Retuning an old whistle

I’ve heard a lot about tweaking whistles to reduce noise, clarify tone, etc, but I’m not aware of people generally retuning them. (Forgive me if not true, I live a sheltered life…)

Anyway, I found reason to do just that and thought it might interest others. The tuning graph below will show why I needed to - the whistle (as found) followed the navy blue trace down instead of coming in as a perfectly horizontal line. After several iterations, it now follows the green trace, a dramatic improvement.

The story of getting there is at http://www.mcgee-flutes.com/Tin-whistle-retuning.htm.

Be aware I have just migrated my web page to a new server, so I’m told it might take up to a day for the changes to propagate through to the darkest corners of the Internet. (Some would argue I live in one of the darkest corners of the Internet!) In other words, if you see nothing, try again tomorrow!

Terry

Hi Terry! Nice work. Actually, knowing your website, I’m surprised you didn’t do that 40 years ago.

I’ve adjusted the holes on several cheap whistles I’ve had. I know there are at least a couple others on this forum who have as well.

I use a small round file, and some fine sandpaper to smooth the edges. I test the tuning using Flutini (thanks!).

In some cases, I’ve also widened the octaves using a very small Tipple-Fajardo wedge (https://sites.google.com/site/dougsflutes/thetipple-fajardowedge). I cut an acetal rod of approximately the same diameter as the ID of the whistle’s tube at a steep angle (around 3-4 degrees), and then cut that down so that I have a wedge of 1" - 1.5" inches long for a high whistle. Then I file a bit of a concavity into the flat side of the wedge, so that it looks a bit more like Mr. Tipple’s example (it seems to help). I start with a larger wedge than I think I’ll need, and test it out to see. If the octaves are now too wide, I shorten the wedge and try again. I’ve found that the best spot for attaching the wedge is at the inside upper edge of the top of the tube, where it fits into the plastic mouthpiece. I put it on the upper side of the tube so that there’s less moisture in contact with it, potentially weakening the glue. That might not be necessary, IDK.

Depending on the specific dimensions of this wedge, it can improve octave tuning, brighten the tone (due to “improved harmonic lineup”–a McGee phrase, I believe), and also potentially alter the scale of the whistle somewhat. So, you may want to try installing a wedge in a tube which you haven’t yet tampered with, so that you can make the minimum necessary adjustments to the holes, once you have the two registers set where they’re comfortable.

A McGee whistle, eh? That would be a thing to have, now wouldn’t it…

I clearly should have! But I largely abandoned the whistle in favour of the flute. So my sins are coming home to roost.

I’ve adjusted the holes on several cheap whistles I’ve had. I know there are at least a couple others on this forum who have as well.

Ah, very good. It would be interesting to know whether other players are finding whistles that are as far out as mine, or if that’s pretty common. If so, those who do whistle reviews might well include tuning among their comparisons. Flutini offers an easy way to do that fairly.

Depending on the specific dimensions of this wedge, it can improve octave tuning, brighten the tone (due to “improved harmonic lineup”–a McGee phrase, I believe)

I probably snitched it from Prof. Neville Fletcher (ANU), or Prof Joe Woolf (UNSW). I probably should have termed it “improved modal lineup”, given that harmonics are by definition harmonic and are thus already lined up. Calling them partials doesn’t really get us much further - as soon as you start to play, the jet-switching action forces the most way-out modes to become harmonic. But you get the picture anyway - if the modes of oscillation line up without forcing, the strength and quality of oscillation increases.

So, you may want to try installing a wedge in a tube which you haven’t yet tampered with, so that you can make the minimum necessary adjustments to the holes, once you have the two registers set where they’re comfortable.

Heh heh, I think that’s the step I’ve been avoiding. But maybe I’ll wilt…

A McGee whistle, eh? That would be a thing to have, now wouldn’t it…

I said maybe I should make one, not lots! I think there might be one or two whistle-makers out there who have the jump on me!

Terry

Seems that it’s rare to find whistles that are perfect as they come, particularly the inexpensive mass-produced Generations and others of that sort.

What you found there, flat F# and sharp G, is common on that type. Also common is having the entire 2nd octave flat.

Back around 1980 I acquired two fine whistles which I still use, a Feadog MK1 and a Generation C, but neither were acceptably in tune as they came. I packed the heads of both to raise the 2nd octaves and did some carving to even out the scales.

The C was well in tune save for that sharp G (actually F, it being a C whistle) and I played it for years with a bit of tape on that hole. I often saw that, a taped G, back in the old days. One day I said “enough!” and I sawed off a bit from the bottom and carved out all the other holes, and I’ve had a wonderfully in-tune whistle ever since.

It’s quite easy to carve out holes on that sort of whistle with an Exacto knife. No machines needed.

A curious tuning approach made itself apparent to me with several of the various Generation whistles I owned back in the late 70s and early 80s. That is, the notes of each octave got flatter as you went up the scale in a fairly even progression. Add to this the fact that the 2nd octave on these whistles was quite flat, and what you have is an even flattening across the entire two-octave gamut, logical in its own strange way.

For argument’s sake let’s say that the scale of the low octave was something like this:
D 0
E -3
F# -6
G -9
A -12
B -15
C# -18
and that the 2nd octave was around 20 cents flat. So, we have an even progression, meaning that you had to blow Bottom D very softly, and increase your blowing at a steady rate over the entire gamut, to play in tune.

I don’t think the makers ever consciously thought of any such thing! But it was the observable, measurable effect of the way these whistles were tuned.

I’ve come across many curiously tuned whistles over the last 40 years, sometimes high-end whistles from famous makers, who (one would think) wouldn’t have such problems.

I’ve owned two Overton Low Whistles on which the bell-note, Bottom D and Middle D (or the equivalent notes for whatever pitch whistle) were quite flat of all the notes made with the fingerholes, a quartertone flat, approximately. No, skillful blowing could not fix the problem. Partially covering all the fingerholes with tape, or sawing off some of the bottom of the tube, were the only ways to get these whistles in tune. I did the latter and achieved great-playing whistles.

Strangest of all have been the twenty or so Susato whistles I’ve owned, in every key (yes even G flat and A flat). Almost none of them gave a true scale as they came. The oddest thing was that each whistle had its very own unique bad scale; there was no pattern to it. Some had a very flat bell-note but were otherwise good; some had a very sharp bell-note; some had certain fingerhole notes which were very flat, or very sharp, or a combination of both. One in particular was so bad that I ended up sawing off the bottom, drastically carving out a couple of the holes, and yet still needed to have one hole half-covered with tape.

I should have charted all these whistles before modifying them! It would have made an amazing display. But my goal was to have whistles which were professionally usable, and I modified them all quickly without thinking about making written records.

Happily Susatos are made from soft plastic which is easy to saw and carve. As I recall the only Susatos that didn’t require modification were a couple of the Low Ds.

I’m actually surprised when I pick up a whistle that’s perfect as it comes. I think this is becoming more common nowadays, with so many musicians using whistles with ordinary instruments and demanding that the things be in tune. New makers such as Burke, MK, Reyburn, etc produce whistles with quite perfect scales in my experience.

I should mention that there is now an iPhone RTTA system, but it will cost you a staggering $3.99

https://itunes.apple.com/us/app/rttatuner/id870470418?mt=8

I can’t comment on it for the very simple reason I don’t have an iPhone!

Regarding the iOS version: it’s every bit as good as Flutini. In fact, I like the interface better! When you factor in that you pull it out of your pocket when needed, it is one seriously useful and convenient app.

It’s not just for whistles/flutes, either. I used it last week to create a compensated bridge for my mountain dulcimer, for example.

The link to the Mac version of Flutini on the Flutini page is broken :frowning:

Maybe I should write an RTTA for android. FT of the input to determine peak frequency should be simple enough, but does anyone know the averaging algorithm?

This is interesting. Doug states not that the purpose of the wedge is to widen the octaves (that is, raise the pitch of the 2nd octave as a whole) but to raise the pitch of the “upper part of the 2nd octave”

“One of the inherent shortcomings of cylindrical bore flutes is that the upper part of the second octave begins to go slightly flat.”

I wonder which notes, specifically, are meant. Why it’s interesting is that on some cylindrical whistles, particularly some Burkes I have, high B is actually sharp compared to low B. Michael lays out the holes on these whistles so that the whole 2nd octave is in tune to itself, and to the low octave, which leaves low B as the odd man out, being flat.

On these Burkes I’ve enlarged hole 2 so that the entire gamut from Bottom D to high A is right in tune, which means that high B is sharp. (BTW uilleann Concert Pitch chanters usually have a sharp high B, making such whistles great for playing along with the pipes.)

On many inexpensive whistles, on the other hand, the 2nd octave as a whole is flat of the low octave.

BTW Terry’s post-tweak graph is still, to me, unacceptable. I want whistles to graph like this:


:slight_smile:

A wedge can make a difference over the whole second octave. I’ve carved a number of fipple blocks with a tail on the end that sticks out into the bore like a wedge. Does help even out the octaves, although if the tail is too thick it affects the tone.

Sure, but you may have to blow the whistle like this:

/_/–_//-_//_///_/\

:wink:

Hi, Terry.

That looks like a Soodlums/Waltons Mellow D, which is a wide body (key of C diameter) D tonebody. Mellow D’s have a notoriously sharp bell note. When I worked with those tubes years ago, I would lengthen the foot by a couple of millimeters and leave it at that. Eventually, I began making my own wide body D tubes.

Looking at your chart, it appears all the notes are low except the bell and octave D. That makes me think if you were to start by lowering the bell note, the other notes would not need as much adjustment.

The other thing I see is, you’re showing the third to be even more flat compared to the rest of the scale. Are you measuring intonation based on equal temperament? Many of the old mass produced tin whistles are set up with a third that’s flat to equal temperament, though still often a little sharp to just intonation, so the player has some room to blow the note up or down according to their preference. (I set all the tubes I make to just intonation.) A just intonation third is fourteen cents flat to an equal temperament third, so it will be a matter of personal taste whether that pitch is considered in tune or flat.

Best wishes,
Jerry

Edited to add: I also notice the C# is flatter compared to the rest of the scale. This is also a matter of just intonation versus equal temperament. A just intonation major seventh is twelve cents flat to equal temperament. Also, if you raise the seventh to equal temperament, you get a sharper Cnat that’s near impossible to blow in tune, so the just seventh tends to serve better than the equal temperament seventh in that respect.

I just can’t do that. I know players who can! They can pick up any whistle, no matter how bad the scale is, and play it perfectly in tune without having to think about it. It’s amazed me a couple times over the years, doing ‘legit’ gigs (non-trad, sitting in the orchestra pit etc) where there’s a “reed man” (sax/clarinet/Boehm flute) who ‘doubles’ on whistles, to hear the guy pick up a cheap whistle and play with quite perfect intonation, and later get a chance to try that whistle and find it horrible. These guys have perfect pitch, hear the pitch they want in their head, and that exact pitch comes out of their instrument, whatever it happens to be.

I have to have all my whistles tuned flatline, so that I can play them without thinking, on an even breath, and they play in tune. It’s players like me, not great players with perfect pitch, who expose any problems with a whistle’s scale.

Jerry is the master at making whistles perfectly in tune. I got a Generation Bb from him and yes, it was flatline up through the scale on an even breath, no breath gymnastics required. Every Reyburn I’ve had has been like that, and every MK and Burke has been very close (Burkes having that quirky flat low B, yet an in-tune high B.)

I’ve found that some tubes require shortening the bell end to bring the bottom note up to pitch. For some reason, that’s common. The Waltons tubes are the only ones I’ve encountered that need to be lengthened. (I built a tubing stretcher for this purpose, but the results were inconsistent. I also tried moving the toneholes, but with only limited success.)

Here’s what I do to the tubes:

D Feadog: Shorten top 1/16”
Eb Generation: Shorten top 3/16” Shorten bell 3/32"
C Generation, C Feadog: Shorten bell 3/32”
Bb Generation: Shorten top 5/32” Shorten bell 5/32”
D Generation: Shorten top 1/8”
G Generation: Shorten top 1/8”

I shorten the tops on some keys so there’s enough leeway to tune sharper. Some of the tubes as they come from the factory are just barely up to pitch with the mouthpiece on all the way.

On untweaked whistles, this should make a big difference. However, one of the reasons my whistles play more easily in tune is, there’s a broader range of breath for each note that will produce a stable, clean tone. Because of that, a consistent breath will produce a good tone all the way up and down the scale where the same breath might not be in the “sweet spot” of some notes on an untweaked whistle.

Best wishes,
Jerry

Ah, thanks for the positive ID. And the confirmation that mine isn’t the only dodgy one out there!

Looking at your chart, it appears all the notes are low except the bell and octave D. That makes me think if you were to start by lowering the bell note, the other notes would not need as much adjustment.

I did end up lowering the bottom note, but I was suspicious of the small holes on the rest of the instrument with regard to its bore diameter. So, If I lowered the bottom note too far, I wouldn’t need to increase the hole sizes as much, and the top B would still end up too flat. So I took the approach enlarge the holes and then drop the bottom note to suit.

The other thing I see is, you’re showing the third to be even more flat compared to the rest of the scale. Are you measuring intonation based on equal temperament? Many of the old mass produced tin whistles are set up with a third that’s flat to equal temperament, though still often a little sharp to just intonation, so the player has some room to blow the note up or down according to their preference. (I set all the tubes I make to just intonation.) A just intonation third is fourteen cents flat to an equal temperament third, so it will be a matter of personal taste whether that pitch is considered in tune or flat.

Yep, I’m aiming to play with others, so ET is the order of the day.

Edited to add: I also notice the C# is flatter compared to the rest of the scale. This is also a matter of just intonation versus equal temperament. A just intonation major seventh is twelve cents flat to equal temperament. Also, if you raise the seventh to equal temperament, you get a sharper Cnat that’s near impossible to blow in tune, so the just seventh tends to serve better than the equal temperament seventh in that respect.

Hmmm, isn’t there the opposing theory that the leading note should tend sharp? Anyway, again, ET rules, for all its sins, so ET wins!

Terry

That would be a wonderful thing, DrPhil. I’m hearing of more and more people using one or other of the versions for a wide range of applications. Making a compensated bridge for a dulcimer, retuning concertinas, some are even using them for flutes and whistles!

Dan has said he’s happy to help anyone wanting to do an Android version, so if you want to contact me, I’ll put you two in touch. Or just contact him via the iPhone app site.

Terry

All of this suggests that the companies didn’t do much in the way of user feedback before rushing into production!

Some older readers might remember the period (early 70’s?) when the moulding went all pear-shaped at Generation, and many or most whistles were close to unplayable. I note my diary entry for Tuesday 3 December 1974:

“Drove to Barnes & Mullins [London] & bought box of 10 D Flageolets for GBP4.10. Retail GBP6.50. Agent in Sydney is Payton & Sons, Clarence St.”

I remember the chappies at Barnes and Mullins were a bit surprised to have an impoverished antipodean drop in unannounced. Especially when I opened one of their boxes and tried the whistles without asking permission. I did not want a bunch of duds.

From then on I bought them directly from Payton & Sons so as always to have fresh stock - the shops were still full of unplayable ones. To buy from the wholesaler, I had to pass myself off as a retailer, but that wasn’t a total fabrication. I supplied them to students of the W.T.McGee Academy of Flageolation which convened weekly in the Student’s Union reading room at ANU. Funny idea of a reading room. Completely bare but for a table and a dozen chairs. Served us well though - the price was exactly right and the adjacent bar served the best Tasmanian cider!

Dear, oh dear, I just Googled Payton & Sons, now just called Payton’s. (I guess Dad Payton didn’t survive the intervening 40 years.) They are still agents for Generation Flageolets (plus Clarke, Feadog & Howard)! Hmmm, I wonder if I’m still listed as a dealer? Now where is that old letterhead?

On untweaked whistles, this should make a big difference. However, one of the reasons my whistles play more easily in tune is, there’s a broader range of breath for each note that will produce a stable, clean tone. Because of that, a consistent breath will produce a good tone all the way up and down the scale where the same breath might not be in the “sweet spot” of some notes on an untweaked whistle.

Yeah, I can believe that. I didn’t think to mention it in the article, but when I looked closely at the head, there were some pretty obvious moulding issues, particularly a sharp-edged moulding flash projecting into the window. I cleaned all that sort of stuff up before proceeding with the tuning. I guess, at the price they charge, personal attention to such things is not going to happen. Great that it has spawned the after-market improvement industry!

Terry