Sax on the Web Forum banner

A Very Informal Experiment on Body Tube Vibration

24K views 97 replies 20 participants last post by  kymarto 
#1 ·
I'll be interested to hear what those of you with strong physics/acoustics background think about this. I was surprised how strong the neck vibrations on 5th-line F were, and also was interested by how much the (very tight) neck tenon screw vibrated. Further, the strength of the bell vibrations (which were negligible on low Bb) on high Bb were fairly pronounced. Of course, I realize that these may or may not make any audible contribution to the perceived sound of the instrument, but the vibration of a standard thickness saxophone body is very much in evidence here--based on my rather non-scientific study, of course.

 
#2 ·
Without tone holes the scenario changes to something more like a brass instrument, where the whole body tube is participating, as all the sound is radiated from the bell. Did you try this with a regular sax? In brasses body vibrations are quite evident, and even in saxes and flutes it is measurable in normal playing (especially at pressure antinodes). However as to it affecting sound, it doesn't appear to. Read this:

http://perso.univ-lemans.fr/~jgilbert/ArticlesPageWeb/2008_Nief&al_JASA_Influence_wall_vib.pdf

It is interesting to note that unsupported, flared bells of brasses do definitely change the sound, by as much as 2 dB at some frequencies, and this is well within the threshold of perception, but in blind tests with these different bells, ten top trombonists all failed to detect the difference.
 
#3 ·
Without tone holes the scenario changes to something more like a brass instrument, where the whole body tube is participating, as all the sound is radiated from the bell. Did you try this with a regular sax? In brasses body vibrations are quite evident, and even in saxes and flutes it is measurable in normal playing (especially at pressure antinodes). However as to it affecting sound, it doesn't appear to. Read this:

http://perso.univ-lemans.fr/~jgilbert/ArticlesPageWeb/2008_Nief&al_JASA_Influence_wall_vib.pdf

You have to be very discriminating to be able to hear a 2 db difference. Most people can hear 3 dB.

I can hear 1 dB, but it's gotta be headphones, and a good day.

It is interesting to note that unsupported, flared bells of brasses do definitely change the sound, by as much as 2 dB at some frequencies, and this is well within the threshold of perception, but in blind tests with these different bells, ten top trombonists all failed to detect the difference.
You have to be very discriminating to be able to hear a 2 db difference. Most people can hear 3 dB.

I can hear 1 dB, but it's gotta be headphones, and a good day.
 
#4 ·
Interesting topic. I have "visited" this type of discussion before and feel fortunate to have someone as knowlegeable as "kymarto" to direct us to some of the scientific date on the subject. I recently puchased a tenor from a friend who is assembling these saxes here in Canada. One of the things he does is "tune the bells". He personally tunes each saxophone bell to a specific pitch just as a Tibetian bowl is tuned. Initially, I was not buying into the tuned bell thing much. However, after a couple of months of blowing this sax I can certainly feel a difference in the way the tube vibrates as compared to other saxes I play, most notably my Silversonic. Now, this makes a difference in the way I "feel" the sax and really is an enjoyable connection to the instrument. I am not convinced it translates into a huge difference in the actual sound that is produced but, I must admit to liking the "good vibrations". The scientific data does not appear to support a large difference in tone colour for a given change and that would seem to include neck tenon screw tension, thumb rests or even ligatures. I'm sure we won't solve this quandry quickly. Like Toby, I suspect, as players, the difference we percieve in "feel" translates to a difference in timbre in our minds. For now, that is good enough for me. Nice experiment! I enjoyed hearing the sax without tone holes regardless of the outcome of the experiment. Thanks.
 
#5 ·
The bell of the sax is, in fact, the only point where significant vibrations take place, but even these are very minor compared to what happens in brasses. Given that I have a huge collection of bells from all over Asia, which produce beautiful, sustained tones, and when I tap my sax bells they go "clack" or at best "thunk", I am interested in hearing how your guy "tunes" bells.

It should be noted that to "tune" a bell presupposes that it has one basic, strong resonant frequency, otherwise it would not know what pitch to play when you tap it. That would, one supposes, mean that it would be excited mainly at that frequency, and you would experience a strange sympathetic vibration at only that frequency and a couple of sub-frequencies. No way you are going to get a broad spectrum vibration if you "tune" something.

On top of that, tuning a bell is total BS in another sense: you set a bell to ringing by tapping it at a single point. This introduces an elliptical distortion, which actually makes the tube oval, In this dimension, a metal tube has rather low resonant frequencies. OTOH the air column standing wave introduces a cross-sectional expansion--pushing out against the metal equally in all directions. In this dimension, a round tube, even flared, has extremely high resonant frequencies, which cannot couple and sustain at playing frequencies. Think about tapping a bell with eight mallets simultaneously around the circumference. Do you think the bell is going to ring? Not hardly. So as far as I can see, tuning a sax bell? That dog don't hunt. That bell don't ring.

Whatever you like about your instrument, I can pretty well guarantee it has nothing to do with "tuning" the bell. Forgive me if I seem harsh, but I just hate all the marketing hooey surrounding so much about wind instruments...

Toby
 
#7 ·
The bell of the sax is, in fact, the only point where significant vibrations take place...
In this informal experiment, it was the neck, several cm out from the mpc, and (oddly enough) the tenon screw that seemed to have the strongest vibrations. Even on a low Bb, the bell barely vibrated. In fact, the strongest bell vibration occurred on high Bb, rather than on low, as I would've anticipated. I still make no claims about whether the vibration of the body contributes to the sound, detracts from it, or has no effect whatsoever. I'm going to buy some lead tape (for golf clubs) and tape it to the horn at various points to see what, if any, effect that might have.

Addendum: I did try this a bit today with my regular saxophone, Selmer Serie III. I could really only play left-hand notes, so that the right was free to touch the metal soprano mouthpiece to various points on the horn. In general, the vibrations were less, but were present. Strongest vibrations were on the neck, just in front of the upper octave vent pip. Those vibrations varied in strength somewhat depending upon the pitch being sounded, but were consistently present. I suppose this is the kind of observation that leads some to experiment with adding mass at those points, just to see what might be affected.
 
#8 ·
Might be interesting to swedge all the keys down and repeat the experiment of the keyless horn and contrast with a normal horn.
 
#12 ·
helps when all your marking is being done by a sessional who lives in a van down by the river! :)
 
#11 ·
This is a great demonstration. Thanks for taking the time to put it on video and share with the rest of us. I have a few observations:

1. The mouthpiece and reed set-up you are using to "track" the vibrations has its own natural resonant frequency---like a blown bottle.

2. There is a principle called "sympathetic vibration" whereby if a source of vibration vibrates at an object's natural resonant frequency, it will vibrate "sympathetically with the source.

3. The body of the entire saxophone was in vibration as you played the harmonics, however it was only when you touched the mouthpiece to a spot that was vibrating at its natural resonant frequency (or a whole number multiple thereof) that you got the greatest response, i.e., the buzz of the reed.
 
#16 ·
Here is an Idea (and I only have a self taught understanding of acoustics)

If we are talking Overtones and harmonics - We are talking different divisions of the primary sound wave. There is the fundamental wave (the full Tube), The splitting the sound wave in half (The Octave), Split into 3rds (the P5), Split into 4ths (The second Octave), and so on....

If you took a dry erase marker and a ruller and marked a ruler on your saxophone, would those "hot spots" for resonance be in relation with the harmonic being produced? It would support that the Octaves would resonate more because you had 4 complete wave forms reinforcing the fundamental.

But then again - I am very much a self taught acoustical hobbyist and if somebody discredits my thoughts with some major factual information - believe them - not me.

Charlie
 
#17 ·
That is an interesting thought. Benade says that the closely in tune harmonics that are below "cutoff" frequency "talk" to one another thereby adding energy to the "regime of oscillation. The areas of strongest vibrations in the body of the tube would be at the pressure anti nodes where the waves are exerting the greatest "push" against the wall of the instrument.

This raises a question about something that I don't know. Do harmonics have their own harmonics themselves? That might be the topic for another thread.
 
#18 ·
No, partials are always based on the length of the tube. The name of the game is to get the partials harmonic. All of the impedances for a set tube length will "decide" on a main frequency for the regime of oscillation depending on their relative strengths, and that regime will be mode-locked into harmonic partials, since a periodic wave demands that except in some special cases.

But once you jump to the second octave, for instance, by limiting the participation of the fundamental, its influence on the regime is eliminated, and so if the second partial was not in correct integer relationship with it, that second partial is them free to vibrate at its own strongest resonance. Then all the other partials will again form a new regime based on their respective energies, and a new mode-locked harmonic regime is formed.
 
#19 ·
DSP's experience with the greatest vibration occurring at the neck tenon on an alto is consistent with the properties Conn used to design the 6M with the double socket neck. The double socket neck isolates the mass of the neck clamp from the vibrating tube, allowing the tube to vibrate more freely at an antinode critical for evenness of timbre.
 
#20 ·
Sorry, I don't get it. First, how does a double-socket neck "[isolate] the mass of the neck clamp from the vibrating tube"? Arguably, the extra metal at the tenon would add a lot of mass there, inhibiting vibration.

It needs to be said here that there is no good evidence that any part of the bore vibrates enough to make any difference to the sound. If you know of something I've missed in the literature please let me know.

Now about antinodes--they shift for every note, so "critical" to what at what frequency? It is known that when a bore is made to vibrate (it's not easy to do so and is not believed to happen under normal playing conditions), the effect is that of a local enlargement of the bore. How does this affect "evenness of timbre"? For that matter, what do you mean by "evenness of timbre"?
 
#23 ·
Your ad hom attacks aside, nothing you say is persuasive for me. I own a King double socket tenor, and it has the same amount of tenon contacting the same amount of socket; the only real difference being that the clamp is part of the neck and not the body. Plus there is the extra mass of the inside sleeve. More mass equals less vibration, no? Please enlighten me here.

I really, truly am interested in getting past the hype on this question. What you perceive as chest thumping is my recitation of the relevant literature. You clearly find this inadequate, and seem to imply that the researchers in question are incompetent, deluded or worse, but for all that criticism you do not seem able to point to any controlled studies that contradict them. What evidence you cite--all anecdotal--I would put in the same category as Cannonball claiming major differences based on whether their key touches are made of peridot or turquoise.

In order to bring this discussion down to earth, we need to get past magic and discuss possible physical mechanisms whereby body vibrations might affect sound. Give me more than a "hint" that it does not have to do with effective enlargement of the bore. Likewise, just what do wall losses have to do with anything? The true figure appears to be more like 99%. Those losses are thermal and viscous, and are indeed somewhat frequency dependent, I have read. You seem to be proposing another mechanism by which vibrations affect wall losses. What is it? If it exists it must be testable. Enlighten us, please.

And it now appears that you have narrowed down the improvement of timbre from the double-socket neck to the G-A interval. Based on a different pattern of vibration? Why would that affect only one interval? Or is the vibration changed only in that particular spot? I would suggest a much more mundane, Occam's razor, explanation: Conn changed the dimensions of the tenon in the double-socket design. Or the upper pip. Or both.

And regarding this information about Conn's move to the double socket and its effects: where does that info come from? Is it available in the original?
 
#28 ·
Your ad hom attacks aside, nothing you say is persuasive for me. I own a King double socket tenor, and it has the same amount of tenon contacting the same amount of socket; the only real difference being that the clamp is part of the neck and not the body. Plus there is the extra mass of the inside sleeve. More mass equals less vibration, no? Please enlighten me here.
Sorry if you take my reference to chest thumping as an ad hominem attack, but that's my honest opinion of how you tend to approach these discussions. You showed the same tendency in a discussion of the role of the vocal tract in voicing (although, to your credit, you did eventually acknowledge being corrected). I must say that I find it kind of rich that you make that complaint in the same post where you imply that someone who disagrees with your understanding of saxophone acoustics believes in "magic." Take a look in the mirror before you complain.

The early 6M neck differs from the Super 20 neck in that the neck tenon is recessed away from area of the clamp pinched by the clamp screw. Whatever strain is imparted to the body tube by the clamp is not imparted to the neck tenon, quite unlike the conventional clamp design. Unfortunately, the recessed neck tenon makes it impossible to get a good photographic representation of the gap between the clamp and the neck tenon. But a 1965 10M with the neck tenon flush with the clamp gives us an opportunity to see how large the gap is between the clamp and the neck tenon. That would have to be on hell of a thick body tube wall to fill that gap. It's pretty obvious that the clamp pinches down to contact the body tube in a ring at the base of the joint.
I really, truly am interested in getting past the hype on this question. What you perceive as chest thumping is my recitation of the relevant literature. You clearly find this inadequate, and seem to imply that the researchers in question are incompetent, deluded or worse, but for all that criticism you do not seem able to point to any controlled studies that contradict them.
Hype? We have a difference of opinion as to which bodies of experience shed light on the issues surrounding of body vibration and the sound of a saxophone. It's only been addressed in the most tangential way in the academic literature, and there is a tendency to generalize way too much from it in search of Ultimate Authority. The discussion in Fletcher and Rossing is a good one, and they don't draw any hard and fast conclusions. Sorry if you see critical evaluation of scientific literature as some kind of assault on competence, sanity, and character, but anyone who needs to glean useful information from academic literature for applied science can't regard anything as gospel. In engineering, misconstruing a scope of applicability can get you cited for incompetence.
What evidence you cite--all anecdotal--I would put in the same category as Cannonball claiming major differences based on whether their key touches are made of peridot or turquoise.
Very telling statement. You're looking for excuses to be dismissive of results on the applied end. Conn had an extensive R&D effort going in the 20s and 30s. When high end neck manufacturers (at least two in the modern era, other than Cannonball) stake their professional reputations on ugly looking innovations involving neck mass distribution to improve sound, it warrants serious consideration. And you've proven yourself more than willing to cite evidence you would otherwise call "anecdotal" when it suits your purposes.
In order to bring this discussion down to earth, we need to get past magic and discuss possible physical mechanisms whereby body vibrations might affect sound. Give me more than a "hint" that it does not have to do with effective enlargement of the bore.
I don't recall ever invoking the supernatural. Did you misplace your copy of Fletcher and Rossing? In a nutshell, anisotropic elements of the body and neck result in anisotropic strain response of the body to the stress input from sound waves, and the stored elastic strain propagates shear waves displacing the inner and outer surfaces of the body, which in turn generates compression/rarefaction in air. In the opinion of some, those shear waves lack the amplitude to have an audible result, but that opinion is based on fallacious reasoning.
Likewise, just what do wall losses have to do with anything?
Jeez......EVERYTHING! The thermal losses are the end result of mechanical energy imparted to the horn body. Pretty basic physics. The more elastic the behavior, the slower the conversion of mechanical to thermal energy. The more viscous the behavior, the more rapid the conversion of mechanical to thermal energy. Joule's mechanical equivalent of heat was derived by stirring water (purely viscous), measuring the mechanical input, and measuring the increase in water temperature.
The true figure appears to be more like 99%.
As I recall from Fletcher and Rossing, that 99% loss figure includes the 90% of the player's energy that doesn't make it past the reed. That's equally as mind-boggling as the 90% of the energy within the horn performing mechanical work on the horn body (and eventual thermal loss), but not as germane to the discussion.
Those losses are thermal and viscous, and are indeed somewhat frequency dependent, I have read. You seem to be proposing another mechanism by which vibrations affect wall losses. What is it? If it exists it must be testable. Enlighten us, please.
The conversion of mechanical energy to thermal energy has been discussed above. You are hereby enlightened....well, that is if you're willing to accept it.
And it now appears that you have narrowed down the improvement of timbre from the double-socket neck to the G-A interval. Based on a different pattern of vibration? Why would that affect only one interval? Or is the vibration changed only in that particular spot?
You had surmised in your earlier post that there would be an effect on specific notes, and I essentially told you that you were correct. Is there a problem?
I would suggest a much more mundane, Occam's razor, explanation: Conn changed the dimensions of the tenon in the double-socket design. Or the upper pip. Or both.
You can scramble for such an explanation if you feel that it MUST be about something other than how the neck vibrates, but changes to the mass impinging on the neck tenon is consistent with other experience concerning the sonic effects of neck vibration. And when you look at the recessed neck tenon of the early 6M, it is separated from the grip of the clamp.
And regarding this information about Conn's move to the double socket and its effects: where does that info come from? Is it available in the original?
It was either a forum post or an article that gave me a "huh, that's kind of interesting" moment a couple of months or so ago. My searches so far have come up empty.
 
#25 ·
The thing that get's me is that for the pro vibration tone people, the vibrations are always good and never awful.

That in itself points to some bias being involved.

The vibrations could never possibly add distortion or alter the harmonics in a way that ends up sounding awful.

The vibrations are always adding gloriously to the tone and are just perfect.

Every guitar vibrates, even electric ones and every violin vibrates, and the results are not always glorious and if the saxes tone was affected by vibrations then it would be a similar situation and not all vibrating saxes would sound great if the vibrations were possibly an integral part of the tone.

A musical instrument that vibrates does not always sound great even if the vibrations play a role in the timbre or tone.

In fact vibrations that do affect the tone for certain instruments are another source of more complex variability for the instrument maker to deal with and it's much easier not to have to deal with vibrations at all.
 
#29 ·
The thing that get's me is that for the pro vibration tone people, the vibrations are always good and never awful.

That in itself points to some bias being involved.

The vibrations could never possibly add distortion or alter the harmonics in a way that ends up sounding awful.

The vibrations are always adding gloriously to the tone and are just perfect.
Nobody is making such claims. I'm sure there are poorly made instruments that vibrate horribly. Your snarky little straw man arguments are about as relevant as a yapping puppy. And I'm a nasty man that likes to kick puppies.:evil:
 
#26 ·
Indeed. Can you imagine the maker of high-end speakers advertising that their enclosures are made of thin plywood, whose free vibration adds resonance to the sound? The fact is that you want the enclosures as resonance free as possible, so that the drivers can work as designed, without added coloration.

A woodwind air column is not exactly a neutral speaker driver, but it is directly connected to the active driver of the reed, and is perfectly capable of radiating sound into the ambient environment witout the need of exciting supplemtary passive elements such as the wood of a guitar body that strings need for amplification.

And this leads to another interesting point: body resonances actually sap energy from the driving element. The sustain of an electric guitar is strong precisely because the energy is retained in the vibrating strings and is not transferred to the body via resonant coupling.

It takes energy to flex the walls of a saxophone, and that is energy that is not going out the end of the horn. Which do you think is more efficient: letting the reed create vibrations in the air column, which are radiated directly into the air, or making the air column vibrations flex heavy metal walls, which then transfer their vibrations to the surrounding air? Which is the more direct transfer of energy into sound?
 
#30 ·
It takes energy to flex the walls of a saxophone, and that is energy that is not going out the end of the horn. Which do you think is more efficient: letting the reed create vibrations in the air column, which are radiated directly into the air, or making the air column vibrations flex heavy metal walls, which then transfer their vibrations to the surrounding air? Which is the more direct transfer of energy into sound?
True, and that energy flexing the wall is why there is 90% thermal loss within the body of the horn. Saxophones are hardly a paragon of efficiency.
 
#27 ·
Another point worth discussing: GFC12 claims that studies done on wall vibration in other woodwinds aren't relevant to saxes because of "multiple scaling issues". Just what are those issues? We need to be a little more specific to determine whether they are relevant or not.
 
#31 ·
I'm looking forward to your working model of a bee the size of a bird that can actually fly and of an ant the size of a cat that can actually walk. Crush a 4" copper pipe beneath your foot - easy. Crush a 1/2" copper pipe woth the same wall thickness beneath your foot - hard. All scaling issues.
 
#34 ·
ths is interesting. i cant always seem to find these tech things in the whats new.

thanks for the references to collect.

surely for the math purposes, the horn walls appear close to infinitely stiff for a normal horn. can some people hear the difference? or are they thinking there is a tone difference based on what their hands feel? it seems hard to answer. the studies kymarto has sent me in the past say trained ears cant reliably perceive a difference. I wonder if tuning the bell helps tweak geometry imperfections and hence improving the tone?

edit: I cant quite accept its all a well understood done deal but it does seem a lot is known and tested. its generally good stuff. im not ruling out there are things not known. if we knew it all, what fun would that be?
 
#40 ·
I made a claim I recalled from my University days and was certainly set straight by kymarto. Although I am not a physics major, I am a curious sort. I did read the studies he suggested and he was gracious and patient with me, also giving me good information by p.m. I have no reason to believe he has suddenly become oblivious to the point of chest thumping. As I stated earlier, I do feel we are fortunate to have an individual who not only has the knowledge but the willingness to share with us. Although it is possible for anyone to be mistaken, I know my errant ways were shown to be just that and I thank him for pointing me to the correct data that clearly supported his statements. This discussion does appear to be another one of these situations (although I have not poured through the links this time as I did when I felt I had a valid point to make..). The point where the discussion seemed to "go south" for me was the suggestion about the Conn double socket necks which really doesn't make any sense to me either, more specifically, on a particular pitch (the interval between G and A)... (I have a Silversonic tenor and can't think that is the case for my sax).
 
#42 ·
Scientists conduct experiments in order to answer questions, or to replicate results in other's research. Experimenting because it's 'fun' doesn't add to basic knowledge, which is the purpose of research.
 
#45 ·
Obviously an engineer.:)

The first thing an experimenter does is a literature search. There's no point in re-inventing the wheel.

What, then, is the reason for designing an experiment, if not to answer a question?
 
#51 ·
"Toby:Your shear wave idea is interesting, but does it have any bearing here? Without evidence, the idea of shear waves introduced by anisotropism in the body, and further those waves having any bearing on the sound produced, are no more than intruging ideas. For the record, I have never seen any mention of such a phenomenon in a metal-bodied instrument.

GFC12:All waves in a solid body that displace the surface are shear waves. The ovaling of a round tube described by Nief et al. is a limited case of shear waves that occur in a tube that was constructed to be as isotropic as possible. In that case, the shear wave has a wavelength of half the circumference of the round tube and lacks an audible effect on the air column because of phase cancellation."

What is the level of the "audible effect"? How do you know that it is strong enough to actually affect the sound perceptibly? And why has no researcher ever mentioned such an effect? If I read you correctly, you dismiss the ovaling waves in Nief because of their isotopic nature, but you are still getting an area change, which is of importance. I'd like to know a little more on your assertion that phase cancellation will nullify the effects of such shear waves on the compression wave in the bore. You got some math on that?

One more point: AFAIK it is generally agreed that in order to have significant vibration in body walls, a resonant frequency of the tube has to couple with a frequency being produced by the air column. In a tube of circular cross-section, even a very thin tube, all resonant frequencies lie well above any significant frequency component of the air column. It is admitted that where the tube loses its circularity, e.g. around tone holes, areas with lower-frequency resonances could exist, and anomalous vibrations in those area have been shown to exist.

Well and good, this is an area where further exploration is called for. But let's look at the tenon and neck. Here we have tubes of circular cross-section, unperturbed by tone holes--pretty isotopic, all in all. The ratio of wall thickness to circumference is extremely high, raising resonant frequencies even higher. In addition many necks are reinforced at the bottom, further increasing stiffness and raising resonant frequencies. The tenon itself is reinforced in all socket designs.Therefore it seems logical to assume that whatever shear waves are produced in the tube will be at an absolute minimum in that area.

Or?
 
#54 ·
The Couesnon Monopole has a weird neck joint thing going on which I don't really like but it's a great sax.

What happened to mine was that the previous owner(s) had somehow managed to push the end of the octave lever into the neck making a dent but the dent went right through to the middle neck sleeve thing or whatever it is.

The Couesnon blows away my Mark VI as far as I'm concerned and all of the Mark VI's vibrations (fact or fiction) adding to the tone don't come close :mrgreen:

Something can vibrate in a certain way as much as it wants to but if it doesn't sound good, who cares.
 
#55 ·
The Couesnon Monopole has a weird neck joint thing going on which I don't really like but it's a great sax.

What happened to mine was that the previous owner(s) had somehow managed to push the end of the octave lever into the neck making a dent but the dent went right through to the middle neck sleeve thing or whatever it is.
That must have been one hell of dent! You can see from the photo in my review that the gap's quite large:

http://www.shwoodwind.co.uk/Reviews/Saxes/Alto/Couesnon_Monopole_II_alto.htm

Fantastic horn though, the Monopole.

Regards,
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top