There has been a good bit of discussion about making your own tubes lately, so this information should prove useful… and spare a lot of headaches!
This is a revised version of the formula I have been using… revised because I found some exceptions to the rules laid down in the previous formula.
Just follow these instructions, and you’ll get a tube that’s nicely balanced and in tune, and has a comfortable hole spacing.
Using one of the several available tone hole position calculators or spreadsheets, start by getting the top hole (#1) as close as possible to the optimal local cutoff frequency… slightly lower than optimal is desirable for more accurate cross-fingering.
Holes #2 and #3 should both be the same size, and should be adjusted until the top three holes are evenly spaced center-to-center.
Hole #4 should be 1/64" smaller than hole #1.
Hole #5 should have almost the same local cutoff frequency as hole #4… adjust the size as needed to obtain the closest match.
Hole #6 should be the same size as hole #2 and hole #3 for high whistles, or the same size as hole #1 for low whistles… that’s cheating to allow easier reach, but you can get away with cheating on the bottom hole.
Most people are aware that the embouchre hole position given by a tone hole position calculator or spreadsheet is just a rough estimation, but many don’t know that the bell end position is also slightly flawed.
The original calculations were designed for flutes with a vented foot, so the exact placement of the bottom hole wasn’t terribly important.
The modified versions made for six-hole instruments without a vented foot are slightly off in their calculation of the bell end position, but this is easily corrected
I find that letting the ruler overhang the bell end of the tube by 3/32", and then marking the hole positions given by the calculator or spreadsheet compensates for the error.
I like it fine right here. One can see from the TITLE of the thread what the subject matter is. Any one can skip reading it. There is not that much of it. For those that want to know what makes their whistle whistle, It is interesting…And for those that are looking for something to complain about…there is the PUB.
Good stuff Gary. I guess my big mistake on my one C whistle was on the first step. It was PVC so it wasn’t a thickness problem but there was no difference between the sharp oooooo and the natural oxxooo. Otherwise it was in tune and played well but too breathy for me. I think all my holes were too small. I will try again.
I have had similar findings and the second d to be sharp or of a differant tone than the rest of the notes. It was not the tone tube but the fipple. Now I keep a good tone tube to test fipples. And a good fipple to test and tune tone tubes.
The problem with the bell end position is related to the the “boundary layer” (a thin layer between the bore surface and the air column) in relation to the bore’s diameter. The end correction formula uses a constant for this and what’s needed is a correction based on bore diameter.
I once saw a whistle construction article where the holes were layed out by first dividing the body into thirds. I dont remember much about it, but I do remember that the holes were layed out and then drilled larger to bring the instrument into tune.
Has anyone seen that article? I would be grateful if someone could post a link.
I know the article you are referring to but the link to it is no longer any good. I printed out the article but I don’t have the file anymore. The article was by Erik Torp-Olsen, Whistle Physics and How to Make Your Own.
I have had success with the following (measured from the fipple blade, hole 1 being the top one L being the overall length from the blade to the end of the whistle. This seems to work whatever the key)
Hole 1 = L x 45%
Hole 2 = L x 52%
Hole 3 = L x 61%
Hole 4 = L x 69%
Hole 5 = L x 75%
Hole 6 = L x 83%
I drill the holes small and enlarge to tune. Generally speaking, holes 1,2 &3 are the same size, hole 4 is considerably smaller, Hole 5 is considerably larger and hole 6 is a fraction smaller than 1,2&3.