In Search of the Fountain of Chiff

Whistle
1

Hi all,

Working on pinpointing the source of a very desirable quality in my dad’s old 1977 Generation. I, like many before me, find the attack of many notes comes with a brief and pleasing burst of overtones and scratch before certain notes, most often G and D, before falling into a smoother and more pure tone. While several of my other whistles have mouthpiece dimensions similar to the Generation, none have the same attack. The only one that comes close is a Dixon trad, which, while having a wider voicing window than the Gen, does have a similar if much more subtle attack.

In examining my whistles ranked by similarity to the Generation, I noticed several things.

Number one, while the depth of the hollow behind the fipple block can alter playing characteristics for the better (as in Carbony) or worse (as in most Generations), it is not the source of the attack. Second, while material of both the mouthpiece and body can accentuate or reduce certain tone qualities, the physical feedback I feel during the attack suggests it is a product of the way the air is disturbed as it creates the sound.

This instability is what made me compare the location of the soundblade in relation to the windway. I noticed that the whistles I had ranked as purest and least-attack had the soundblade closest to the floor of the windway, whereas the closer in sound to the Generation had the edge of the soundblade higher up in the air stream from the windway. The Generation’s soundblade is just below the middle of the windway.

Since the whistles I was comparing all had nearly identical voicing window lengths, I do not think voicing window dimensions alone produce this effect. I also notice that one of Jerry Freeman’s tweaks for Generation-type whistles involves extending the soundblade to where the edge is closer to the floor of the windway. Freeman-tweaked whistles do not produce the same attack, and for this are a good comparison. I have not yet been able to account for the effect of a curved windway versus a flat windway.

Anyway, in the next couple days I’m going to make a few whistleheads with interior dimensions to the Generation as close as I can get them, and come out with a low, medium and high blade placement relative to the windway. If this experiment bears anything interesting, I’ll post some pictures and sound samples. I’m very curious if any whistlemakers (looking at you Hans, since you take such a scientific interest in all this) have tried this.

As an aside, I can see why most ‘high end’ whistles end up with the soundblade very close to the floor of the windway- most mouthpieces are assembled from interlocking cuts of tubes, which does not lend itself much to raising the soundblade without undercutting. Second, I have not yet established the effect of blade placement alone on overall tone quality and playing characteristics. It may well be that the negative qualities of my Generation, namely rattle and hiss, are inseparable from the attack so highly prized among vintage whistle enthusiasts.

Cheers,
Madman

2

Good luck with your experiment. I look forward to hearing the results.

My experience suggests a low windway floor relative to the bottom of the soundblade leads to harshness in the second octave. This is consistent with Guido Gonzato’s observations in his low-tech whistle instructions.

3

Noticed another thing. In all the whistles with the attack chirp, underblowing D (OXXXXX) just right gives a sudden leap to around 2400 hz. The ease of reaching this sound correlates with the degree of attack chirp the whistle possesses among the whistles I own. Here’s an (admittedly poor quality) recording of the attack chirp and the 2400 Hz crossing noise. It sounds like I’m doing either tonguing or cuts on the D to A transitions, but I’m not. That’s the attack chirp.

https://soundcloud.com/ravencake/demonstration-of-attack-chirp

Anyway, making the whistleheads has been fairly easy so far, except for increasing the interior diameter of the PVC tubing to accept a brass body. I’m looking into tools that would do the job faster than filing/sandpaper.

4

The spectrum of that sound shows two peaks, one around 890 Hz and another around 2400 Hz. The lower pitch is not unusual: I can produce it on every whistle that I’ve tried it on. The higher pitch is unusual. I tried to reproduce it last night an a handful of whistles, and on most of them, I couldn’t. The most obvious success was on a Feadog Mark 1. I did see some evidence of it on a PVC whistle when I replaced the fipple block with one that had a much lower windway floor, and left it farther than usual from the soundblade. There may some connection between the height of the soundblade in the airflow, as you suggest, but there were other whistles that appeared to have a relatively high soundblade that didn’t produce that higher pitch sound.

On the Feadog, I can get a similar squeak in some fingerings if I let one of the closed holes leak a little, as might happen in the course of a cut: for example, the transition between XXX XXO and XXX OXO when cutting the E. Do you notice something like that? Is that contributing to your sense of chiff?

5

Yes …

My standard test for chirp when trying out an unfamiliar whistle is to play a sustained 2nd register D xxx xxx (not vented), then quickly slurring to a 1st register “lazy” B xoo xxx while at the same time suddenly dropping the breath pressure. Whistles with the right character will “snap” cleanly to the B with a very pronounced chirp. On other whistles the chirp may be anywhere from anemic to non-existent.

It’s basically what you’re doing with your D-to-A transitions at the beginning of your recording, but easier to control consistently. Applying the same test to any whistle lets me compare apples with apples. And I find that the D-B test is an excellent predictor of the overall chirpiness of the whistle where expected throughout its compass.

6

Interesting. I tried that on the vintage Gens and didn’t hear it, but I did hear it on my Carbony, which possess a different kind of quick-response quality, but have very little of the Bb/D leap.

7

Well, it takes a bit of practice to get the breath and finger coordination just right. So maybe try again. You’ll know you’ve got it when you hear it. I own a slew of vintage Gens, and every one of them passes the D-B test.

Sorry, what is that?

8

After some practice, I can fairly consistently produce a multi-phonic while partially covering L2 and L3: XDD XXX. But only on the Feadog, and the PVC whistle with the lowered windway floor, can I consistently produce a higher pitch well above middle D, around 3000 to 3600 Hz. Does that sound like what you’re looking for, MT?

When you say “chirpiness”, do you distinguish this from chiff?

9

I’m not even going to mention chiff, because people have so many definitions of it. I’m focusing only on the very defined and desirable acoustic elements at the attack of notes.

10

The acoustic elements at the attack of notes are also referred to as pop/response.

11

Yes, halfway between unvented D and lazy-fingered B. Most whistles I’ve tried can produce the multiphonic if you hit the partial hole closures just right. But that’s not the test. The test is how cleanly and crisply that multiphonic becomes a distinctive chirp when fingered at normal playing speeds and with the right breath control.

12

Have you noticed a distinction between whistles where the multiphonic is B5 plus D6, and those where the multiphonic is B5 plus something much higher, above D7? At playing speeds, I would expect the former to show a smooth transition from B5 to D6 with no chirp.

And to get back to MadmanWithaWhistle’s original post, once you’ve separated your sheep from your goats, if you look at the position of the floor of the windway relative to the bottom of the soundblade among them, do you notice a pattern?

13

I have in my hand one of my early attempts at a Bb whistle. It has many faults, but it does exhibit the chirp MTGuru describes above. And the floor of the windway is just about even with the bottom of the soundblade. There goes that hypothesis. :puppyeyes:

14

You know, you’re probably right. If the only tones sounded are the two actual fingered pitches, and no chaotic multiphonic, then probably no chirp. But as a player not a maker, I’ve never been quite so analytical about it. :wink:

15

What? You mean you’re satisfied to find a good whistle and play music with it, without analyzing why it’s good? Where’s the fun in that? :wink: