Flute Tone Investigations 2

Irish Flute
1

OK, I’ve put up Flute Tone Investigations 2. I’ve made this a separate topic in case anyone still has comments on the first article.

In section 2, we move to a conical flute played by an Irish player, and we start to delve into measurements and analysis of the tones he produced with some very interesting results. An elephant is involved. If you don’t have a head for maths and graphs, skim past all that to the conclusions.

If you have already read the first article, go to:

http://www.mcgee-flutes.com/Flutetone-Analysing_an_existing_tune.htm

If not, you might want to start at the Introduction:

http://www.mcgee-flutes.com/FluteTone-Intro.htm, but remember to add your comments on this one to the original post.

Watch out for the elephant.

Terry

2

Fascinating.
In critical mode I was left thinking:

  1. There would fewer variables and more data if a tune used F# and C# and one set aside (to start with) C and the second octave (including D5) since we know that we are asking something different of the flute in those cases.
  2. Remind us please how the relationship between the first and second octave works. The second octave is not just the first octave with the fundamental and alternate harmonics missing, because we are changing the jet to make it do that. But should it be looked at (or summarized) as part of the background to the tone study ?

And to those of us who use that downloadable tuner you recommend the elephant is an old friend (but I had wondered about software calibration, microphone and room acoustics).

3

Yup. This was just a way to get into the topic, using something we had readily available. The real work is yet to come. I hope it will prove enough to whet peoples’ appetites for the topic.

  1. Remind us please how the relationship between the first and second octave works. The second octave is not just the first octave with the fundamental and alternate harmonics missing, because we are changing the jet to make it do that. But should it be looked at (or summarized) as part of the background to the tone study ?

I imagine so, but I don’t really know either. It could be that most reasonably similar flutes sound pretty similar up there. I guess it will depend on the extent that the cut-off reduces the harmonics of the second octave to inaudibility. That is not a hard test to carry out and one I will definitely do. Could be as simple as taking a second octave note, and using Audacity to play in succession the note, the note with the harmonics removed and the note with the harmonics doubled in volume. If we can hear a difference, we have to take the harmonics seriously. We could apply the same test on each harmonic in turn until we can’t hear a difference.

As to how the second octave works, any fingering (eg xxx ooo) is capable of sustaining a number of harmonically related notes (eg G4, G3, D4, G4 etc), and we select the one we want by roughly matching the jet transit time (the time an air molecule takes to get from lip to edge) to the wavelength of the note we want. We can do that by speeding up the jet (more pressure), pouting our lips forward (less distance) or a combination of both. Since just making more pressure will also increase the volume, we tend to do the pouting thing too. It’s one of the major differences between whistle and flute - try playing a quiet B5 on a low whistle!

Pouting of course also means that more of the hole is covered, which tends to bring down the cut-off frequency, further curtailing development of the higher harmonics. But the ear is very sensitive up in the 3-4kHz area, so they might yet be audible and significant to our perception of tone.

And to those of us who use that downloadable tuner you recommend the elephant is an old friend (but I had wondered about software calibration, microphone and room acoustics).

Indeed. I was thinking that I should report the equipment that I am using, but also compare it to the kind of stuff people have around. The early results seem to support my feeling that almost any microphone will do for flute. The lowest note (D4) is about 300Hz, and the harmonics were about 30dB down around 3KHz. Old landline telephone bandwidth is 300Hz to 3.4kHz within a few dB, so it would be a poor mic that couldn’t outstrip the nasal old telephone. Most electrets seem pretty good up to at least 5 or 6kHz. Essentially, if it sounds natural on voice, it should be fine on flute. Any sound card should be up to it (although I’m not taking risks). And room acoustics shouldn’t be too much of an issue if you use an omni directional mike and reasonably close miking.

So I’m reasonably confident that if we can develop protocols for analysing flute tone, most people could try it at home, without cost and with pretty reasonable confidence. That brings the very real possibility of providing newcomers to the flute with tools to compare their developing tone with the experienced players’.

Terry

4

Thank you, Terry for this work you are doing. Fascinating stuff. And thank you for discussing the more nuts and bolts aspect of mic’ing. I am particularly eager to compare my Dixon, 4-piece acetal and my Serov 6 key based on a Prowse. The Dixon is quite loud and has a booming low D but to my ear is more like a cylindrical flute than a conic one. The Serov has a much more complex tone, very easily overblowing the octave, and having a particularly complex D bell note when ‘well driven’.
My problem, as for almost anyone else, is that I am “closely coupled” to this acoustic system while I am playing it. . . and so have only subjective perception to go on. These tools will definitely help in my development of better tone production!
Cheers,
Bob