Untitled Document
     
Advertisement Click to advertise with us!
     

Saxophone mouthpiece study

Here is a comment I received by e-mail from a colleague which I think is worth introducing to the Some points that may help convince the "stubborn" on the Woof mouthpiece thread:

All of the muscles which allow the tongue to move are attached to the jaw, from just behind the lower front teeth to the jaw "hinge" just in front of the ear. Since the jaw muscles overlap the tongue muscles and all the facial tissues are connected in some way, one simply can not move the tongue without affecting the jaw pressure (while playing, as opposed to position) to even some minute degree. Aside from that, tongue position (vocal tract pitch) doesn't affect played pitch any more than the combined Z of the VT and reed resonance together can affect the negotiations of the considerably stronger tube resonances in deciding what fundamental frequency the regime forms at - minimally.

While there is a very wide range of various manipulations used by players at the various levels of proficiency/sensitivity, accomplished players don't think of using tongue position to change pitch, if they have thought about what they do at any length. Jaw/embouchure pressure controls pitch and any change in pitch requires embouchure corner/vocal tract/tongue position adjustments to optimize VT and reed resonance for tonal control (note voicing). Everything has to be done at once so the sensation of what is happening blurs. As Carina Rascher says, one can make jaw/embouchure pressure adjustments for basically, only +/- 3 cents of pitch correction, after which those adjustments render optimal note voicing impossible. Then a new fingering or a better mouthpiece/horn match must be found.
 
I love these discussions with you Toby. Just like the "good old days" on SOTW they send me back to the literature to learn more about the topic.

Since discussions like this one have a tendency to be come a bit scattered and unfocused with the introduction of peripheral issues, let me restate the original questions as I see them.

1. Can the tongue move independently of the jaw and muscles that form the embouchure?

2. Can a saxophonist lower the pitch of a note in the traditional range by using just the vocal tract?

The answer to question #1 is clearly yes. Regardless of where the muscles of the tongue are attached in human physiology, the tongue can move independently. The proof is simple. Clench the teeth to completely immobilize the jaw, and then move the tongue about inside the mouth. The tongue moves easily including the back of the tongue which can go up and down changing both the volume and shape of the oral cavity.

The answer to question #2 is also yes with a caveat. From the abstract of "Measurement of Vocal-Tract Influence During Saxophone Performance" by Dr. Gary Scavone is this quote.
Vocal-tract influence is clearly demonstrated when pitch bending notes high in the traditional range of the alto saxophone and when playing in the saxophone's extended register. Subtle timbre variations via tongue position changes are possible for most notes in the saxophone's traditional range and can affect spectral content from at least 800 - 2000 Hz.

Under "Pitch Bending" the study says:

On an alto saxophone, bends produced via vocal-tract manipulations are easiest when starting on notes above concert Eb5. Bends that start above a concert C5 tend to have a lower limit around the C5 frequency 523 Hz. In other words, the fingered note controls the starting and highest frequency of the bend range but the minimum frequency is generally always between 500 and 600 Hz.

The upstream resonance frequency is mainly controlled via tongue position variations. The pitch bend is only possible for notes higher in the traditional range, where the air column resonance structure is relatively weak. We speculate that the lower frequency limit on the pitch bend is related to the vocal-tract physiology and player's control of the second upstream resonance within an approximate range of about 520 - 1500 Hz, which is close to that reported by Benade.

So the "caveat" in question #2 is only on those notes in the traditional range that are above high A. Pitch bending below that note must be done the traditional way by "lipping down" the pitch. Scavone's study also notes that the range of pitch bending using only the lip is about half a semitone or a quarter step.

In the Cannonball video by Randall Clark his example of a jazz pitch bend is done on a high B natural which is in the range of frequencies mentioned in Scavone's study that can be bent down using the vocal tract. Randall's demonstration of pitch bending on a high D fits well into the range of frequencies Scavone's study indicates are possible. He runs out of steam at a written A natural (concert C) which is the lower limit given by the study. In his reference to knowing player's who can lower the pitch a full octave, Randall is probably thinking of those who can do so on the mouthpiece alone which is different than on the instrument itself.

Of course the idea that directing the airstream downward toward the reed is responsible for lowering the pitch is not correct. Maybe on flute, but on saxophone. I will need to chat with Randall about that the next time I see him.
 
Thanks for finding Scavone.

What I was mostly reacting to was the notion that jazz players are setting their mpcs high and bending down the whole range of the sax--which is why I specified special cases. It only works when the primary horn impedance is weak, otherwise the best you get is timbre variations. What Randal appears to be doing is--generally speaking--using oral cavity variations as an adjunct to--or a causative factor for--subtle embouchure changes that maintain good dynamic tension and breath support: but the factor shifting the pitch in his demos is clearly "lipping down".

I never contended that there are no effects from oral cavity resonances, but only that they are a weak influence across most of the playing range where tube impedances are strong. Lance is right: each factor gets entered in the "regime of oscillation" mix, and only has as much influence as its relative strength.

Randal's video has value, but it is based on a fairy tale.
 
Last edited:
I have uploaded and attached Dr. Scavone's study for those who are interested. Toby, please review this paper and then we will continue the discussion afterward.
 

Attachments

  • Scavone vocal tract.pdf
    929 KB · Views: 329
All of the pitches bent down by Randall Clark in the video are within the frequencies that Scavone's study show can be adjusted using the vocal tract/oral cavity alone without changing the embouchure, yet you continue to claim that he is "lipping" the notes. To me it appears self evident that if those pitches can be lowered using the vocal tract/oral cavity alone that the player's claims that the embouchure does not change are true.

"but the factor shifting the pitch in his demos is clearly "lipping down".

He lowers palm D all the way down to an A---a 4th lower. Scavone's study found that one can "lip" a note about half a semitone, but that "The use of vocal tract manipulations for pitch bend can achieve downward frequency shifts of a musical third or more."

The mechanism by which changing the vocal tract/oral cavity effects a drop in pitch is still unclear to me. I have tried the exercise of bending the pitch on a high B down on my alto by raising the back of the tongue and it works without relaxing the embouchure. I have also made the discovery that lowering the pitch of a whistled note employs the same tongue movement. Is it possible that changing the direction of the airstream is part of the mechanism. To me seems counter intuitive that raising the back of the tongue thereby reducing the volume the oral cavity produces a drop in the pitch. What are your thoughts on this?
 
Scavone indicates that it is possible to drop the pitch quite drastically using vocal tract resonance, but only on high notes. It is admittedly more than I thought. But the caveat is that it can't be done at all in the first register, and it is possible less and less as you go down towards the first register. So while the palm D can be bent quite a bit, the G5 can't be bent nearly as much, and the D5 almost not at all. I am honestly surprised that lipping is only a semitone or so; I guess I am also using vocal tract effects when I think I am simply lipping.

I'm still skeptical about the assertion that jazz players tune high and pull down the pitch using vocal tract manipulation. That would mean that their lower register would always be sharp in comparison to the second and third registers, and get increasingly sharp in the bottom of the second register, as pitchbending possibilities get progressively less and less. Certainly good players use all expressive means at their disposal, and this clearly includes pitchbends in the regions where that is possible. I think Randal indicates that he plays low to pitch, and that is certainly a possibility in the second and third registers. You can play low to pitch and then bring it up when holding a note.

No, air direction can and does have nothing to do with this. There is not really any direction to the air in the mouth, is there? It is a pressure reservior.
 
Well, I finally had time to actually get out a sax and try it for myself. I was somewhat wrong, to say the least. Pitch bending is not like altissimos--it is a broadband phenomenon, so you don't have to tailor the bend to each resonance specifically as in altissimos. I find that I use it, but usually in concert with some slight change in lip pressure, so it is a combination of the two for me. I can also just use tongue position manipulation, but that is more difficult, and on extreme bends it makes the high notes want to drop an octave without some adjustment. Like riding a bike, it is a coordinated effort between many factors: tongue position, jaw pressure, breath support, etc.

I tried holding a constant tongue position across two octaves. Very interesting. Indeed true that at C5 (fingered A) you reach the limit. So simply holding that position if fingered C1 is in tune C2 is already getting quite flat and C3 is more than a semitone flat. So I'm skeptical that jazz players tune high and play low, unless just a bit; otherwise the lower notes are horribly sharp. Maybe it's good enough for jazz.

Also true that the farther you pull the note south of the tube impedance, the worse the quality gets and the harder it is to maintain a steady pitch, since the tube tries hard to pull it up. So this seems good for scoops and whatnot, but not for steady playing, unless we are talking 10-20 cents nominal sharpening of pitch, where jaw pressure can drop the low notes enough to be in tune.
 
It appears the topic has changed from pitch bending to playing lower on the pitch and pushing the mouthpiece in. I watched the video again paying close attention to Randall's comments in this area and I believe you are correct that it is somewhat misleading. What I know for a fact is that jazz player typically play lower on the mouthpiece pitch. The Santy Runyon target for the alto is A=880 for the mouthpiece alone. This is not so much about the pitch per se, but about the embouchure tension and hitting somewhere in the center of the range of pitches possible. Most classical players play at this higher mouthpiece pitch which requires a firmer embouchure. Jazz players typically play anywhere from a whole step to a minor third below this pitch. They do so by opening the teeth which opens the oral cavity, but which also puts less pressure on the reed from the lower lip ie. relaxing the embouchure. The result is a louder, fuller tone that has more overtones.

My thinking from what I have studied up to this point is that opening up the oral cavity and playing lower on the mouthpiece pitch increases the "effective volume" of the mouthpiece. Remember that Benade and Gerbler's figures using a soprano sax mouthpiece and my own using an alto mouthpiece show the "effective volume" of the mouthpiece to be about 30% greater than its measured geometric volume. One of the effects of moving the mouthpiece farther on to the neck is to remove some of the geometric volume. This not only restores the correct "effective volume" that matches the missing cone, but it also brings the instrument's pitch back up to A=440.

What I am still fuzzy about is that if playing lower on the mouthpiece pitch is accomplished by totally by relaxing the embouchure, the resulting tone quality would sound flabby which it does not. This raises an interesting question: What is the relationship between the mouthpiece input pitch and the reed's interaction with the resonance of the body tube?

An example of the alto sound Randall achieves by playing much lower on the mouthpiece pitch can be heard in this video. It is a bit reminiscent of David Sanborn, but it is certainly not a "flabby" sound per se. [video]http://www.cannonballmusic.com/whoszat.php[/video]
 
Why should a more open embouchure lead to a "flabby" sound? First, I have read and it seems logical that by relaxing the embouchure, the reed is not pressed as close against the lay. This increases effective volume and lowers the pitch. So does playing louder, I believe, since the reed travel distance is increased.

I'll have to look a bit more closely at the effects of reed damping, but basically on the sax, as opposed to the clarinet, none is needed to actually sound the note. Of course an undamped reed sounds raucous. When I studied oboe--the French school as opposed to German--I was taught never to bite the reed, but to open and at the same time create a firm bed of muscle to control the reed. German players do not scrape back the lay, but leave it very solid and bite hard. To do this on a French reed raises the pitch considerably by limiting blade travel and hence volume.

So if one reduces pressure but maintains firm lip muscles, the sound will open, the pitch will lower, but with control there will be no "flabbiness" to the sound, which I believe is caused by quick and unintended pitch variations--wobbliness.

Another way to achieve this is by playing further up the mpc. This allows one to maintain a firm embouchure and play down on the pitch, at the expense of some fine dynamic control, which anyway is basically useless in jazz, as distinct from classical playing IME.

In fact I had to really relearn my embouchure when I got a nino. The upper notes just won't come out reliably without biting way up on the mpc. It's interesting to play used mpcs with teeth grooves--you really get a sense of how different the teeth positions of different players are.
 
Why should a more open embouchure lead to a "flabby" sound? First, I have read and it seems logical that by relaxing the embouchure, the reed is not pressed as close against the lay. This increases effective volume and lowers the pitch. So does playing louder, I believe, since the reed travel distance is increased.

I'll have to look a bit more closely at the effects of reed damping, but basically on the sax, as opposed to the clarinet, none is needed to actually sound the note. Of course an undamped reed sounds raucous. When I studied oboe--the French school as opposed to German--I was taught never to bite the reed, but to open and at the same time create a firm bed of muscle to control the reed. German players do not scrape back the lay, but leave it very solid and bite hard. To do this on a French reed raises the pitch considerably by limiting blade travel and hence internal volume.

So if one reduces pressure but maintains firm lip muscles, the sound will open, the pitch will lower, but with control there will be no "flabbiness" to the sound, which I believe is caused by quick and unintended pitch variations--wobbliness.

Another way to achieve this is by playing further up the mpc. This allows one to maintain a firm embouchure and play down on the pitch, at the expense of some fine dynamic control, which anyway is basically useless in jazz, as distinct from classical playing IME.

In fact I had to really relearn my embouchure when I got a nino. The upper notes just won't come out reliably without biting way up on the mpc. It's interesting to play used mpcs with teeth grooves--you really get a sense of how different the teeth positions of different players are.
 
I guess you have to have taught beginning clarinet and saxophone players for 27 years to truly understand what a "flabby tone" sounds like with an embouchure that is too loose. :)

This may sound crazy, but I am beginning to entertain the notion that a portion of the "equivalent" or "effective" volume of the mouthpiece is added by the space in front of the reed inside the player's oral cavity.

In the mouthpiece "effective volume" study I have attached, I found that the "effective volume" of my Rousseau 4R alto mouthpiece to be 9.13 ml, whereas its measured physical volume was 6.87 ml leaving a difference of 2.26 ml which is close to 33% of the measured physical volume.

I have taken another step and calculated the volume added by the "wedge" between the reed and the facing of the 4R mouthpiece which comes out at .28 ml which is not even close to the 2.26 ml difference in my study. That "wedge" volume was based on the beating of the reed going toward the mouthpiece face and back to its starting position. Even if the beating reed goes farther out than its resting position due to its elasticity, that would only double the volume at most bringing it to about .56 ml at the very most. That is still far short of the 2.26 ml difference.

The idea that the oral cavity becomes part of the mouthpiece "effective volume" would answer how a jazz player can "open up" and play lower on the mouthpiece pitch without overly relaxing the embouchre, and it would be an elegant answer to what makes up the difference between the geometric volume and the effective volume of a mouthpiece under playing conditions.

 

Attachments

  • Mouthpiece equivalent volume study revised.pdf
    37.6 KB · Views: 355
Effective volume must include the volume between the mpc facing and the reed at its furthest point of travel, but I absolutely see how any compliance outside the reed travel could be included. Don't forget that there are mechanical reed effects to be factored in as well. I'm curious as well how you figured out the theoretical missing volume.
 
The attached paper uses the method described below to calculate the "effective volume" of a mouthpiece under playing conditions. A more complete explanation of the process is described by Benade FMA p.466

1. The mouthpiece is attached to a cylindrical pipe of a known resonance frequency.
2. The mouthpiece is blown which produces a frequency lower than that of the natural frequency of the pipe alone.
3. That frequency is used to determine length of an imaginary pipe whose natural resonant frequency would match that played frequency.*
4. The volume of each pipe is calculated and the difference in the volumes represents the volume added by the mouthpiece ie. its "effective volume".

* This includes an adjustment for "end correction" of the blown pipe which does not exist when determining the natural resonant frequency of the pipe.
 
But how did you calculate the volume of the missing conic apex? And given that we know now that oral cavity can have an effect on pitch, as well as lip tension, how did you control for those factors as they would affect pitch of mpc + tube?
 
Sorry I misunderstood your question. The calculations to find the volume of the missing cone are in the attached study. The taper of the neck was used in the computation which I know is debatable and/or controversial. The same computation using the taper of the body tube produced a missing cone volume 51% larger! That is quite a discrepancy, I know. Perhaps the truth lies somewhere in between??

FWIW the calculated natural resonance of the length of the missing cone using the taper of the neck added to the measured length of the neck corresponds very closely with the played frequency of the mouthpiece and neck.

To answer the second question, the frequency of the cylindrical pipe played with the saxophone mouthpiece really "locked into" a centered pitch. I don't feel as if the variability of lip tension played a significant part because of this strong "locking" of the reed's vibrations with the resonance of the tube that takes place.
 

Attachments

  • Missing Cone Volume Study-Revised.pdf
    19.6 KB · Views: 477
Tube impedances in a thin cylindrical pipe of fair length are probably quite strong, and if one doesn't endeavor to change them with the oral cavity are this probably reliable, is my guess.

The neck is a bad measure of missing cone volume for several reasons. First, it is a different taper than the main body (which is again different from the bow/bell). Then, some necks change taper and geometry at the end, so measuring IDs at both ends may not be accurate at all. One really needs to average the taper along the length of neck/body at least I think, but even that does not take into account the effective alteration of taper as caused by tonehole chimneys. The one thing you can take to the bank is that when the mouthpiece volume equals that of the missing cone the thing will play in tune, though I know that is a tautology...
 
(...) otherwise the lower notes are horribly sharp. Maybe it's good enough for jazz. (...)

Not wanting to interfere in this sky-level exchanges, may I respectfully suggest that good jazz players don't necessarily play out of tune...
 
Last edited:
The neck is a bad measure of missing cone volume for several reasons. First, it is a different taper than the main body (which is again different from the bow/bell). Then, some necks change taper and geometry at the end, so measuring IDs at both ends may not be accurate at all. One really needs to average the taper along the length of neck/body at least I think, but even that does not take into account the effective alteration of taper as caused by tonehole chimneys. The one thing you can take to the bank is that when the mouthpiece volume equals that of the missing cone the thing will play in tune, though I know that is a tautology...

All of these points are well taken. I have undertaken the tedious project of measuring and computing the taper on my Selmer SBA which has been used in all of my studies. A look at the attached pdf file will show that indeed the tapers of each section of the saxophone are markedly different. The impossibility of making sense of the body and/or bell measurements is due to the fact that each tonehole makes the bore both larger and longer at that location and that those effects are cumulative. The neck alone is the only "pure conical tube" in the entire system.

As to the issue of the validity of the taper derived from measuring the ends of the neck and dividing by the length, I did the following measurements a while back which are shown in the 2nd pdf attachment. These calculations show identical results on the neck when computing the taper for the entire length (without the tenon) and by adding each sections taper as a percentage of the total.

The answer to the question of which is the most accurate way to calculate the length and volume of the missing cone has eluded me for several years. I believe that I made a breakthrough this week. I though, why not find the natural resonance of the saxophone without its mouthpiece and calculate that frequency's wavelength. Then find the wavelength of the lowest note Db3 concert and subtract the two to find the length the missing cone adds when playing the complete tube. That was simple to do. Then the task was to use that missing cone length and "reverse engineer" the taper required to produce that length using the diameter of the neck opening. The interesting result will be in the next installment.
 

Attachments

  • Saxophone slope and length study.pdf
    4.7 KB · Views: 363
  • Detailed SBA neck measurements revised.pdf
    14 KB · Views: 370
Back
Top Bottom