Phill, I assume that your algorithms will calculate all holes in an equal manner. But for a "Cnat "thumb hole that may not be desirable.
If you add a thumb hole to an existing whistle, you get an extra option for a more powerful and in-tune Cnat. You will also be required to keep the thumb on this extra hole in order to play C#, with the other fingers off. So playing in a “D” scale you just leave the thumb on. This is expected and easier to finger than having to also take the thumb off for the C#.
But using TWJCalc with an ET pattern file of 2,2,1,2,2,1,1 will result in a whistle for which you need to open the thumb hole in order to get an ET-tuned C#. Leaving the thumb on will result in a slightly flattened C#, a different C# than what you get using the six hole standard calculation.
Just thought I mention it. Generally I am happier with a flattened C#, not an equal tuned C#, so I may well like the result better. But I am preferring just intonation in general, and TWJCalc’s ET tuning will give me a sharp F# (sharp to just intonation), which I am not keen on.
I had downloaded TWJCalc and was just going to ask Phill in which file do I insert the new scale 2212211 (I know nothing about Java or the C Language, and little about HTML, but did know how to program in Pascal 25 years ago) when I read your post Hans. Not having any music background I don’t know what it all really means except that the thumbhole calculation in TWJCalc may not give the desired result.
Daniel mentioned in 2008 to use his “whistle-diatonic” spreadsheet, which I haven’t been able to find so far. So is there anything that can be used?
I guess an alternative might be to list the whistles I want to modify (not to worry - I’m sending my high-end whistles to their maker for modification), see if anyone else has added the thumbhole and get their measurements, drill to under the listed hole diameter and finish by using a tuner, however, from reading other posts thumbholes seem to be in a small minority (but gaining momentum recently).
Thanks Hans for that clarification. Yes I was assuming that the thumb-hole was an extra hole for an extra note. I had not realised the complexities involved. So, my apologies PipeBender for falsely raising your hopes.
Let me see if I can grasp this subtlety.
Assume a whistle with a thumb-hole looks like this (on two lines):
xxxxxx
x
with the thumb-hole shown underneath. Then
oooooo
x
is C#
xooooo
x
is B
so
xooooo
o
is the fingering for Cnat? If so, then you are, of course, correct. TWJCalc will not do this directly. If I could specify to the calculation engine that it should calculate C# with the thumb-hole closed and Cnat with T1 closed, would the equations still be valid? [Edit: I mean by specifying the appropriate open and closed hole corrections, and distances being to the next open hole etc. Or do the equations rely on all holes being open below. ]
Would enough people be interested should I modify TWJCalc to do just this? Is this a special case of a general solution?
Phill, absolutely no apology necessary. I admire your tenacity with learning, playing, and building a new model whistle.
Well Phill there is no doubt that I am interested.
I had been editing my last post around the time you had been doing yours, but in any case I believe that many more whistlers are becoming interested in ease of sound quality rather than strictly tradition only. I for one am “turning”. And whistle makers like Hans and Michael Burke also think the same way.
BTW this is a good time to ask - when tuning a whistle for a modification, like adding the thumbhole, what is the best compromise note to initially tune the whole whistle? To tune to an 880 “A”?
However, as I don’t understand really understand the acoustics/physics/maths of this (hence the confusions above) I cannot really say yet that this is possible. I think it should actually be quite easy as a special case, but I would really like Hans’ take on it. For what it is worth, my thoughts are:
The calculations of a hole’s position makes reference to the previous hole apparently assuming that this hole open is all that affects the current pitch.
If I allow the concept of ‘previous hole’ to be slightly more sophisticated I think the solution falls out quite nicely.
Assuming that my understanding of the fingerings is correct (see my previous post) and -----> indicates ‘logically previous hole’ or ‘previous open hole’ and TH is thumb-hole.
T1 ----> T2 ------> T3
TH ----> T2
If that is correct then I should be able to write the modified calculator quite easily. Making it’s purpose clear on the interface may be a little less easy.
Daniel Bingamon had something in TWCalc code that I did not understand at the time, but looked maybe similar to this, as special case code for North American Flute, IIRC. Does that ring any bells with anyone? I will need to go look it up again.
OK a quick test shows that a thumb-hole whistle calculation comes out very similar to a non-thumb-hole whistle. The only discrepancy was 0.46mm difference in the placing of T1.
I used mostly the default values in TWJCalc. Based on D5 major scale, with the added C natural
Phill, I agree with your fingering diagrams.
But I don’t know how the main function works in your calculator, so can’t give you specific advice. Basically when calculating hole 1 (L1) the thumb hole (next hole) should not be considered an open hole, but next open hole is L2. But I think you know this already. Once you got that implemented there should not be any discrepancy for L1 using a six or seven hole calculator. Meanwhile you got that small discrepancy as you report. But I am surprised it should only be ~0.5mm, I thought it would be more. Or is this after you made alterations to allow for a closed thumb hole for C#?
Firstly I am using the findLocations2 form of algorithm from flutomat.
The half-mil discrepancy was after I had allowed for the closed thumb-hole. I have however reduced the discrepancy by not subtracting the closed hole correction for T1 while calculating the thumb-hole. The discrepancy is now 0.01mm.
I wonder what causes this? And would you consider it significant? I was hoping to get exactly the same answer. I am worried that, though the difference is small in my test, some other configuration would give more divergent results.
It is interesting that this use of a thumb-hole is a form of cross fingering. Do you think that there are other significant cross fingering patterns that need to be taken into account? (Native American Flute? I know nothing about them - anyone help?). I am wondering here how ‘generic’ a solution I need or whether my temporary special case code is sufficient.
OK, I have created a version of TWJCalc for testing purposes. Tell me whether you think this works as a method of doing ‘thumb holes’.
I also wonder if the term ‘thumb hole’ is sufficiently precise. There are two features of this particular ‘thumb hole’:
it is operated by the thumb, and therefore on the bottom of the whistle
it is used in a way that could look like cross-fingering. ie when C# is sounded hole 7 (T1) is open but hole 6 (thumb hole) is closed.
I use ‘thumb holes’ on my bass A whistles, but these do not have the cross-fingering implications. I wonder if we need to distinguish between the two concepts?
Anyway, try the new option (it is implemented as a new scale - MajorPlus) and tell me what you think.
[EDIT: if you use the standard settings you will need to increase the size of the B hole to get sensible results].
[EDIT: only works for flutomat calculator - which has become the default]
I am no position to comment knowledgeably about any of this, however, sometimes an observation from one that doesn’t know can be helpful in some way. These are observations not recommendations and something that is not apparent to me now may become apparent later as I learn more about this and/or future modifications are made. So with that in mind I don’t expect any response to any of my observations.
Ok so for the 2212211 scale the one item that strikes me is that there are measurements given for the foot of the barrel to hole position for each note, except there is none for the position of the Cnat thumbhole.
For many people like myself who are considering adding a Cnat thumbhole to a preexisting whistle, I would assume that the final version would need fields for entering the already existing hole positions as well as hole diameters or maybe the algorithm already assumes that correct position is predicated on the hole diameters.
PipeBender, Thank you for trying the applet and taking the time to comment. Any feedback is welcome.
I will answer your question about measurements first. Did you just switch to the new scale pattern without adjusting any hole sizes? That results in an invalid solution to the hole placement algorithm, probably because B and Cnat holes overlap. Try increasing the size of the B hole by half a mil. The problem should disappear.
To answer your second point ‘that’s not the way it works’. I think the algorithm was designed with the whistle designer in mind. The available drill bit sizes limits the choice of hole size, since drilling is the most accurate way to create holes. So the algorithm works by allowing the designer to choose hole sizes and calculates where those holes should be placed for the given set of frequencies.
This is only the starting point for making a whistle, though. A whistle maker may choose to deviate from perfect tuning, or may discover discrepancies in the algorithm.
You could try measuring the holes on your whistle, and using the sizes in a standard major scale whistle design in TWJCalc. If this gives a reasonable approximation of your hole placements then you could proceed to the next step (if you are brave enough) of running the calculation for a MajorPlus whistle design. If you are then confident in the results you could try drilling a hole - preferably smaller than calculations show so that the result will be flat. You could then increase the hole size until the Cnat is in tune.
I would recommend against doing this to any whistle that I valued. Chances are that you would mess up the first few times. This could waste a lot of good whistles.
Phill, I just played with your “TWJCalc for testing purposes”.
With MajorPlus selected it does calculate an additional thumb hole, but the thumb hole is not displayed, even though its position on the whistle is displayed with lines, and the measurements. So it nearly works!
As for the minute discrepancy you reported (0.01mm): it is so small I would not worry. Perhaps it appears as a rounding error?
I am not aware of other cross fingerings needing special treatment. I would call the thumb hole case an “additional thumb hole”, as it is an addition to a standard six hole whistle, and does not influence the fingering of other notes, but just provides an additional alternative to the usual Cnat cross fingering.
Thanks Hans, both for trying it and giving feedback…
Where would you expect to find a thumb hole Hans? Try pressing the ‘Rotate up’ or ‘rotate down’ buttons on the drawing. The ‘+’ and ‘-’ work for zoom in and out, too.
Ooops! It does show the thumb hole, if I rotate the hole from its 180 degree default back position, to the front! Just a bit confusing. How about a see-through whistle body, to see the hole on the back when calculating (use a lighter shade of grey)?
Frankly, I think the display of rotation of holes is of no importance, and a bit of a gimmick. I’d recommend removing this feature. It also will make the choosing of hole sizes less accident prone, as at present one can click the offset controls easily by mistake. But it really is all yours to develop as you like, so please forgive me, I don’t want to be demanding in any way!
The 3d was a bit of techy fun - I have never coded 3d before, and the maths of shape distortion on a cylinder was quite a challenge for me… The serious side to the 3d was that I had more of a feel of the reality whistle. The offsets were there for a more complete description despite their being irrelevant to the calculations, and once they were there it seemed logical to represent them on the diagram.
The most important thing, though, is if the tool helps others, and if the users want a simpler drawing of the whistle then I am happy to oblige. We have one vote for it so far, and no votes against…
re 3D: well you did the work already, and it looks quite funky, especially when rotating the whole whistle! How about just giving any hole on the back of the view (like a thumb hole) a light shade of grey, so it looks as if one sees through the body, with some opacity?
Yes, that should actually work out ok by slowly increasing final hole size from flat that should take care of any variations. What is the best compromise note to initially tune the whole whistle before making a modification? To tune to an 880 “A” or is another note a better compromise between octaves?
I’ll start of with modifying the low end and work up. I’m sending the high end whistles to their maker for the thumbhole addition.
Does one always select Flutomat for both Major and Major Plus?
I have found that Flutomat works slightly better for me -but really its a case of suck-it-and-see. I have only made one whistle, a bass A, and that is atypical, having a 45 degree bend. Certainly TWCalc mechanism has not yet been adapted to use the modifications needed to cope with a cross-fingered Cnat thumb hole. That is on the list if folk reckon that the (w)hole idea is a good one …
In the long run I think that calculation tools are only guides - they give you a reasonable starting point. Perhaps others will disagree… if they do then believe them, because they probably have more whistle-making experience than I.
Hans, where between the top two holes on your whistles did you find the Cnat thumbhole most comfortable for most players? Exactly in the middle or a few millimeters either side? Depending on where the thumbhole is placed on my high-end whistles by the maker will probably dictate where I put them on the rest.
This must be a dumb question but from what I gather the tuning of a whistle when it is made is a compromise between the lower and upper octaves so that is why I was asking if there was a specific note on the whistle like 880Hz “A” as the best point of reference to keep the whole whistle in tune when tuning the Cnat? On second thought rather than using A, it might be best to make sure the B and C# were closest in tune since they are the above and below notes closest to Cnat.
I place the centre of the thumb hole about half way between centre of hole 1 and 2, plus or minus 10% of that distance, and exactly 180 degrees around from the main line of hole centres on the front.
As for tuning: tune the Cnat relative to the other notes.
If your whistle is tuned to equal temperament, you could tune to any note using an electronic tuner, then get the thumb hole tuned fine using the tuner. That means usually having A=440, or 880Hz.
If you tune to just intonation, I would tune by ear playing the G scale (on D whistle).
…