Does windway size affect tone?

Whistle
1

In my quest to make a satisfactory bass A whistle I am investigating windway size. By windway I mean the bit of the whistle that you blow in, right down to where it enters the fipple window.

I have two (whistle) heads of exactly the same design (within the limits of my skill) except that the height of the windway is 1.5mm in one and 1mm in the other.

The smaller windway has higher back pressure, as I wanted.

I was surprised to find a tonal difference between the heads. Is this normal for changing windway height? Or is it more likely that there some other (nu-noticed, unintended) difference? The smaller windway has a less pure tone, more raspy, but the raspiness increases as the hole pitch increases. I like it better than the other up to about halfway through the second octave.

Can anyone help? And any references to observations about the effect of whistle head dimensions on tone would be greatly appreciated.

TIA.

2

In my experience with playing various whistles (not making), the windway height makes a huge difference in tone and playing characteristics. The playing characteristics for short and high windways (respectively) seem to range from backpressure with little air needed to play to virtually no backpressure and needing more air to play. As for tone, the shorter windways (when set up correctly in terms of the relative position of the floor to the blade) will usually yield a more focused, crisp tone. They will typically be sweeter with less breathy or windy noises. The higher windways (thinking Waltons here) will be more breathy sounding, especially as air velocity increases in the second register.

I’ve given examples here of two extremes though I’m not sure how much of a difference you are going to see in a .5mm difference. I’ve played some whistles like Sindt that have a fairly short windway and played great. I’ve also played some hand crafted whistles at meetings that were, for my taste, too short and displayed some of the characteristics you detailed above. They also clogged really easily in an a/c’d room. I’m curious to know where your blade sits in relation to the floor of your windway. When you look down the windway from the tip of the beak, where does the blade sit? Of course, the window size (distance from the end of the windway to the blade edge) also plays a role in these things.

3

Some useful observations, thanks StrayCat82

Because of my construction method, the lowest part of the blade should be exactly in line with the windway floor (windway and blade are curved). I file the underside of the blade so that the sharp edge lies slightly above this level. Because of the limits of my skill- and tool-set the amount by which the edge of the blade is lifted is not as accurate as I would like, but I aim for somewhere about 1/4 of windway height. Because of construction methods it is relatively easy to get this line reasonably even, though the ‘taper’ under the windway is less easy to see and therefore is likely to be less accurate.

Yep, everything seems to affect everything… I have a window approx 15mm square, and a windway 15mm wide.

4

That is a huge distance from exit of windway to blade edge. On my low whistles it is less than half (about 7mm). I expect you have problems with the second octave high notes. Having such a long window may also put extra demands on windway accuracy and internal smoothness, and accuracy of the windway exit. A narrow windway (1mm height you say) may be more problematic here too, unless it is super smooth and straight.

Mind you I have not yet attempted to make a bass superlow A whistle.

5

Ok thanks for that observation Hans. Since I am willing to try any variation, I will make a head with a window 15mm wide, 7 mm long.

For some reason, though I tried shorter windows, I settled on 15mm x 15mm. That may have been influenced by my early 2mm high windway, or my poor body design. It seems to be an iterative process - improvements to the head require a change in the body.

Mind you this is still my first whistle. It has three heads and six bodies, but it is still only one whistle…

6

A while back we had this topic:
A fipple physics question
Some discussion in it about the block and it’s bevel may be interesting.

7

That is interesting, thanks Hans. I feel somewhat outclassed by discussions where measurements are taken to tenths of a millimetre :blush: .

8

I am afraid a tenth of a millimetre does count in whistle construction. :wink:

9

I foresee calipers and micrometers in your future and something spinning. Is that a lathe?

10

Steady hands and a new pair of glasses first.

11

In my own observations from whistle making, it’s not so much the tenths of a millimeter that are important, it’s the hundredths! It is the tolerances between components that matter in the final performance of a whistle once you have settled roughly on the dimensions you want to use. For someone doing one-offs and small runs, the right tolerances are hard to achieve consistantly. Very often they can be better felt rather than measured and that is where handbuilding large runs (50 or so at a time) produces better and more consistant results for me. I can cross test all the components to get the right fit, looking after the hundredths, so to speak.

12

I am not going to get to hundredths of a mil. Never mind, I am getting some decent-ish sounds from my whistle. Although a lot of the enthusiasm can be ascribed to the novelty of a whistle made of waste-pipe, I did get asked twice for another tune at the session last night. Ok it was by the locals rather than the sessionistas, but it still says I am heading in the right direction. Or the right direction by my standards, which is to make music that people enjoy.

The component-fit issue is solved in two ways for me. Plastic plumbing couplers fit very snuggley, with smoothed bends, and the parts I assemble into a ‘head’ are glued together with a glue that fills gaps. The accurate windway height comes from the thickness of sheet polystyrene. The rest is down to hand-carving techniques.

I am never going to produce a saleable whistle, but I am convinced I can make a playable one. Actually, I am only going to make one whistle, no matter how many heads and bodies it takes…

13

Little late getting to this thread - but in a word…

Yes

Along with quite a few other things - many of which interact with each other and make it very hard to tease out generalities or hard and fast rules. I basically agree with all that has been said in this thread (for what that is worth - it agrees with many dozens of prototype whistles I have made). And frequently hundredths of a mm do matter which is not to say that it can not be done by hand without fancy tools - just might be hard to reproduce.

14

Thanks again for that comment about the window length. I have made a head with a shorter window (~8mm from the fipple block). It sounds good right up to the top of the second octave (though wildly sharp at oooooo - needs oooxxx). Third A is achievable, surprisingly not too sharp.

The whistle is loud enough - I made a new body and tuned it for blowing quite hard. The only downside is that the low end of the low octave is a bit weaker.

The xxxooo is in tune but a bit ‘weak’. Ah well, the shop has plenty of pipe to sell me.

15

What is the inner bore diameter of your low A?

16

32mm

17

That’s a big bore and should give you strong bottom notes.

18

Perhaps hole 4 is a little too small?
Did you smooth all your hole edges?

19

I guess it depends upon what I compare with. The longer window gave really strong loud bottom end, but the top notes were unreliable, and needed a lot of push. Using the shorter window has made the top notes easier and more reliable, but the lower notes are less strong than they were. I think they are still loud enough, and the tone has improved greatly. It is more focused(?) whereas before I would have described it as soft and woolly.

Also the comparison is not like-with-like as I changed the design slightly. The edge of the blade is in line with the bottom of the wind way. That means I dont have to try to make an accurate chamfer underneath, where I can not see it.

Yep, I reckon hole 4 is a little small. I already moved it down a bit, and next time I will move it down a bit more. Fortunately it is a thumb-hole, so it is not going to interfere with the fifth hole. TWCalc and TWJCalc fail me for the large holes. I am not sure if it is the algorithm or the fact my whistle has a bend in it, but the larger holes wind up smaller than either program calculates. Also the large holes allow the finger pads to intrude further into the air column, and this affects the tuning of holes lower down. I could double the wall thickness (to 4mm) quite easily. Would this be beneficial?

I am having a bit of trouble with deciding how to tune the whistles - if I tune them for softer/normal playing, then when I push them, say in a larger room, they will go sharp. But if I tune them for harder blowing they sound a little flat played normally. I guess that where a good design would have a ‘plateau’.

As for smoothing the holes, well I am doing a bit, but my hole cutting is primitive. I need to get a drill press, and some drill bits above 12 mm (crimbo is coming). Currently I have to ‘counter bore’ large holes and trim them out with a knife. Some of the holes end up a little on the wonk, and sanding them smooth might be a little pointless. Even normal drill holes go a bit triangular when the bit breaks through. I think that I have a way to go yet tools and technique-wise…

I guess that the surprise is that I have a playable instrument in a couple of weeks, starting with no knowledge or special tools, though I might be benefiting from the ‘singing pig’ syndrome at present. No one notices how badly the pig sings, they are amazed it sings at all.

I am grateful as always that you proffer advice.

20

I have not tested TWJCalc. But here is a comparison for a low A of results from TWJCalc and my own, based on Pete Kosels:

Pipe ID 32mm, wall thickness 2mm, ET intonation
All holes 12mm for simplicity, 
calculations rounded to one tenth
     TWJC     HB-C 
1:  396.5    397.0
2:  346.1    345.4
3:  293.8    295.7
4:  214.5    207.0
5:  194.5    194.5
6:  118.3    116.5

Some holes fit well, but hole 4 shows the most difference. So when hole 4 ends up smaller than designed, in order to be in tune, then the calculation did put it too high.

I don’t think that large holes are an issue regards to fingers protruding into the tube. This happens also on smaller bore whistles, and I have not yet seen that this throws the tuning off.

I prefer thinner tubing for a more responsive whistle. So I would not go for 4mm wall thickness.

The bend should not matter. But TWJCalc shows 774mm length, my own shows 714mm, with a window 8 by 16mm. That is a huge difference. And resizing the window in TWJCalc does not change the result, but it does in practice.