Scientific analysis of whistles

Whistle
1

So it occurs to me, I can make a setup that will put out a consistent amount of air, and also measure the pressure of that air, and blow it into a whistle to test each note in both octaves for air requirement and backpressure, along with a tuner to see the variance in each hole. Heck, I could throw in a decibel meter and get that number, too. Air volume + air pressure + tuning + decibels, for each note on a whistle. Probably expressed as ranges.

Seems like this would be a really fascinating way to compare whistles. Like, I feel there should be a measurable way to distinguish a good whistle from a great one, scientifically (timbre aside).

Has anyone tried anything like this before? Surely someone has, right?

2

I’d agree with the first part. I’d imagine that some makers might find that information useful in designing a whistles that do exactly what they want them to do. I don’t know if it’ll tell us much that we don’t already know about the ways that various factors affect volume, tuning, breath pressure, and other pertinent qualities. But it might yield something useful that could inform how whistle makers approach their craft.

But that leads us to the problem with the second part. Many people have fundamentally different ideas of what makes a “great” or even a “good” whistle. There is no objectivity to it, just what you want and whether a particular whistle matches that. Makers build whistles to suit their own idea of what they should do, and great makers make whistles that do exactly what they set out to have them do. But that doesn’t mean that any given player will agree!

I’ve played lots of whistles that were obviously well-made and were very successful in doing what the makers wanted them to do, but weren’t at all what I was looking for. Does that make them bad whistles? No! In fact, some of those makers are praised to high heavens here and elsewhere. Some players love them, I’m just not one of them.

So, I wouldn’t get too hung up on some objective measure of the perfect whistle. Just find what you like, and let other people worry about what they like.

4 Likes
3

Oh I get that. At the end of the day I’m measuring a tube of aluminum. The music comes from the musician. “Better” would always be subjective in that sense. But I’d still like to be able to understand why some people have called the Oz Vambrace the best whistle ever made. Arguable, certainly, but why would they say so? The answer should lie at least in part in things one can measure

4

Yes: Blowing machine .

I feel there should be a measurable way to distinguish a good whistle from a great one

No. Best you can hope for is judging how feasible it is to play a whistle in tune over its range. There are too many other variables that go into a great whistle
 including the opinions of the player.

1 Like
5

Great thread. Long. A lot to absorb but interesting. Thanks!

6

But the point of the above is that that says a lot more about the player than the whistle, at least when it comes to whether or not it is the “best.” You can measure a lot about a particular whistle, but whether or not it is the “best” is entirely subjective. All the measurements would do is potentially give you a set of standards that you could match other whistles to, an “if you like X, you’ll like Y”-type of comparison. That won’t give you the “best,” though, it’ll just say that a certain group of people like those attributes. Whatever that Oz whistle does, it obviously does well enough for people to say that. But I’m sure there are plenty of others who wouldn’t say the same.

Take 3 makers that routinely get heaps of praise here and elsewhere: Goldie, Burke, and Sindt. These 3 makers are all making whistles, so there are obvious similarities, but they have very different designs and specs. Goldie and Burke even do different designs themselves! You can measure everything about them and get a set of numbers, and that will tell you something. But ultimately, you’re not going to be able to objectively figure anything out about their relative quality. All 3 are impeccably crafted, all 3 have great players who swear by them, all 3 are “great whistles.” Some like one, some prefer another, it’s all what you expect out of the whistle.

1 Like
7

I have observed that sometimes whistles that sound louder in my ear actually put out less volume as measured by an SPL meter. I think it has to do with the harmonic content.

I think that when you seek out a “best” whistle, you’ll probably incorporate things that aren’t so easy to measure, like how comfortable it is to hold and how responsive it is.

1 Like
8

Yes some things are easily measurable, such as tuning.

Would some players prefer an out-of-tune whistle? They might do, but their session-mates, or band-mates, or employers not so much, I reckon.

And volume. I’m evaluating a couple dozen Low D’s at the moment, their 2nd octaves all have good volume but maybe a third of them have low octaves so quiet that you couldn’t hear yourself play at even a fairly quiet session or with a typical band or ensemble in an acoustic environment. Would a player prefer that?

And what I call “air efficiency” which I measure by how many seconds I can sustain High B, the air-hoggiest note on Low D whistles. Some of the Low D’s I have to hand require twice as much volume of air pushed through them than some others, meaning you have to take breaths twice as often and can sustain a note half as long. But, being no louder than some others twice as efficient, all that extra air isn’t accomplishing anything. It’s simply being wasted. Would a player prefer that?

Or a Low D whistle that has a Low E and Bottom D so feeble you have to really underblow them so they don’t break to the 2nd octave, and are indistinct and barely audible. Would that be a player’s preference?

“Tone” (or more properly “timbre”) would be the aspect most subject to taste, I would guess. But I have several Low Ds in front of me that have a clear voice in the low octave but in the second octave have a thin metallic whine, or audible hissing air noise, or strong presence of the low-octave note (multiphonic). While not measurable these things are equally audible to all. Would a player prefer a whistle like that?

All of that stuff goes into my deciding whether or not a Low D is suitable for professional uses or taking to a session.

1 Like
9

You’ve all raised some important points that I didn’t think through. I think what I had in mind might still have some value, but I’m gonna have to chew on it for a while.

I intended to eventually get some sort of numbering system in place that would tell you in a quantifiable way what whistle reviewers are trying express, and maybe some things they didn’t look at.

For example, the octave jump. Some whistles require more pressure to hit the 2nd octave than others. If I measured 100 whistles and got ranges for the octave jump on each note in each whistle, then I should be able to have a 1-10 scale of ranking. Same with variance between note pitches at a consistent air pressure. I could rank them 1-10 based on the variance needed, with 1 being basically none.

Now imagine you’re looking for a whistle and they come with stat sheets:

Octave Jump Pressure: 4
Pitch variance: 1
Bell Note Pressure: 2
High B pressure: 6
1st octave volume: 3
2nd octave volume: 6

Just spitballing, since I’m sure you could come up with like 20 things that could be measured this way. And if you really wanted to get tricky, I bet you could find a way to quantify the difference in timbre between the 1st and 2nd octave by looking at the complexity of the spectral image of the sound. The idea would be that if you were looking for a certain type of whistle (easy vs hard blower) you’d have a reliable way of knowing HOW easy or hard it was, and what the tradeoffs are.

Furthermore, I strongly suspect you would begin to see clusters of measurements in what are considered top tier whistles. There might be 5 things that always rank high or low across various types and makers, which are consistently correlated with people saying they’re great. Those things would presumably also be similar across different whistle from the same maker. I don’t know what those might be, necessarily, but I bet they exist. Maybe I shouldn’t have said anything about “best whistle” because obviously, everyone has something in particular they’re looking for (and Goldie et all make varying whistles on purpose), but I’m still not ready to give up on the idea that some whistles are simply objectively better than others and it should be possible to quantify and understand at least a major portion of the difference.

This might all be silly, though. Smart people tried this before me and it didn’t result in what I envision.

1 Like
10

You mentioned somewhere earlier that you are a singer. Consider the challenge of quantifying the human voice, such that for any singer you can assign a single metric so people can easily see which human has the best voice. It would be difficult for many of the same reasons it is difficult, and perhaps nonsensical, to quantify the merits of musical instruments. Instrument design is always a complex trade off among many factors.

2 Likes
11

That’s a very good point. How on earth could you arrive at declaring “the best voice?” It does seem nonsensical.

Although I do want to explore it a bit. Limit it to “better vs worse opera singing”, which, actually, there are a lot of things you could conceivably measure. As for which voice you find most pleasant and inspiring, that’s certainly a matter of taste. But you can definitely talk about a tier of “the best”.

In practice, assuming an adequate vocal instrument (big assumption), most of it results from technique, which you theoretically could measure with some kind of Star Trek scanner. And I don’t just mean artfulness at singing a particular passage, I’m talking solely about sound production. The good or bad of one’s technique comes out in ways a trained ear can hear, which is why direct tutelage is necessary for opera. And some of it you might never expect, like the speed of a vibrato. There’s a correct rate of quavering. And in fact, you can scientifically measure a lot of a singer’s voice quality. Consider the following chart:

image
image1772×1145 538 KB

There’s a big difference, and you can measure and explain biological structure it takes to do this. This kinda goes with what I was saying, where I expect to find a cluster of traits among the most-respected whistles, from “perfect” generations on up. I don’t know what it’d be, but I bet it’s there. Honestly I’m tantalized by the idea.

One thing I can certainly take away from your remarks, is that very smart and experienced people are telling me I’m on a trail nowhere, and others have been down it before. And I do think you’re probably right. The opera comparison makes your point, not mine. Opera is an extremely narrow field, in terms of what considered correct. You could even say there’s an ideal and we just argue about who gets closest to it, and how. In contrast, from what I can gather about ITM and whistling in general, the ideal is unique to each person. There’s a genre, and it has rules, but if you’re inside that circle, the ideal is your personal expression, not a platonic Perfect Reel.

I think I’ll shelve this idea for a few years until I have the playing acumen and experience to even participate intelligently in the discussion, and see what I think about it then.

Even so, if anyone else has thoughts, I’m all ears. I also appreciate the responses. I learn something every time you guys weigh in!

12

If you scale back your expectations, I think there may still be something of value. Rather than looking for “the best whistle”, you can ask, “What does it take to play this whistle in tune?”

I’m not sure that anyone has charted pitch vs. pressure over the full two-octave, nine-fingering-pattern, range for a variety of whistles. And then, how does that correlate to whistle dimensions. I, for one, would love to see such a study. I’d try it myself if I had the time and resources.

1 Like
13

I have to say, if you’re boiling opera down to spectrogram charts and vibrato rate, I think you’ve more or less lost the entire point of opera. But laying that aside, I would hope that you’d agree that a good Orfeo or Aeneas requires a completely different sound to a good Parsifal or Lensky. You wouldn’t want to see a Tonio that sounded like Siegfried or vice versa. And that’s basically what we’re talking about here, you could get two whistles that are well-made, hit all the criteria that @pancelticpiper talks about, but are still very different. And at that point, how do you then judge “best?”

This is entirely a side point, but I’d highly recommend you check out a new recording of “Die Schöne MĂŒllerin” by Georg Nigl. It is a very different baritone sound than a Goerne or Quasthoff, and yet at least to my ears is quite remarkably fitting. Because of the economics of the singing world, operatic technique is what tends to be focused on, and that’s what we get when hearing lieder as well. But that’s not the only way, and it’s important to realize that there are more ways to skin the proverbial cat. Just try to get past the period keyboard instrument, which makes me very thankful that we now have modern pianos.

1 Like
14

@gruevy I applaud your effort to quantify things. Huzzah!
Given the right testing equipment and enviroment, it’s possible.

If there were a database of your results, and an instrument scored in a specific way on factors such as ‘ease of overblowing’, ‘average tuning across both octaves’ or ‘difference in decibels between octaves’, then your work would be of great worth if it informed the purchase of a prospective whistle buyer.

If I might sway the thread off-topic in another direction than opera; I see motor vehicles as an adequate equivalent.
There are clearly measureable, objective factors at play when considering the a car or bike, there are always lots of numbers to look at, which can be extremely important to us.
And of course we consider the way the vehicle feels to ride/ drive - which is subjective.
The latter does not discount the importance or relevance of the former.

1 Like
15

What I’ve found playing loads of these things is that it really just comes down to three notes

-Low octave E and Bottom D

-2nd octave B

On most Low D’s, in the low octave, you can go down the scale B, A, G, F# and they’ll have pretty much the same volume, timbre, and breaking point but when you hit E they’ll usually be a dropoff in volume, a less-clear timbre, and lower breaking point, in short Low E is weaker. (The obvious reason is that Hole 6 is the most acoustically out-of-postion.)

This has a practical effect on playing ITDM, say, a reel that goes (in ABC notation)

Key: E dorian |BAGF EFGA | B~E3 etc

You’re playing this passage but when you hit E it breaks to the 2nd octave unless you very quickly tailor your blowing to accommodate that feeble note. I don’t want to have to think about that, I just want to play a musical phrase.

What’s interesting is that Bottom D seems to vary independently of Low E. Yes some Low Ds have matching Low E & Bottom D (for better or worse) but the Burke I played many years ago had the most feeble Low E and the most powerful Bottom D in the business. (The new Burke I have on loan now has fixed the Low E issue.)

And I’ve come across Low Ds with a decent Low E but feeble Bottom D. Ideally Bottom D should be the most powerful note in the low octave.

In the 2nd octave, with nearly every Low D in the pile I have on loan now you can play up E, F#, G, A and all have similar volume, timbre, and blowing requirements but when you hit B it’s louder, harsher, and requires more pressure. On several Low Ds High B is so loud and the timbre is so harsh (often with strong presence of low-octave B) that for me the whistle is unplayable. With others the difference is subtle.

(A fingering sidenote, for High B you have to make sure you don’t have an anchor finger down, the fingering

xoo|oox

impacts the clarity of High B on many Low Ds. On some Low Ds you simply can’t overblow Low B to High B with that lower-hand finger down.)

Rare is the Low D where High B sounds smooth and civilised.

Well, you just have to “blow out” High B you say- give it enough extra pressure and some of the harshness goes away.

While generally true, the issue is that on several Low Ds I have on loan High B is a bit sharp already, and “blowing it out” makes it sharper yet. Remember the #1 thing is being in tune. (It’s odd how whistles with an in-tune or even flat Low B will sometimes have a sharp High B.)

What I did learn is that the matter of Low B requiring far more force to make it overblow than the rest of the low octave doesn’t seem to have much practical impact when playing Irish reels. I’m sure there must be tunes that pop directly from Low B to High B but none came to mind. With all the reels with lots of hopping around between the octaves I could think of this wasn’t an issue. It’s more of an issue IMHO in playing airs when you go from High A to High B and the latter note is twice as loud and has a harsh timbre with lots of undertones from the low octave. It just sounds bad.

1 Like
16

It sure would be interesting if John Sindt did Low D whistles! I’m sure they’d be up there with the best. Or has he now?

I’m wondering what’s meant by the “relative quality” of whistles.

Do you mean the craftsmanship, or how they play?

Most of the performance factors that make a great-playing whistle can be measured or if not measured at least perceived by most people.

The thing is, long before I began coming up with quasi-scientific categories of attributes I would just pick up a Low D and quickly conclude whether it was “great” “good” “mediocre” or “unplayable”. These were simply gestalt-y impressions by playing a tune or two.

It was only later, when I was buying & trying as many Low Ds as I could get my hands on, that I began understanding what specific things went into the gestalt. I wasn’t changing these things, only becoming aware of them.

Anyhow yes Burke and Goldie have arrived at really good-playing Low D whistles with many of the same attributes, though through quite different means.

I’ve been spending a lot of time recently going back and forth between a couple 2026 Burkes and several 2026 Goldies. Any pro could take any of these to a paying gig and be in good hands. The only big playing difference between the Goldie and the Burke is that the Burke requires nearly twice as much air to be pushed through it, for the same output.

2 Likes
17

I’ve never regretted doing that kind of investigation, even if it doesn’t lead to any practical application. If it appeals to you, then maybe give it a try with the whistles you already own.

I listened to a new (to me) CD recently and the first thing I thought (because of these recent discussions) was “Well that’s not a Generation
”

1 Like
18

By “relative quality” I do mean craftsmanship, and basically how well it executes what it seems to be trying to achieve. I’d also say that a very important part of overall quality for makers is consistency; some are better than others at reproducing the same exact feel and sound in each whistle that they make.

But again, that’s the point, what they’re trying to achieve may or may not be what someone wants. In the case of those three, there are people who love one and hate another, and that has very little to do with any objective sense of “quality.” All three make whistles that play very differently (Goldie himself makes very different whistles!). I would say that a Burke narrow bore and a Burke session bore have the same basic “quality,” but they are trying to do something different, and some people prefer one or the other. A Burke session bore is likewise not trying for the same sound or playability as a Sindt, but that is not a marker of “quality” so much as what the player is hoping to get from the whistle.

Also, I’ll be honest, I don’t really care too much about low D whistles so I’m mainly referring to their high whistles (this topic, as far as I can read, never specified low D). I have a Burke low D but it rarely gets played; for what it’s worth, it’s a nice whistle and while I’ve sometimes thought about selling it on given how little it is used, I always end up keeping it around.

1 Like
19

For sure High whistle are a totally different thing, which I’ve not paid much attention to.

I don’t play those much and I still use the vintage Generations and Feadogs that I’ve had since 1980. From what little exploration I’ve done it seems that Sindt has captured the essence of the best of those vintage whistles and my early Killarney clone is really close.

1 Like
20

One way to “fix” the problem with the E is to increase the size of the lowest tone hole and move it down the bore. This will make it speak better, but it will also make the ergonomics worse by increasing the bottom hand stretch. For some people this is better and others it is worse. Different makers have different goals.

1 Like