Intonation...

Whistle
1

As far as intonation goes, what should one be aiming for? The upper and lower range of my generation seems verry unbalanced, the low D plays verry sharp unless at near-inaudable volume, yet the second octave tends to play rather flat unless pushed, then it gets really loud. On my concert flute the two are much more balanced, and it’s generally easier to keep in tune. On a side note - is that why flutes have a taper headjoint?

Also are whistles generally tuned to equal temerment? I’ve seen posts on this site sugesting that they are closer to just intonation. What should I be striving for?

2

My Jerry Freeman Mellow Dog whistle is balanced and in tune.

There is a lot of discussion about proper intonation. There’s not right answer (un?fortunately).. I vote for just intonation, because it generally sounds better – even if (especially if) you are playing with fretted or keyboard instruments.

The only place you really get into trouble with intonation is if you are playing in unison in the same register. Then you need to match the other instrument or they need to match you. It’s more important to match frequencies than it is to match some theoretical standard.

3

my generation seems verry unbalanced, the low D plays verry sharp unless at near-inaudable volume,

Maybe twiddle a bit with the position of the head. On all Generations I have the interval between A (@ 440) and D is pretty much perfect.

4

Some whistles are tuned towards just intonation (Bracker, Reyburn), and some are tuned towards equal temperament (Susato, Burke, MK). Some, your guess is as good as mine.

In comparison to flutes, whistles do need to be blown harder in the second octave than in the first, both to get them to sound cleanly up there and to make the octaves play in tune. Consequently, the second octave will always be louder than the first. Some whistles have the registers tuned more widely, and some have them tuned more narrowly.

If you haven’t already, you might consider trying the poster putty head cavity tweak. In my experience, it can make the second octave a bit more stable and slightly, but noticeably, more balanced with the first in terms of tuning.

5

As far as intonation goes, what should one be aiming for?

Well, this is going to sound silly, and self-evident, but one should always be aiming to play in tune in the given situation. In other words, being “in tune” is situational, not absolute. So for example this issue of Equal Temperament versus Just Intonation: if you’re playing along with fretted strings and/or keyboards (which are in Equal Temperament) you would only be “in tune” if you played in Equal Temperament, but if you’re playing along with an uilleann piper (which will most probably be in Just Intonation, more or less) you would only be “in tune” if you were playing in JI.

The upper and lower range of my Generation seems very unbalanced, the low D plays very sharp unless at near-inaudible volume, yet the second octave tends to play rather flat unless pushed, then it gets really loud.

Yes that’s how the octaves of Generations and many other whistles are tuned: the 2nd octave is intentionally tuned a bit flat, meaning that you must blow the 2nd octave fairly strongly, and somewhat underblow the 1st octave, to keep the octaves in tune. This might be thought of as the traditional octave tuning, and not only Generations and other “cheapies” but also high-end whistles like Overtons were tuned this way. The trouble with this design is, as you point out, that it tends to exacerbate the already-present volume differential between the octaves (on whistles the 2nd octave is louder than the 1st).

The other extreme is the octave tuning of MK whistles, on which the 2nd octave is tuned much sharper, meaning that to play the octaves in tune you have to blow the 1st octave very strongly, nearly to the point of the notes breaking, and you have to rather underblow the 2nd octave. (I’ve owned six different MK Low Ds and they have varied a bit in their octave tuning, but even the one with the flattest 2nd octave had a 2nd octave far sharper than an Overton, Goldie, Dixon, Generation, or any other make I’ve had.)

In the middle are Reyburns, Burkes, and Susatos, and I’m sure others.

But a good Generation shouldn’t be as bad as you describe! Good Generations have full, round 1st octaves and sweet pure 2nd octaves and can easily be blown into tune by the player. Quality control with Generations seems to be lax or nonexistent and you might have to play through dozens of them to find a really good one.

is that why flutes have a taper headjoint?

You can read about the “why” of Boehm flute design in the inventor’s own words, by reading his book:

The Flute and Flute-Playing
by Theobald Boehm
1871

In which he says

"I had prepared, in 1846, a great number of conical and cylindrical tubes of various dimensions, and of many metals and several kinds of wood, so that the relative fitness of each as to pitch, ease of sounding and quality of tone, could be investigated.

The most desirable proportions of the air column were soon found. These experiments show:

  1. That the strength as well as the quality of the fundamental tone is proportional to the volume of air set in vibration.

  2. That a more or less important contraction in the bore of the upper part of the flute tube, and a shortening or lengthening of this contraction, have an important influence upon the production of the tones and upon the tuning of the octaves.

  3. That this contraction must be made in a certain geometrical proportion, which is closely approached by the curve of the parabola.

  4. That the formation of the nodes and segments of the sound waves takes place most easily and perfectly in a cylindrical flute tube, the length of which is 30 times its diameter, and in which the contraction begins in the upper fourth part of the length of the tube, continuing to the cork where the diameter is reduced one tenth part. "

The notable thing about Boehm is that all of this stuff was arrived at by experimentation, not by the application of pre-existing theories. He also had a flute-tube made with movable holes so that he could arrive at the ideal location of each hole by experimentation.

Also are whistles generally tuned to equal temperament? I’ve seen posts on this site sugesting that they are closer to just intonation.

Pretty much all of the different whistles I’ve owned have come more or less strictly tuned to Equal Temperament. However the F# on some of my Generation D’s has come rather flat (where it should be, in JI) and I’ve carved that hole out a bit. I need my whistles to be in ET for the sorts of gigs I do.

There’s really not all that much difference between ET and JI. Take your ET D whistle, throw a bit of tape on the F# and B holes, and voila! you now have a JI whistle.

6

That really is not true.

Consider: You have a just-tuned D. The piano plays an A. If you play your just-D, then you are “in tune” with the piano. If you play an ET D, then you are, necessarily, out of tune with the piano. If the piano were to play a D instead, then the situation would reverse.

In all the celtic-rock and celtic-folk bands that I listen to, the pipers tune their chanters just, to their drones, and to me they sound great against the ET instruments. It just depends how the ET instruments are played maybe. You have to leave room for the solo voice or it /can/ clash. If you look at a Beethoven violin sonata, you will see that he pretty much avoids putting the violin and piano in unison – more so in pieces where the violin would tend to be further (intonation wise) from the piano.

If you listen to any of that music from the 70’s with brass sections they are playing just over electric guitars and pianos, and it sounds awesome. The brass sections are always in-tune with themselves, which means they must (necessarily) be playing just intervals, which cannot be ET.

If you listen to any orchestra I can pretty much guarantee that the violins are using one intonation system that is not ET and not the same just system the winds or brass would use. But they make it work. Finally, if you examine the intonation that Robert Plant sings with, well first you’ll go crazy, but then you will realize that what sounds good isn’t /always/ what theory would predict. Theory is a way of trying to explain what musicians do, not the other way round. ET make theory a lot easier to understand, but it rarely makes music sound better.

7

I think that playing in tune is situational too, and that “playing in tune situationally” can be as simple or as complicated as one is interested in making it. Ultimately everyone’s preferences are their own. Here are mine:

I play ET whistles: MK low D, Feadóg tube high D, carved Gens, PVC instruments I build myself. I prefer the ET scale for whistle because it’s more predictable for me, leaving me the job of playing in tune situationally, instead of trying to do it for me.

This isn’t to say that I prefer ET tuning over JI–quite the opposite. The held ET thirds and sixths of accordions or electric keyboards really annoy me. But a JI instrument playing in a key it was not tuned for annoys me too–and I like to play each whistle in as many different key signatures as I can: D, G, A, some E; C, F, maybe a little Bb.

With an ET whistle my approach to playing in tune in any of those key signatures is the same: I can blow any of the held minor thirds and sixths a bit sharp to make them just, and I can underblow or shade any of the held major thirds and sixths a bit flat to make them just. All the fourths and fifths are correct by default, no matter what key I’m in. If I used a JI whistle, I’d have to blow to correct for the JI-in-D tuning of the whistle while attempting to blow towards JI in E or in C or whatever.

And yes, I understand that I’m in the minority for taking my whistling outside of the comfortable one and two sharps. :slight_smile:

8

Sorry, but without more information the statement above doesn’t make much sense to me. (A “just-tuned D” might be at any pitch whatever, depending on what key you’re in.)

People overstate the difference between ET and JI. With the 4th and 5th it’s a mere two cents. The main differences are the Major 3rd and Major 6th, which are around 15 cents flat of ET in JI.

As you point out, acappella choirs, brass ensembles, violin groups, etc will probably be playing in JI whether or not the musicians realise it. One professional musician I know, who played brass in a number of symphony orchestras, dismissed the issue as being a very simple one, saying “yes, you flatten the 3rds a bit”. People with good ears are going to play in tune in whatever situation they find themselves in, adjusting quite unconsciously.

A good violinist I know was telling me how her “default” finger positions on the violin’s fingerboard are ET but she adjusts pitch to the situation. She mentioned how she’ll hit a chord and for a split-second the 3rd will be sharp, but nearly the moment she hits the chord that finger will make its slight adjustment.

About playing Highland pipes tuned in their usual JI way against ET accompaniment, it can sound really bad, and anyone with a decent ear can hear the clash.

I have often been hired to play Highland Cathedral with a brass ensemble or pipe organ, and it puts the piper in a tough situation: the first section is the chanter heard only over the drones (for which the chanter will only sound “in tune” if tuned to JI) but then the brass or organ comes in (for which the chanter will only sound “in tune” if tuned ET).

Most of the chanter’s scale is close, but F# (at -14) and High G (at -31) will sound clearly flat against the accompaniment. It points out exactly why ET was invented in the first place! Because the -16 F# is a gorgeous consonant with the A drone (being a Major 6th), and still fine when the organ is playing a D Major chord (because now the F# is a Major 3rd, at nearly the same spot tuning-wise) but while you’re holding that long F# in the first part of Highland Cathedral the organ chord switches from D Major to F# minor! Well, now that F# is the root of the chord, and should be at pitch (neither flat nor sharp of the organ’s overall pitch). It sounds really bad to have the organ play that F# minor chord at concert pitch with the pipes defiantly holding its note sixteen cents flat! (Unless somebody is tone deaf.)

High G at -31 is obviously the worst, but that note doesn’t appear in Amazing Grace (so as not to trouble generations of pipers, organists, and church-goers) and in Highland Cathedral it’s only a brief passing-tone.

What I do, when playing Highland Cathedral with organ or brass, is to tune my F# a hair higher, so that, true, it’s not sitting in the lovely JI “pocket” with the drones during my solo introduction, but when the brass or organ comes in I can “blow it out” a hair so that note blends. I also tune the entire instrument a tiny hair sharp so that D (not A) is bang-on to the organ or brass… this helps the flatness of C#, D, F#, and G a tiny bit (but of course exacerbates the sharpness of B and E a hair).

9

Take a theoretical GHB and tune it perfectly with low-a = 440.

Take a theoretical piano and “tune” it perfectly to ET at a440.

  • Piano plays A; GHB plays D. They are perfectly in tune with each other.

Piano plays A and D. They are, by definition, out of tune with each other (2 cents), because they are ET, which detunes all the intervals.

Piano plays D and GHB plays D. They are two cents out from each other.


People overstate the difference between ET and JI> . With the 4th and 5th it’s a mere two cents. The main differences are the Major 3rd and Major 6th, which are around 15 cents flat of ET in JI.

agreed!

A good violinist I know was telling me how her “default” finger positions on the violin’s fingerboard are ET but she adjusts pitch to the situation. She mentioned how she’ll hit a chord and for a split-second the 3rd will be sharp, but nearly the moment she hits the chord that finger will make its slight adjustment.

It would be very interesting to measure what she really plays. I would be quite surprised if she actually plays ET, because normal practice reinforces a mixture of just systems. The instrument itself is tuned to perfect fifths (pythagorean tuning). The resonances built into the instrument reinforce playing pythagorean intervals. But they break down when you have double-stops (or chords) and you go to just intervals.

I saw a really neat demonstration of this by the fellow who makes these videos:
http://violinmasterclass.com/en/masterclasses/intonation
I saw him do the demonstration live at a string teacher’s convention. It was really fascinating. He would derive each note in the two different systems (pythagorean and just) and then have his student play them in the context of a Bach solo. Someone asked a question about ET and he dismissed it as something some people thought they did but in reality he did not think it was possible.

About playing Highland pipes tuned in their usual JI way against ET accompaniment, it can sound really bad, and anyone with a decent ear can hear the clash.

I agree that it /can/ sound bad, but it can also sound good. An orchestra /can/ sound bad. You have to have good arrangements.

10

Each note on the piano has two or three strings, and those strings are already tuned a couple of cents apart from each other. The GHB is no more out of tune with the piano’s D than the piano’s D is out of tune with itself.

That’s assuming the piper can hold the note within 2 cents for long enough that anyone could notice.

11

When a GHB D is two cents out, you can hear it beating. At 586 Hz (rounded off just D), two cents is 0.68 hz. If you have that going along with the A, you will hear it beat at an interval of 2/3 of a second. Any reasonable piper is going to blow steady enough to hear that. [It’s actually the 880 A, the fifth harmonic of the chanter D that beats against the A harmonics in the drones. At two cents out the 880 harmonic would be 881, so it would beat once per second, not 2/3 of a second, but stick with the 2/3 to make the next section consistent]. Less steady blowers will be on and off the in tune, but they will still spend more time in the middle than on the edges. With radically unsteady blowers there’s not much point to tuning :astonished:

Anyway, when I play the D above middle C on a piano I never hear it beat on a 2/3 second interval. Those strings might not be perfectly in tune, but they’re not /that/ out of tune. Fractions of a cent, and most would be my guess. I can (and I think most people can) perceive beats down to maybe 10 per second. At 20 per second they become a 20 Hz tone. Twenty beats per second implies 0.05 cents, which is 0.01 Hz at 440.

12

Whatever sounds right! Which, despite all the technical minutiae obsessively discussed on C&F by folk whose ears can’t all be significantly better than mine, is not just the most pragmatic approach to most real-life situations but probably also the most accurately effective…

13

How many notes last 2/3 of a second ? How many cycles are needed to perceive a beat ?

14

Yup. I enjoy obsessively discussing technical minutiae as much as the next guy, but when actually playing, the rule is: it’s in tune if it sounds in tune.

15

If you only play solo whistle, then you’ll never hear any beating. :wink:

In highland piping we frequently hold notes for several seconds, and tuning is considered really important because we always have the drones going.

If you are playing airs on whistle and you have an accompaniment, and if you are playing in unison on some parts, then being two cents out of tune could be pretty annoying. In this context it wont make any difference if it’s tuned just, equal, or randomly, if they aren’t exactly the same, they will be out of tune.

In our pipe band we tune our high-g in a way that is not traditional, and not just; our high-g is actually out of tune with our drones, but they are all tuned the same; we normally get good comments from the judges on our tuning. Which is an illustration of the “if it sounds good it’s in tune” maxim.

16

Just to put all this discussion about 2 cents difference in context: Most accordions that are not tuned totally dry have more than a 2 cent difference between the sets of reeds. 3-5 is quite common and the really wet tuned ones are much further apart. Most orchestras seldom achieve 2 cent accuracy unless they are amazingly good about pitch. When Boulez had New York he achieved amazing pitch accuracy, though I can’t speak to what the temperament was. As already pointed out GHP are heard most often in relation to their own drones where the often two octave difference makes anything at all off the overtone of the drone sound very bad very quickly. You don’t need to wait for the beat against the fundamental. That’s why the pipe scale is what it is. The chanter is designed to be spot on with the drones.

One other comment: A very famous composer who better remain nameless wrote a work for the National Symphony Orch. (Washington DC) for pipes and orchestra. A friend of mine played, I believe at the premier. All the composer knew was that pipes music was written in “A” So, he wrote the pipe part in A major. Nuff said :boggle:

17

So you would blow it into tune. 2/3 of a second is plenty of time to adjust.

18

If anyone else has anything to add to this, feel free. But having begain as an ocarina player, and with the ocarina being an insanly unstable instrument, I’ve found it can be better to hold something off tune then fix the following note. Of course this depends on the length of the note, how off it is and the situation generally.

And of course bending something up/down a few cents will be far less obvious than a larger change, so immidiate compensation shoulden’t draw too much notice.

19

Pancelticpiper is on the right track. I have a tweaked generation that played the 2nd octave flat - until I inserted a short thin tube at the top of the whistle shaft to constrict the bore. My beloved Burke whistle has a narrower bore built in at the top, and the octaves are very well in tune with each other.

20

Interesting, where could you find a tube of the right size?