Playing Harmonics

Irish Flute
1

I’ve been working on embouchure control by playing harmonics, which has benefited my tone, volume and consistency enormously - both in the low notes and the high ones.

After some persistent practice, I’ve noticed something kind of interesting on my flute which is an antique Rudall. A lot of the harmonics play clearly and initiate easily, but on some notes and some harmonics it is helpful or even necessary to depress some lower-keys. For example:

C, C#, D & Eb play 1st, 2nd, 3rd and 4th harmonics very easily and with very good intonation. There’s no flat foot on my flute.

My low E is in tune without the Eb key, but richer (stronger and more resonant) with Eb. The 2nd harmonic (E2) requires the Eb key to come into good intonation, something that is known and expected on these antiques. 3rd and 4th harmonics are easy with the Eb key, but lose intonation dramatically without it.

F-nat plays fine in the first & second registers without the Eb key, but the 3rd & 4th harmonics really need the Eb key.

My low F# is completely in tune without using the F-nat key, but stronger with the Eb key. 2nd & 3rd harmonics are helped by Eb. 4th harmonic not easy for me, but possible if I add the F-nat key.

On G, I can only get the first three harmonics, but they are pretty easy with or without additional keys.

This was the era of Beethoven, and these flutes were the concert flutes of the day. Therefore they were designed to play into the third register, up to G3 or higher. I’m thinking that the way the harmonics work is beneficial for the quality of the lower notes, and perhaps necessary for the high notes to function at all. The flute makers of the 1800s certainly rested on the skills of earlier generations.

I’ve heard it said that modern flute makers sometimes replicate antiques (warts and all), or perhaps more often, they do their own optimization. Recognizing that the market is mostly for traditional Irish music, some makers might optimize only for the first & second registers.

Do other people find that their flutes do or don’t play well in the third register?

Are easy-to-sound or well-intonated harmonics beneficial for the tone quality in the 1st and 2nd register?

Do some flutes easily play the harmonics, or are the 3rd & 4th harmonics difficult? Is that correlated playing in the third register?

I’m sure these are not an easy questions, but something that all modern flute makers are trying to figure out.

2

Woah, so many questions, tstermitz! I’ll try to cover a few of them…

Generally, we expect the low notes on the tube (D for instance) to be able to support more partials than those up the top of the tube (A etc). (We should ideally use the term partials, rather than “harmonics” as they are not always harmonic!). This is due to the better venting at the low end, and the very poor venting further up. Those notes vented by small holes (eg A) or have small holes and venting holes a long way away (eg E) do particularly badly. Or put another way, really enjoy it when you open a key just below them!

G does pretty well, as you noted. The first open hole (R1) is often reasonably big, with R2 even bigger, and not too far away.

The upper registers don’t only rely on the partials, they are the partials. But it’s not always easy to spot which partial of which fundamental. So for example, third D can be played as the fourth partial of low D (oxx xxx or similar) or the third partial of G (oxx or similar). Sometimes one is far more useful than the other.

And sometimes they need either reinforcing (eg oxx rather than xxx for third D) by opening or closing a hole to narrow down the range of viable partials. And othertimes they need pulling into tune by opening or closing a hole elsewhere.

And yes, some flutes are harder to find the upper notes on, probably for a number of reasons. Small-holed flutes seem better than big (probably cracking nuts with sledgehammers - note how sometimes halfcovering a big vent hole is better than opening “the lot”). But wear on the embouchure hole edge can be a factor too when you are trying to favour the upper partials by directing the jet.

The people who make this an artform are the Latin American Charanga players, see http://www.mcgee-flutes.com/5keyfingers.htm

They prefer small holed (French 5-key) flutes, but often enlarge the embouchure hole considerably to give themselves better venting and more flexibility at that end.

3

Ooops, realised I’d missed one (heh heh, at least one!)

“Are easy-to-sound or well-intonated harmonics beneficial for the tone quality in the 1st and 2nd register?”

Yes. You mentioned your period flute doesn’t suffer from flat foot. Many do, and you can see the effect. Even if you repad and restore them well, if the middle D (say) is in reasonable tune with most of the other notes, but the low D is dramatically flat, then that low D will also be weak. It’s because we’re sending mixed messages to the jet - the return pulse of air that is intended to lift the jet out of the flute becomes a range of differently timed pulses from the different partials. If the restorer then modifies the flute to bring the low D up (not always easy!), the strength of the note, the quality of the note and the intonation of the note all improve.

Which is why players probably back in Nicholson’s day and certainly Irish players back in last century came up with a work-around. Offset the jet at the embouchure to direct most of the energy into the 2nd partial. It’s well tuned and responds powerfully. Because the spacing between the partials is still based on the now silent bottom D, we still hear it as a bottom D, not the 2nd D. We call it “the Hard D” after the piper’s trick, but science refers to it as “missing fundamental”, https://en.wikipedia.org/wiki/Missing_fundamental

It’s instructive to watch the spectrum on a spectrum analyser while blowing xxx xxx from “naive” to “hard”. On “naive”, blowing across the top of the hole, most of the energy goes to the fundamental. As I offset the jet to go to “hard” (by blowing downward rather than across), I see the fundamental just about disappear, and the partials up to about the 5th really go wild. I can see some activity up to about 3KHz, which would be the 10th partial! But the spacing of the partials never alter. So it still sounds like D, just harder.

(You can do this on a well-tuned flute just as easily. It’s just that on a flat-footed flute, you have to do it!)

4

Fascinating stuff, Terry. Here’s another question for you: suppose I overblow D1 (xxxxxx) and I am now perceiving D2. Is it because the harmonics are now (approximately) multiples of D2 (i.e. all the odd multiples of D1 have disappeared) or the spacing of the harmonics is still determined by D1, but there is a psychoacoustic effect that leads us to perceive the sound as D2 (perhaps because most of the energy is redirected at D2, overriding the missing fundamental effect? :confused: ). This is a question that has been bugging me for some time…

5

Thanks Terry. Interesting comments.

I know my question was mal-formed, but it has become more and more apparent to me that good (easy?) harmonics, or partials as you say, are important for tone quality.

6

No, not malformed. I’m just being picky!

I do well remember the late (Prof) Neville Fletcher ever-so-gently picking me up on the use of “harmonics” for partials that were very inharmonic - we were talking flat-footed flutes at the time. Neville was great like that. One of the world’s leading musical acousticians, but never rubbing your nose in it. And always on the lookout for new understandings. A curious question from me “what actually happens when I revoice the clappers on the carillon” ended up with him, me and one of his PhD students researching “what really happens” and publishing the results in the peer-reviewed Journal of the Acoustical Society of America, the first group to have looked into the science of it. What a privilege!

And yes, “good and easy”. Good in that they are in good tune, and easy so we can make use of them!

7

The former, gwuilleann. When you finger xxx xxx but convince the flute to play D2, you are sending the flute instructions to ignore the D1 it would naturally play. The easiest way to do that is to lift the first finger oxx xxx - then the flute has absolutely no choice as D1 is no longer viable. But sticking with xxx xxx, you can overblow it (either more air or a thinner faster jet), or redirect the jet upwards to favour that octave, or all of the above. We tend to do it subconsciously, so you may not realise you are doing it!

It’s worth checking it out on a real-time spectrum analyser (you can download audio spectrum analysers for computer or phone). Interesting in particular to push xxx xxx to D2, but then carefully let it fall back so that D1 is just starting to reassert itself. You can hear and see it happen at the same time. And as the D1 peak starts to become visible on the spectrum analyser, so do the intermediate partials (3, 5 etc) based on it that were missing in D2 mode.

So it’s definitely real, not psychoacoustic. Indeed the fact that we can notice it both aurally and visually at the same time is a real compliment to both our ears and the spectrum analysers!

8

Thanks, Terry! :slight_smile:

9

Yes, thanks Terry. For me an explanation highlighting the partials not necessarily being harmonic helps counteract the impression from most basic descriptions of musical acoustics that never get beyond partials that are harmonic. I guess that, apart from instruments that are hit by something, tubes with holes in are where it is most important.

Say we have a flute where the resonances are such that a partials of a note don’t want to be a multiple of the fundamental frequency. When we blow the fundamental do we get partials that are inharmonic but weak because they are out of time with the jet or partials that are harmonic (because of the jet) but weak because they are not quite at the resonant frequency ?

10

Yeah, ideophones (things like bells, marimba bars, etc which vibrate to make the sound) and aerophones (flutes etc) are probably the problem children. But you do see minor effects in other instruments. The strings in pianos are thick because they have to be able to accept the energy from the hammers (unlike the strings on harpsichords, which are plucked). And that thickness means that they have significant stiffness at the ends, meaning they look shorter to the upper partials than they do to the fundamental. Which is why the piano’s tuning needs to be “stretched”. So they are “almost harmonic”. Whereas the thinner harpsichord strings do not require a stretched tuning.

Say we have a flute where the resonances are such that a partials of a note don’t want to be a multiple of the fundamental frequency. When we blow the fundamental do we get partials that are inharmonic but weak because they are out of time with the jet or partials that are harmonic (because of the jet) but weak because they are not quite at the resonant frequency ?

The second case there. The partials present as exactly harmonic, because the jet switching pulls them back into line every cycle. But they do so reluctantly, and the jet switching is less efficient than in a well tuned flute. You can sometimes hear and feel this as a slightly forlorn sound and sullen response. I have some Pakistani made “Irish flutes” here that exhibit this characteristic well. I keep them as a warning to others. If you play the fundamental and look at the spectrum, all the partials are spot-on multiples. But run through the individual partials one by one, and they are all over the place.

A simple cylinder flute is another good example. Note only are the octaves painfully out of tune, but the notes have that forlorn “I’d rather not be here” sound to them.

11

I heartily disagree about the cylindrical flutes. I have perfected this design for years now and it plays well in tune. You just gotta know what you’re doing and how to construct it. It’s a very delicate balance between stopper position, hole size, tube diameter, and embouchure size. Tested it against a number of other flutes by now (from renowned brands and makers) and also played together with a number of different instruments (amongst them - I played with my neighbors who are classically trained professional musicians) and my flute was not out of tune.
But I guess I gotta send you one, to convince you.

12

Oooh, fighting words! Good on you!

I have perfected this design for years now and it plays well in tune. You just gotta know what you’re doing and how to construct it. It’s a very delicate balance between stopper position, hole size, tube diameter, and embouchure size. Tested it against a number of other flutes by now (from renowned brands and makers) and also played together with a number of different instruments (amongst them - I played with my neighbors who are classically trained professional musicians) and my flute was not out of tune.

OK, you have our attention. It seems improbable that you have managed something makers since the renaissance have failed to do, and that caused Hotteterre (according to legend, it could have been another French maker) to invent the conical bore, and then Boehm to come up with his tapered head bore. Tunborough recently did some modelling to show that a simple cylindrical reduction around the embouchure could achieve good tuning, but that’s still different from a plain cylinder. But hey, it’s possible.

But I guess I gotta send you one, to convince you.

Oooh, I’d like that, but that’s a bit much to ask. You don’t have to convince me. But it raises the question, how does one publically establish the qualities of their flute design?

Coincidentally, I’ve been mulling a related issue. I make 5 different models of conical flute, from the very small-holed and small-bored Grey Larsen Preferred, to the very large-holed and large-bored Prattens Perfected. And I’ve been putting some extra effort into having at least one of each in stock to show people who can drop in to try out flutes. But then I remembered that I have also made Boehm bored flutes with our familiar fingering. Should I have one of these in stock too? Argghhh!

Then your “outrageous claim” that your plain cylinder flute could be in tune then made me wonder, if we compared it with my 5 conicals and my Boehm bore flute, what would we find? Where would it fit in? Oh good grief, this is taking on the Labours of Hercules dimensions! And raises interesting questions, how do we fairly measure the output level of the various flutes? Simply a case of measuring the sound level produced by the same player on each flute? Or is that too simple?

Now, assuming that you’ve achieved good tuning from the plain cylinder (using all the tricks you’ve mentioned), what’s your feeling on the power or efficiency of the resulting flute? Do you feel it matches what we can get out of a good conical bore, or Boehm’s bore, or is there a price to be paid in volume or ease of playing terms?

This all makes me feel like Ratty, in The Wind in the Willows. “Believe me, my young friend, there is nothing – absolutely nothing – half so much worth doing as simply messing about in flutes”.

13

Oh, there are definitely downsides to the design. For one, you need big hands to play it. And the real problem was not to make a cylindrical flute play in tune (I knew it could be done from making Quenas for a few years) but to make it sound like an “Irish” flute with all the “oomph” and being able to make it “bark” and stuff like that. That’s what took a while. I’m currently typing on my phone but I can describe all the details once on the computer (with a proper keyboard). Maybe tomorrow.

14

I feel like I might be stepping into treacherous waters here, not to mention getting in the middle of someone else’s argument… however… :slight_smile:

If we’re just talking about the single issue of playing “in tune” and nothing else like volume or tone, then posting the results of a RTTA test run for a few tunes would let everyone know how well intonated a given flute design is. I discovered the joys (and pains) of RTTA on your web site years ago, and it’s been an eye opener.

It wouldn’t satisfy all concerns because it’s just the way one person blows the flute, and says nothing about how difficult it was to blow in tune. We know of examples of flutes with poor intonation that famous players have managed to overcome. But a RTTA result would demonstrate if a flute can reach the goal of 12TET intonation at all, even if just by one person.

I’ve used the TTTuner app on my Android phone to discover how my flute intonates by recording and analyzing tunes actually played several times through, and then looking at the averaged pitch of each note instead of “chasing the needle” on a tuner by blowing individual notes. Definitely an eye opener about what the flute is actually doing compared to what I think it’s doing.

P.S. I know the subject of 12TET intonation is somewhat fraught when it comes to traditional styles of music played on diatonic instruments, but I still think it’s a good baseline for discussion, even if we assume a few quirks like the slightly sharp “Piper’s C” with a cross-fingered C natural, the flat foot notes, and so on.

15

Sure, I used TTtuner quite a lot to check tuning. And a number of different tuning software plus a proper actual tuner – a Korg OT-120 “orchestral tuner”. I also made a ton of recordings on my phone but so far have not uploaded any videos with the newest design, the one with the optimized embouchure, on Youtube. I might have to do a separate thread about all that.
Just a quick link – but that was an older design, I have since tweaked it some more, so it can be played fully chromatically with a usable half-holed Eb. And I now use a slightly smaller embouchure and different material for the lip-plate.
https://youtu.be/DO2cq1G_duI
And this is the TTtuner-analysis:

But, like I said – the new one is tweaked a bit more but have as of yet not uploaded a video of it.
And of course it might be possible that I am “blowing it in tune” but if that is the case it’s not much effort. So far I only noticed the C#6 might need a minimal push. But I am pretty happy with the design. Never play anything else. My other flutes are basically gathering dust.
Edit: not sure what happened with the E4 there at -20 cents however. And some of the notes are from rolls or more like “artifacts”, as there certainly was no E3 in the tune. Shows the problems of that software, I guess, when not playing just a clean scale but recording an actual tune.

16

Yes, the one time I’ve posted a TTTuner graphic here, I had to edit out the artifacts in Photoshop and just show the main notes of the tune because it will show every partial it picks up. Looks like a good result there.

17

Yeah, as much as it would be nice to be able to play a tune with all the grace notes into an RTTA system, I think we have to cut it some slack. I usually play bits of scales, not from one end of the range to the other, but all over the place so as to simulate a tune. Or I play some song tunes or something else slowish, and consciously avoid grace notes.

18

I just assume that they are part of what is coming out of the flute, so the software is doing it’s job. With me I don’t don’t know if they are partials or if blips from cuts between repeated notes are being picked up.

The only cylindrical flute I have played regularly is a Tipple G and I don’t have much problem blowing the octaves into tune, it’s just part of playing that flute. Much more noticeable is that the shorter sounding length means the holes are positioned so as to give a more even scale, so its mainly between the two octaves that requires embouchure gymnastics.

19

Hello everyone.

Terry.

I was considering joining here to post on this topic previously after reading somewhere (might have been your own site) that you didn’t get on with renaissance flutes for the cross fingering. As the discussion is up again I decided to comment.

In short, I have played the guitar for a long time, and came around to learning baroque music, and so immersed myself in the instruments of the day, then learned of the evolution of the flute and that barrier which is “Boehm flutes are complex technical masterpieces for professionals” was overcome and I set about making a renaissance flute… mostly after hearing Bartz play, but I think there are other nice soundclips at some makers sites also. None of them show if cross fingering is nescessary though. Anyway I first made a Swiss pfife and then copied the dimensions for a Verona flute… I think Puglisi noted those down and there are a few to be found of his online but not his entire work… or was it using the dimensions of one drawn of the pfife article… either way the wood I had was too short so I had to guesstimate a semitone higher and the flute is in C# base note.

I’ve not played or made flutes before, in fact it took me a month to realise the third and top note were a whole note low, i.e. I had picked the “wrong” scale to tune those flutes to :slight_smile: , which is good really because it showed me how the character of the music I was improvising changed completely after I corrected those. I might build some flutes with different scales just for that. Now I have built a pole lathe also which makes it all a bit easier, and might post on my adventures in making flutes so far… but I’m hesitant, partly because I understand a lot of technique is kept quite by makers and I would not want the local chapter of the flute makers guild to send a consort to play out of tune music outside of my window at night to remind me in my dreams ?

What I think though, is that for ten pounds of material costs (a steel rod to turn into a bore bit, some sandpaper) anyone with normal tools (cordless drill, plane) and a bit of patience is able to produce something much nicer and personal than a plastic pipe versions often suggested online. None of this should be considered competition to professional makers, as I’m sure those are a different level altogether…the opposite in fact because if people actually like picking up their (wooden) flute some are more likely to eventually search for more capable models.


Anyway, as it would be I ended up learning traditional Irish music more than baroque. I tend to go for tone before anything else, and the renaissance flute has a nice tone when played that way. Also, I don’t cross finger at all, and though two lower notes aren’t perfect at unforced playing (another story) the top three are all very close or right on in the second octave when compared to their lower octave, and without effort -I could not adjust them without changing the lower octave also. So in the next week or two I will plain record some scale and possibly a reel on a very basic setup and post it so you are able to verify this, and if so I will send you or post the actual dimensions for you to try to replicate if that is of interest.

You look for ideas at your site, and one that I thought up, which must have been thought up before as well, is that for some tunes the scale uses a half note before top, and half fingering is possible but not fluid, and is a bit slurred or awkward… so I made a wooden disc that fits the top hole, scalloped the outside inwards so you don’t notice it, then bored it to the note required. It push fits to place, and pushes through to take out, and works well…worked well for a long time, I only made it out of a chip of unstrengthened wood and ended up breaking it, will make another.


I just made a “Pratten” as well, I wanted to see what they were all about, incredible how the box at top acts like an amplifier, picks up on details. It is a trial version of olive wood, made up of sections of misdrilled (another story) wood and with embouchure set 3cm out by error away from start of taper… which sounds ok up to first three second octave, then the top three are literally a whole note out in second octave, with first octave all in tune… :smiley:.

20

I’ll throw in the only short recording I have for what it’s worth, recorded half a year ago (and have only been playing for a year now), open rythmic piece, with a cheap mic balanced on my knee to record, just as recorded, etc - doesn’t do the flute or its tone justice - but it is just to show that traditional music is playable on renaissance flute, and though for second octave it doesn’t include top two notes (which also play in tune with their first octave), it gives a rough idea of how close the tuning is. I’m sure before baroque flutes made their way to Ireland people were playing renaissance style flutes, but what music exactly before 18th century or so I don’t know exactly…and surely that has found its way somehow through to what is now traditional Irish music.

https://ln4.sync.com/dl/b49cdb1a0/26q2gicj-gwpyrznb-rvrsetug-p84eujx9

I don’t enjoy recording much, but I will try to record some clearer example, or scales.