Is there a program equivalent to TWCalc that calculates hole placements, hole diameters and taper for conical bore whistles?
Thanks,
Mike
No, but I’ve been thinking of making a mouthpiece for a sweetonebarrel. Has anyone tried that?
Rob
I have been working on one on-and-off for about a year. Anyone want to help?
Do you mean on a design program, or on a mouthpiece for a Sweetone? I’d like to help on a program. I started doing some research on the web yesterday, but then work got in the way 
. From what I saw (and I admit I didn’t have time to dig into it), you need to use the bore diameter at each hole center in the iterations. It might be that just adding a taper parameter to TWCalc and a couple of lines of code to calculate the local bore diameter as the holes move would be all you’d need. Is source code for TWCalc available?
I made a Whistle using conical Elder and it ended up having pretty much the same distribution of hole positions and sizes as a cylindrical except the hole pattern was shifted toward the fipple. This might be a one off, but maybe a simple linear term could be added to the cylindrical case to correct it for conical bore.
John S
I have been working on a program.
You are right AFAIK, that each hole has to have its own bore diameter. The software that I have written is menu driver in text mode, so it should be absolutely portable between platforms. The last time I worked on it was August 2004. The UI is klunky, but it allows a whistle to be saved and retrieved, modified and re-saved. It also stores all parameters for each hole separately. I was originaly intending to use it to test some ideas about optimization that I had, but I never made it that far. Work takes over my brain like a body snatcher and play is ARRGBNNDSgssvd,
Anyhow, you (and anyone else) are welcome to this source code. The makefile requires a unix-like environment (Cygwin under the windyhosejob works well) However, the source should really compile with any C compiler. (Borland springs to mind). Send me an email and I will attach it to a reply. You can redistribute it however you like or use it in a commercial application. This release of the source code is PD. You are hereby advised that you are getting what you are paying for. <>
TWcalc was written by Daniel Bingamon. You can go to his web site to contact him. It is http://www.tinwhistles.us/
In theory at least, the distribution of holes would have to be non-linear. If you are tuning each hole separately, then you would not notice the difference because we are talking about increments of less than .05". In practice, it may well be that a simple value added to each hole position would work. The question is how do you get that incremental value?
I have considered putting tapered pipe capability in TWCalc. I suppose anyone who would make a taper would be be from top to bottom. Then there is the tuning joint, which is not tapered. So, The question is how to convey the taper and straight section information. Would you put in the beginning and ending tubing diameters or a starting diameter and the number of degrees of taper.
Another possiblity is to a have sample measurements to profile the bore every so many millimeters.
This requires a lot of thought to make it work with various whistles.
BTW - Source code to TWCalc is available to anyone who wants to work on it.
What I had originally intended was to specify a length and the beginning and ending diameter. You could then calculate the bore size at a given point fairly easy. Also, I think those are the numbers I would actually use.
Do you think it is reasonable to assume that if part of the bore is not tapered, it would not have much effect on the tuning of the rest of the instrument?
I don’t know how people have approached this with whistles, but I took a look at several recorders (3 sopranos and a tenor), and the bore taper goes all the way up. The joint is not tapered, but the inner bore is. So maybe this becomes a non-issue.
In any case, it appears that the local bore diameter is a relevant paramter only in the neigborhood of hole, in that it affects f_c for that hole. I’m assuming the joint would not be in the neighborhood of a hole, so maybe it doesn’t matter. It just means that maybe you’d need to specify bore diameter along the length more specifically than just giving the two ends and the length. This might be a cool thing to do anyway, so you could use this to design bores with discontinuities in them, like telescoping whistles or David Daye’s “infra-whistle” made from three different diameters of tubing.
- Mike
AFAIK, there is no software that supports a tapered bore. (yet )
I have another question along these lines. Has anyone ever tried to predict cross fingerings using software? (silly question perhaps) I am wondering if anyone has tried using different combinations of open and closed hole corrections to try and predict what cross fingerings will actually work. Does that seem plausable?
Hi Ctilbury, as you correctly surmised I do tune holes individually, the reason I thought it relevant, is that all the holes ended up more than 1/2" nearer the fipple.
The Elder went from at the fipple 12.5mm down to 11.5 at the end, but this is the natural taper of the material so I don’t know how linear it is.
TTFN
John S
Just as an aside…but I cannot get TWCalc to function on my WinXP machine. Something about a 16 bit Windows subsystem. Is this a TWCalc error, and is there any way around it? Has it been written as a 16 bit program for older DOS based machinery?
I’m busting to try it, cause I found some of the hole placements on the http://science.univr.it/~gonzato/whistle/ site a little off for the A whistle.
but I cannot get TWCalc to function on my WinXP machine
TWCalc is compiled under the Win 16-Bit API. But so is Regedit, command.com and large number of other windows functions.
There is some spyware (malware) or virus that disable 16-bit programs. One of these programs is called ‘cool.exe’. If you try to run REGEDIT and it keeps shutting down, the computer is infected. Deja Vu, this is the second time I’ve heard this problem and last time it was this virus or spyware.
I’ve run TWCalc on Win 3.11, Win 98, Win ME, Windows XP and even under Linux using WINE Windows Emulator. In runs in all of those settings.
I ran the numbers from the website you mentioned with TWCalc and here are the results:
From Bottom to Top
Hole 1 6mm dia. position 57.6 mm
Hole 2 8mm dia. position 90.6 mm
Hole 3 6mm dia. position 109.3 mm
Hole 4 6.5mm dia. position 142.9 mm
Hole 5 6.5mm dia. position 170.5 mm
Hole 6 6mm dia. position 196.8 mm
Quite a different set of number than the website.
If you reverse the website number to measure from the bottom, you get:
48 74 106 131 159 183
The explanation for this is simple.
The web site defines drilling the holes using a pair of scissors.
Now, even if you have a micrometer that measures the hole in the thousandths, if will still come out different if you drill the holes with a normal drill. The scissors cuts a tapered hole.
Air has a viscosity to it and the shape of the hole influences the way that air flows in and out of the hole. This has been known for ages by many instrument makers. In physics, the term is coefficient of discharge.
So basically, to use the measurements given at that website, you need to use the scissors as instructed.
If you make your whistle, it’s best to use thin tubing to avoid tonehole chimney area influencing your design - you can do all this detail if your going into the whistle business but it’s a lot simpler to use thin tubing for firsttime works. Given that, to keep a whistle in tune with itself - make the fipple and body with no toneholes first and see if it will play the bell note in tune and trim material until it plays the right note.
After that, if you alway measure toneholes from the bottom rather than the top, you stay out of a lot of trouble.
The window of a whistle is complicated to express mathematically. It’s not just a square hole. The inside streamlinedness, the outerwalls and other geometry around the window influences its area.
If you make a whistle the same way every time, with the diameter tubing. You can use the area as a constant to bias the ‘cork to end length’ variable to figure the extact location to put the block on a whistle in a different key. The main thing is ‘area’.
I like that word…
I’ll try to get this up and running some way, somehow. Regedit works, but I know I did have an infection a while back.
Thanks for the advice Daniel, most appreciated.
EDIT: Found my problem…autoexec file not present due to previous bug. All works now!