TubeNet

I'd be curious to hear what other players find on this. It's an interesting idea, because it bears fairly directly on whether fixed is better than removable, and indirectly on a lot of ideas about construction of brass instruments.

So I gave it a try. The collar fits tolerably well on my Holton, so I just pulled it out all the way, and tilted it a bit so as to make a visible crack where you could look into the bell. My experiment was only good enough to look for gross differences, of which the only one I found was that if the crack was pointed towards me the sound was much clearer since it was right there in my face. If the crack was on the other side, I couldn't pick out any difference. But you all out there with the ears and acoustic environment to hear grass grow, might be able to do the experiment at a more interesting level. (Carefully - the bell needs to be supported by one hand or of course it's a sure thing it will fall off.)

I suspect that all bell tenon/receiver sets were made in sets and paired so they left the factory with a good fit and minimal leakage. Of course... tolerances had to be taken into consideration. I've had bell tenons that fit loosely into receivers and some that fit 'just right'. But... I've never had a tenon that was simply too big to fit into a receiver as long as it was truly round.

I make detachable upright bells for Mirafone 186 tubas and I've had several that pretty much 'fell in' to the receivers. Not having a snug fit could leak air but more importantly (in my opinion) is for it to fit snugly so vibrating energy isn't lost. My personal 186 has both a recording bell and an upright bell. The upright bell fits very loose and I'm always checking to make sure the bell screws are still tight. To help effect a seal between the tenon and receiver... I keep a rubber band around the tenon that compresses into the space between the tenon shoulder and the receiver. I don't think it's my imagination but the horn 'feels' more alive and resonant (there's that word 'resonant'). We all pretty much know what it means but cannot explain it!

I've owned one detachable bell tuba before, a Marzan. The bell was soldered on, and I think it had 1 or 2 of the 3 bell screws. To me, no big deal, seemed fine to me. I sold that tuba, and a few owners later, it ended up at Lee Stofer's shop. After being repaired there, I heard that there was an air leak at the bell since it was missing screws, and after fixing that it improved some of the quirkiness.

Go figure.

No more removable bells for me!

Reynolds and olds bell tenons fit so tightly you don't even need the screws.
Conn bell tenons are pretty sloppy by comparison and I was always suspicious of those.

bort wrote:To me, no big deal, seemed fine to me. I sold that tuba, and a few owners later, it ended up at Lee Stofer's shop. After being repaired there, I heard that there was an air leak at the bell since it was missing screws, and after fixing that it improved some of the quirkiness.

Go figure.

No more removable bells for me!

See, this is why I think it would be interesting to do a little direct experimentation. We have the basis right there (above) for a myth, that a removable bell detracts from your sound. Myths don't strictly have to be false, the point is just that they're understood through frequent repetition, as opposed to any verifiable fact.

My "null hypothesis" is that bell attachment is non-critical - it doesn't have to be perfectly air tight and it doesn't have to be mechanically very tight. That's a poor hypothesis because it implies, and I admit, it's a matter of degree. Everything has some effect - a dent, a speck of dust, a music folder that fell into the bell and is still there - and a leak or loose collar will have some effect. Like you can put a grain of salt in a glass of water, and say for sure it's saltier - but at what point does it matter? Well, it depends, so like I say, not a very good hypothesis. My basis for the hypothesis is that, at this point where the bore of the tuba is fairly large, small acoustical effects are relatively less important.

But try it for yourself, and you'll know. I did the same thing with my King, where it's a little easier physically. I think, maybe, I could hear a little difference, with only the slightest opening between bell and body (but still far beyond an ordinary leak, again the bell is just resting on the top of the collar, and tilted slightly to create that opening.)

If Lee Stofer put his hands on my tuba and it came back with fewer quirks, that's about what I would expect in any case! While I was fooling around with my King, I held it in an unusually vertical position for the sake of balancing the bell, and that pushed a water key against my leg. Now, that did make a radical difference in sound. I suppose that as usual, the lower piston knuckle on the 1st valve loop had some water in it. The tubing is all dirty and sludgy. Everything has some effect; some things are sure to get fixed at the same time as your bell collar. Without doing the experiment, it seems to me you have no way to know.

Donn wrote:My "null hypothesis" is that bell attachment is non-critical - it doesn't have to be perfectly air tight and it doesn't have to be mechanically very tight.

That makes a lot of sense. For air to leak, it has to be at higher pressure inside than outside. I think the amount of pressure in a bell, a few inches from that huge "leak" at the end of the bell, has to be incredibly low relative to the outside air.

lost wrote:

You would think so until you try putting a hole in the middle of your bell and listen with your ears.

I would think so, even after listening.

Consider how large the tone holes are on a saxophone when they are close to the bell. Even considering that the pads cover much of that opening, those are huge openings.

Consider the transient effect of the frequencies involved. What makes leaks have an effect is that the sudden rise in pressure at the peak of the pulse is enough to push some air through that opening instead of sending it down the bugle. That sudden rise in pressure is relative to the pressure outside the bugle. The closer the opening is to the bell, the less pressure we are talking about. But the air at the opening also fights back against that pulse, and that's a reason the bell flare improves the quality of the sound--it attenuates the upper and noise harmonics and reflections resulting from the impedance mismatch.

I would say: It would take a mighty loose bell tenon to make much difference out front.

Rick "loose screws" Denney

lost wrote:Out front maybe little difference, but to the discriminating player operating on feel, a whole ton...

Arnold Jacobs: "Play by sound, not by feel."

And there is a fine line between "discriminating" and "deluded". I'm not discriminating enough to know where that line is, however.

Rick "but I understand what you mean" Denney

lost wrote:I'm pretty sure Arnold wasn't speaking to a horn with a hole in the bell.

Your use of the word "discriminating" was not complimentary to those who disagreed concerning the importance of the effect, including those who had conducted some of their own experiments. Think about it. Jacobs wanted people thinking about the music, not about the brass and it's myriad of minute effects.

I'm sure saxophone players are really careful that their pads don't leak, even the ones by the bell. So, it must have some effect. A big leak from an eccentric in the attachment ring? Sure, that's on the same order as a pad that leaks over half its circumference, and it should be corrected. But a slightly leaky bell screw that might not hold static pressure as long as the valves? I sorta doubt it, having never experienced that with my various interchangeable-bell tubas. And if the tenon is really loose, tightening the screw might make the leak worse in any case--I have experienced that.

Rick "stupidly arguing with a moderator" Denney

I can't believe I just read the phrase "null hypothesis" on TubeNet.

I come to TubeNet to get away from grad school and stats classes!

Rick Denney wrote:I'm sure saxophone players are really careful that their pads don't leak, even the ones by the bell. So, it must have some effect.

You're right about saxophone players, but not so sure that you're on solid ground past that. There is a very pronounced effect, but it's very different than what happens in a brass instrument. A leak in a woodwind makes it hard to produce the intended pitch, at all. Brass instruments respond very differently - you can open a water key in the main tuning slide, and of course that's going to cost you a lot in terms of tonal quality, but you can still play; not so with a woodwind, where a leak will utterly defeat the note. So it appears to me the acoustics are very different in just this respect.

I just doubt a very small leak that far down the bell loses much air.

Of course there is air moving very slowly down the horn, and a sound wave moving much faster - I'm pretty sure the air flow speed is essentially zero compared to the sound wave. Possibly an impedance mismatch causes a reflection to the sound wave simultaneously with a leak. I dunno. I know on a trombone an open spit valve affects some notes but not others. I think a leak there causes turbulence and an impedance problem. The flare of a bell is an impedance change, at different points for different frequencies.

I've measured static pressure in trombone crudely with a DIY manometer, years ago. The differential pressure between a point just past the mouthpiece (I inserted mouthpiece into a tee-fitting that sealed to the trombone) and the outside air was about 1 inch of water, or about 0.032 psig. A tuba would have to be even less. (The leak at the end of the bell is VERY large.) This measurement would be easy to do on any tuba with bits, just use a PVC tee from the big box store. Probably cost less than a dollar USD.

A bullet hole in a tuba? Until I looked at some photos on this site, I had never seen a tuba bell that wasn't heavily dented. Those still seem to play.

It's not about air flow leaks, but pressure leaks. The sound wave is a series of pressure fronts whose spacing determines the resulting vibration of the ear drum and the perceived pitch. Each harmonic is a different series of pressure fronts that are superimposed on each other. The higher the frequency, the steeper is the increase in pressure for a given amplitude (that steepness is called slew, and when removed from the time-series of the waveform can be treated like a transient.) Some of them may be strong enough and sharply enough defined to create sufficient pressure to push through a tiny opening in a bell stack. This has the effect of changing the impedance--that frequency will be undermined so some extent. The question is to what extent. Given that the pressure front where the bell stack is 6 or 7 inches in diameter is huge, a tiny leak won't undermine much of it before it has passed. And that's assuming the leak is sufficient to affect a frequency that is important in the sound. Hence, the small effect. A water key is an opening that is a significant percentage of the diameter of the tubing, and therefore affects a much wider range of frequencies (lower frequencies have shallower transients that the water drain will still pass) and also will affect them more profoundly. In the smaller tubing, the static pressure (which is the result of air flow) will also cause a noticeable leak.

Trumpets have much higher frequencies in the harmonics, and therefore much steeper transients and higher instantaneous pressures. But I doubt a pinhole in the trumpet bell area would have much effect compared to further upstream in the bugle.

Rick "static pressure is only part of the story" Denney

I do wonder though if air flow or static pressure could have something to do with stability of pitch. If you buy the notion that air flow is considerably less in a saxophone, than a tuba, and that the saxophone's pitch resonance is much more easily destabilized by a leak. Let's see, where's my ophicleide - oh dang, I never did get one!

I agree from experience with the necessity of a good fit for a detachable bell. I used to have a sawn-off St. Pete upright bell attached to the Miraphone tenon on my BBb 186 (taken off the stock recording bell) before I converted the bell. 1st ledger line Eb - 1st valve was always iffy. I don't know if it was a leak, or if the extra mass of the tenon/receiver interfered with a node or anti-node, but I can tell you that once the Besson bell was retrofitted the intonation of the Eb solidified, as did some other notes.

OTOH, I purchased another tenon for the stock recording bell, so I could have both, had the recording bell assembled, and when I used the recording bell, I didn't have the intonation problems with Eb. But I did have the typical 5th-partial flat notes, that were not so bad with the upright bell. I attribute this to dumb luck: with the St Pete bell, it was too short, and I had to use a longer main tuning slide. With the recording bell, I used the stock tuning slide. I attribute the "fixing" of the flat 5th partials with the upright bell to having just enough cylindrical tubing in the right place - about two additional inches on the main tuning slide, that it helped sharpen those notes.

When I changed out to the Besson bell, its tail was slightly smaller in diameter than the Miraphone, so when my tech trimmed it, it came out to exactly the same height as with the cobbled St. Pete bell, so I used the longer tuning slide. Again, the 5th partials were much improved, with B nat, 2nd space C, and Db good, and mid-line D only slightly flat, and with no adverse consequences to the intonation of any other notes. I consider myself very, very lucky, indeed!