3d-printed whistle head: strengthening lower octave

[Tunborough]

Gah, sorry, I made a typo - I meant to say I'm using A, not B (I haven't fundamentally redesigned anything since I made my original post - that head is still current).
...
I feel like I must be doing something wrong or otherwise missing something -- everyone is telling me I'm using a reverse taper, and even the sample data file with WIDesigner with "a Fajardo-style wedge under the window to improve octave balance" has the first bore point smaller -- but I can't see it ...

We may need to look at your choice of optimizers, constraints, and starting profiles. For bore optimization in particular, the default constraints often need some adjustment. In your last two samples, the second bore point is less than 1 cm below the top of the bore, which isn't very much room for a useful profile. If you attach the instrument file to your query on the WIDesigner site, https://github.com/edwardkort/WWIDesigner/issues/147, I can take it out for a spin and see what I get.

[Nazrax]

You could do what makers do normally to adjust the space in the windway. Burkes, Reyburns, MK, Etc, all take up space by having the block on the inside continuing on the sides, and go forward as much as you need to, to adjust the tuning how you want. Taking up more space by having it longer makes the upper second octave more sharp. Having a little cave in the block or mouthpiece like in gens, feadogs, etc, will have an extra flat upper second octave. No spacer or cave will be in the middle.

Would you consider what I'm doing the same as the "cave" (just on the extreme end)? Or is it something different?

If you attach the instrument file to your query on the WIDesigner site, https://github.com/edwardkort/WWIDesigner/issues/147, I can take it out for a spin and see what I get.

Attached - thanks!

[Narzog]

Would you consider what I'm doing the same as the "cave" (just on the extreme end)? Or is it something different?

Are your A B and C pictures showing the empty space, or are they the whistle shape? Not 100% sure what I'm looking at. But if you are increasing the empty space inside the mouthpiece, this is like the cave example. Most makers do the opposite, which makes the second octave upper notes easier to play in tune for most people.

Also in a few posts you mentioned conical bores, but I think that you are using a cylindrical tube. To my knowledge most of the benefits of having a conical bore, comes form the bore, not the headpiece. The inside of the mouthpiece doesnt need to be conical, unless that gives some benefits I don't know about. But I haven't dealt with conical whistles enough to have a valid opinion of them. But unless I'm mistaken you can get very good results with a normal cylindrical inner mouthpiece, and tweak it from there.

[Nazrax]

Taking up more space ... makes the upper second octave more sharp. Having a little cave in the block or mouthpiece like in gens, feadogs, etc, will have an extra flat upper second octave. No spacer or cave will be in the middle.

As I thought more about this statement, I realized that it makes a lot of "intuitive" sense (bigger air column means lower notes, etc) and also explains why my "cave" works when it "shouldn't": my older fully-cylindrical designs had a sharp second octave (which I'm gathering is actually unusual). Given whatever weird thing I'm doing that ends up with extra-sharp second octaves, extra space in the head to flatten them sounds perfect.

Are your A B and C pictures showing the empty space, or are they the whistle shape? Not 100% sure what I'm looking at. But if you are increasing the empty space inside the mouthpiece, this is like the cave example. Most makers do the opposite, which makes the second octave upper notes easier to play in tune for most people.

Those are really rough exterior views of "hypothetical" heads. For an accurate view of the head I'm actually working with, see the first post.

Also in a few posts you mentioned conical bores, but I think that you are using a cylindrical tube. To my knowledge most of the benefits of having a conical bore, comes form the bore, not the headpiece. The inside of the mouthpiece doesnt need to be conical, unless that gives some benefits I don't know about. But I haven't dealt with conical whistles enough to have a valid opinion of them. But unless I'm mistaken you can get very good results with a normal cylindrical inner mouthpiece, and tweak it from there.

So, many commercially-available whistles seem to use a straight bore head and a conical bore body (just based on external pictures I've found - I haven't run across any pictures of what the insides look like). Since I don't have control over the body, I'm instead using a straight bore body and an appropriately tweaked head (conical, parabolic, fractal (j/k - I hope), whatever I need) to get things in tune. Some what I've read indicate that the conicalness doesn't have to be full-body, that it can be "compressed" into one end and have much the same effect (hence the effectiveness of things like the Fajardo wedge), and that's what I'm aiming for with my heads.

[Tunborough]

my older fully-cylindrical designs had a sharp second octave (which I'm gathering is actually unusual). Given whatever weird thing I'm doing that ends up with extra-sharp second octaves, extra space in the head to flatten them sounds perfect.

Based on some experiments, the small window you're using makes the difference. Your window is about 12 sq mm. With a small window, you can see a flat first octave and a sharp second octave, even with a cylindrical headjoint and cylindrical body. With a more conventionally-sized window, a cylindrical body will lead to a flat second octave unless you include some narrowing in the headjoint. For comparison, a Feadog has a window more than 40 sq mm.

Another observation ... With walls around the window, you'll want to make the bottom hole unusually large, to balance the Ds and Es, especially in the second octave. That large hole isn't necessary with a lower window height.

[Nazrax]

So, I went back into WI Designer and added an extra bore point, so now I have the block at -4, the labium at 0, the bottom of the taper, and the end of the whistle. The new design might look more like what everyone's been expecting:



I've also found that it's possible to strengthen the lower octave too much: with designs with higher walls, my blowing pressure has to jump significantly to actually reach the second octave. Looks like what I really wanted was to make it harder to underblow the lower notes. My favorite attempt so far has a low wall and a small block extension, but testing continues.

[Nazrax]

So, my attempt to make the above head went really wrong, if fascinatingly so. I did my usual process of cutting the pipe a little long, doing a bell note calibration, calculating hole spacing/sizing, drilling holes with bits smaller than predicted, and then tuning the whole thing. This whistle is incredibly resistant to sharpening! I had to cut the pipe a good bit shorter than predicted, and the holes ended up 3-4mm larger than predicted. Additionally:

• The OXXOOO C-natural fingering only flattens C# by about 15 cents
  The OXXXXX D6 fingering actually plays a sharp C-natural
  XXXXXX plays a pretty decent D6
  If I play a D5 and gradually increase pressure I hit a weird "dead" spot that doesn't make any sound at all; as I keep blowing harder I get a D6

Here's a comparison of a whistle using the head design from my top post (on top) vs the whistle with the latest design - they're both roughly in tune, but look at the hole sizes on the new one!


It looks like momentum on this thread has died out, so I won't post any more updates (unless there's renewed interest from someone else).

[RoberTunes]

Wow, I had no idea that the width of the windway, window and blade was so incredibly small, roughly 1/4 the width of the bore. I can just imagine how difficult it would be for the instrument to produce good tone from the air going into that eye of a needle. I had previously thought that if a blade/window area was narrowed, it would still be at least 80% the width of the bore, and every aspect of the entire mouthpiece area would have to be perfectly selected, for it to function. The reverse idea is when, to gain loudness, tonal range and expressive options, designs flatten the top of the bore at the mouthpiece area, to make the blade (and therefore windway too) as wide as the 100% of the inner bore.
I'd immediately try widening that window to at least 80% the width of the bore, and see what happens.
Perhaps quietening a whistle should not be based mostly on mouthpiece design, but on other factors.

[Nazrax]

The main factors I'm aware of which affect volume are window width and tone hole size - and thanks to cutoff frequencies there's a physical limitation as to how small those can get and still work. Is there another factor I'm missing?

I'm actually pretty satisfied with my 3mm windows - but I want to see just how far I can go and have an instrument that still "feels" like a tin whistle (e.g. not that crazy thing I just posted about!)

[Narzog]

Those look quite nice, and pretty unique, whic hsi also nice. One of my biggest thigns for wanting to maek a 3d printable / injection molded head is uniqueness. Making a whistle out of tubes like everyone else kinda always looks the same.
Print quality is also nice. Did you have to use internal supports? Those really killed me when I was trying to fdm print heads.

Are you buying wooden tubes? The easiest way to test whistles would be to use PVC because its insanely cheap and easy to drill. You can even enlarge the tone holes by twisting some scissors in the hole.

That's really interesting how the new one has massive holes. Not what I would have guessed. and when your goal is to make it as quiet as possible seems very problematic haha.

You should keep posting updates. The forum just doesnt get a whole lot of use these days and most people aren't whistle makers so amount of people posting will be limited.

[Nazrax]

Those look quite nice, and pretty unique, whic hsi also nice. One of my biggest thigns for wanting to maek a 3d printable / injection molded head is uniqueness. Making a whistle out of tubes like everyone else kinda always looks the same.
Print quality is also nice. Did you have to use internal supports? Those really killed me when I was trying to fdm print heads.

Heh, thanks. I print it with the "sleeve" on the bed and the windway pointing up. The only supports are for the ring where the ID makes a sharp change where the sleeve meets the bore (10mm in). Everything else higher up that isn't supported is straight across, so it's able to bridge with no issues. I print with a .16mm layer height (I tried .12, but something weird happened with the blade; and .16 seems good enough and is a little faster).

Are you buying wooden tubes? The easiest way to test whistles would be to use PVC because its insanely cheap and easy to drill. You can even enlarge the tone holes by twisting some scissors in the hole.

I mostly use CPVC, though I've tried PVC when the bore is closer to what I'm looking for - and its extra thickness doesn't really hurt sound quality since I'm not trying to make the head out of it. I usually sand it down and round out the holes a bit to make it nicer to hold. I've never tried any other material.

That's really interesting how the new one has massive holes. Not what I would have guessed. and when your goal is to make it as quiet as possible seems very problematic haha.

Yeah, working with that whistle was really weird. I ended up skipping 2 or 3 bit sizes at once - which I never do - trying to get the thing in tune! But, it did teach me that the internal design of the head - even if it's only a couple of centimeters - can make a huge difference not only to the sound quality but also to the overall tuning of the instrument.

You should keep posting updates. The forum just doesnt get a whole lot of use these days and most people aren't whistle makers so amount of people posting will be limited.

Thanks for the encouragement

[Narzog]

I have tried printing windway up but would have lots of issue with sagging on bridges. Your printer and setting seem very good to not need supports. Being able to print no supports is a huge plus, supports suck haha.

For some reason in the picture it looks like wood haha. You can buy black pvc which looks extra nice.

[Huhupat]

The finger holes line up only because you used the same pattern at different starting points. If you line these up window to window, or end to end there is a better context over hole placement vs size. Are your measurements linked to a spreadsheet yet?

[Nazrax]

The finger holes line up only because you used the same pattern at different starting points. If you line these up window to window, or end to end there is a better context over hole placement vs size. Are your measurements linked to a spreadsheet yet?

I tried to line up the windows (give or take a couple of mm), but my point to show the radical difference the weird head design caused; I wasn't trying to line up the holes at all (the hole locations were calculated by WI Designer).

I'm not sure what you're asking about regarding a spreadsheet.

[Huhupat]

The reason you need big holes is because the hole pattern is too far toward the bell end.