on their bell note with this bore profile.

Yeah, I think that's more common in my experience too. This flute wanted to be a clarinet.

I'd be a bit suspicious of a flute whose minimum bore diameter was at the bottom end of the right hand section, near the tenon, for reasons of possible tenon compression. But I suppose if the entire foot section has a reverse taper that must indicate that it was reamed from the bottom end entirely. I haven't seen one like this though.

Agreed about the need for caution, and I reckon we can see the telltale signs of tenon compression in the Firth and Pond flute I showed. The Firth and Pond (Grey's flute) is the bold trace in the graph below. Look at the right hand end of the LH section (bold, navy) - I reckon that is clearly compressed. The right hand end of the RH section (bold, pink) is a little less clear cut. It has the same squished look as the LH one, but if you hold a rule up to the screen, it does appear to be pretty symmetrical around a minimum at say 315mm.

Note that the reamers I hurriedly used to make the "test copy" of Grey's flute before he left town (shown on the same graph but with thin traces) also has its minimum near the end of the RH section. Those reamers were for a French five key by Noe Freres. I often wonder if the early US makes might have taken some inspiration from French makers. We'd need to compare with English, French and German short-footed flutes to see if we could pin that down.



I should mention for those who are feeling a bit mystified by all this that the article about my collaboration with Grey Larsen where this graph comes from is at https://www.mcgee-flutes.com/Grey.htm

[Kedster]

and get out of the way....

That's the plans...

[Tunborough]

On a tangential note, I kinda of want to build a machine learning ("AI") program that would make it easy to model how various shapes, materials, thicknesses, bores, flares, holes and all effect the pitch.

Google "woodwind instrument designer". You'll find a Java application and YouTube videos on how to make it work.

https://github.com/edwardkort/WWIDesign ... WIDesigner

The coding is the easy part, but I have a very rudimentary knowledge of acoustics, worse yet, I really don't know where to begin teaching the program. But any guidance on that would be greatly appreciated.

Pitch calculation is an acoustics problem, suited for mathematical models, not machine learning models at all. A good place to start are the acoustics reference documents in https://github.com/edwardkort/WWIDesign ... bliography. Online, there is http://www.phys.unsw.edu.au/music/. (Be warned: this is a deep and fascinating rabbit hole.)