Thursday, March 30, 2017

Does Mouthpiece Material Matter? Part 2.

In Part 1, we looked at the idea that a certain "perfect" hardness of vulcanized rubber effects how a mouthpiece sounds.  With the help of a mouthpiece expert, the preliminary conclusion was that it does not.  We will now look at the published statements of another materials expert on whether mouthpiece material matters.

I have used Mr. Otto Link as an expert before.  We first find him working for Alexandre Selmer in New York before WWI.  We then find him working at an instrument repair shop with Frank Meyer (later of Meyer Brothers mouthpieces) in 1923.  Then he opens his own musical repair shop and (in small print) offers mouthpiece refacing.  Finally, years after Goldbeck had received a patent for a metal mouthpiece (in 1920), Mr. Link begins producing what appears to be an identical metal mouthpiece under his own name.  

But there is a problem.  The production cost of Mr. Link's metal mouthpiece is fairly high.  Competitor's mouthpieces cast out of hard rubber are undercutting his sales.  So he begins production of an hard rubber piece.  To quote a 1940 Link sales pamplet: "By embodying the famous "LINK" TONE CHAMBER heretofore found only in our Metal Mouthpieces we have created a very popular Hard Rubber Mouthpiece which exactly fills the needs of of those musicians who prefer a hard rubber mouthpiece." Great, Mr. Link created an identical mouthpiece made of rubber.  

But don't think that Otto Link was going to ignore the whims of those who desire mystical vibrations from materials.  His advertising goes on to state that his metal mouthpiece was made of "bell metal."  Presumably, the idea was that bells resonate, ergo, a woodwind mouthpiece made of bell metal (a fairly specific bronze alloy) could ring like a bell or something.  Wait a minute.  How can Mr. Link possibly make a rubber mouthpiece to mimic his bell metal mouthpiece? What is the sound of a rubber bell?  None of this makes any sense.

Not to worry.  We don't need to make any sense if we can come up with some convincing gibberish.  Otto Link simply ignored the inherent contradiction in metal vs. rubber vibration claims.  He realized that when promoting a product to musicians both material science and common sense can be ignored.  Superficial hype is all that is required.  Simply make an unfounded allegation (e.g., Link metal mouthpieces are made of "bell metal" with the understanding that the term is inexact and nobody will ever test.)  Let others imagine what bell metal does for a mouthpiece.  Likewise, just claim that Link hard rubber is "eburnated" and walk away.  Remember, musician's don't analyze matter, they just repeat claims that material matters. 

Let's assume that we are not manufacturing or selling mouthpieces and examine a "bell metal matters" claim.  Common bell metal (a bronze that is 78% copper and 22% tin by mass) has a nicer ring to it than common brass (copper and 32-39% zinc) when formed into the proper shape of a bell.  The shape of the bell is all important in the clarity, length of resonance, and pitch of the bell.  Mr. Morgan, in our first expert's article referenced in Part 1, alludes to this when he states "Obviously an alloy with less copper content will be harder and more dense, with a greater capacity to resonate. If you have ever compared the sounds of a fine Zildjian cymbal with some of the lesser quality ones available, the difference is heard immediately."

But wait a minute.  Brass always has less copper content than bell metal (61% to 78%).  Therefore, according to our first expert's theory, brass would be harder, more dense, and has a greater capacity to resonate, all of which is completely wrong.  I think we have further evidence that one of our materials experts is not an expert.   

Let's unpack the ultimate claim that a brass and a bronze mouthpiece would "sound" different by conducting a test. I bent a Zildjian cymbal into the shape of a saxophone mouthpiece.  It made a poor mouthpiece and was no longer a good cymbal.  I then bent it into the shape of a bell.  It appears that the shape was more important than the material for either a cymbol, a bell, or a saxophone mouthiece.  Yes, the material of a cymbal or bell matters, but only after it is formed into the proper shape.  We now have evidence that it is primarily mouthpiece shape that effects the sound and not whether it is 78% copper or 76% copper.

Of course, there is no sound at all unless you hit the mangled cymbal with the drumstick.  Here, we can get philosophical, since we never actually strike a bell metal saxophone mouthpiece.  What is the sound of an unrung bell, grasshopper?  If you want to get really hippie dippy, watch the YouTube video linked above all the way through.  (If you can hear the difference between a mouthpiece alloy with 78% copper and one with 75% copper, this video is for you.)  It appears that bell metal doesn't matter on a mouthpiece.  Or more precisely, bell metal only matters if formed in the shape of a bell or cymbal and struck with a hard object.  

Even when bell metal is the proper shape to resonate, it generally has only one responsive pitch (some metals, such as Monel, have additional harmonic responses for a given shape).   A "bell metal" mouthpiece does not have what anybody would consider a pitch.  Trust me, no need to whack your Super Tone Master with a drumstick.  But if a Link metal mouthpiece could resonate at a certain frequency, as claimed, what would that best pitch be?  Neither Mr. Link or Mr. Morgan tell us, so we will have to choose one.

How about Bb for tenor and Eb for alto?  Can you find them in this picture?  It is very important.  Not.  What is important to understand from this picture is that metals, copper, bronze, silver, etc., don't resonate at a certain frequency.   Various metal alloys, when a certain size and shape resonate at a certain frequency.  Have you noticed that none of these bells are the size and shape of a saxophone mouthpiece?

In the last web site linked above, the exact dimensions of a middle C tuning fork made of steel (the common material for tuning forks) was recreated using different metal alloys.  Not surprising, each alloy has a different pitch.  Also not surprisingly, most tuning forks had only one pitch per alloy.  And also not surprisingly, each metal alloy resonated only when struck with a hard object.  So a "bell metal" mouthpiece could conceivably produce a single short duration "pitch," depending on its shape, but only when struck with a clapper (or drumstick).

Let's take a closer look at a famous Link "bell metal" Super Tone Master.  But first, here's a brass bell.  Yes, they make bells of both brass and bronze.  Mine's just brass.  You can see that, although brass isn't supposed to be magnetic, it responds to a rare earth magnet and slight magnetism is one way to differentiate between brass and bronze.

Next, I'm going to use the magnet on bronze.  This is a high quality bronze wood screw, sometimes called a boat screw because of it's anti-corrosion properties.  

No reaction at all.  Next is an old Otto Link.

What?  That's not supposed to happen.  Let's take a closer look.  In the first video, you can see that the brass bell is brass colored.  In the second video, you can see that the bronze screw is bronze colored.  That's the other way to tell if something is brass or bronze.  Brass and bronze alloys have a different color and get a different patina with age.  Brass is more golden, bronze more reddish.  Here is a look at the metal used on an old Otto Link (balanced on the bronze screw).

Now what?  Metal Otto Links are made of two cast halves brazed together.  And you can see that this one is made of two different alloys.  Which one is bell metal?  I'm going with the right side.  That's the side that always sounded best.  Just kidding.  This is further proof that material doesn't matter and Mr. Link knew this.  Brass, bronze, whatever is most convenient to cast and machine at a reasonable cost. 

Apparently, different Link casting batches used different alloys and nobody could tell when playing them.  Just claim that it's made of "bell metal" and let the suckers consumers make up the rest of the story.  To his credit, Mr. Link never proceeded further with claims about the magical properties of "bell metal."  I'm sure that he knew that there is no pitch or resonance produced without the shape of a bell, and being struck with a clapper, and being a particular size.  And he knew that it was possible to make a great mouthpiece out of either brass, bronze, or hard rubber.  So much for materials matter. 

Mr. Link did not, however, do any thing to stop the myths surrounding his "bell metal" and "eburnated ebonite."  Our other expert, Mr. Morgan, treads where even Otto Link, the heretofore King of Hype, dared not go when discussing metal mouthpieces.  "Metal mouthpieces are not subject to the same tonal changes we note in hard rubber, since the average thickness of material used can do nothing but act as a damping effect on the reeds (sic) vibration."  

Let me understand this; if you want a damping effect, use metal not rubber.  Forget about rubber baby buggy bumpers.  Metal dampens more than rubber?  That contradicts both my personal experience and the understanding of the manufacturers of rubber products.  A rubber bell resonates longer than a bronze bell?  In what world?

Next, Mr. Morgan claims that players want to use metal mouthpieces "because a long list of prestigious players used that type of mouthpiece."  But these unwitting rubes don't realize that "they have only heard a performance using all manner of electronic enhancement and simplification, either live or on CDs/records/tapes.  Therefore, what we hear may be from what the player, mouthpiece, and instrument actually sound like, due to the whims of the sound engineer, etc., altering the true sound."  

Okay, that's a lot to unpack.  First, we can assume that the player has heard himself play when not using electronic amplification and prefers the metal piece.  Second, we can assume that sound engineers can also alter the sound of a Shore 85 "perfect" ebonite piece to make it sound as good as a metal mouthpiece.  Third, a lot of performances aren't miked and the patrons don't turn away when somebody plays a metal mouthpiece in a completely acoustic setting.  So much for that materials matter claim.

Finally, with yet another claim that metal mouthpieces "naturally provide more damping of the reed than hard rubber," Mr. Morgan states that metal mouthpieces can actually work if "the basic design be the result of much acoustical and aerodynamic study."  Aerodynamic study, not this bogus claim again!  

At what speed does aerodynamics come into play on a woodwind mouthpiece?  
A.  30 miles per hour.  
B.  150 mph.  
C.  Never.

But my favorite statement is "Experiments show that a mouthpiece properly designed and made of good hard rubber will produce 30% more sound overall and play with a more centered sound."  What is 30% more sound?  More volume? More overtones?  More inharmonics?  50% more highs and 20% less lows?  And "centered" around what?  Is any of this good?  

And why do other experiments, experience, and common sense show that metal mouthpieces are appropriate and appreciated?  The listening audience, which is responsible for creating the "long list of prestigious players," apparently groove on metal pieces.  What good is hard rubber's claimed "more centered sound" if fewer people want to listen?  Clearly, we have established that material matters . . . it can create unfounded and unsubstantiated prejudice.

Finally, Mr. Morgan makes a specific mention of the "distinct quality of silver mouthpieces.  One metal which does have the capacity, given the correct alloy, to produce a distinct clarity of sound is silver . . . which resonates at frequencies conducive to the production of a richness of sound not present in most other metal mouthpieces."  And what is claimed to be the "resonant" silver alloy?  Sterling silver (92.5% pure silver) of course.  Why is sterling silver so good?  Maybe because sterling silver was once used for money.  That sounds expensive, so it must be good.*

Sterling silver is always 92.5% silver, but the rest can be an alloy of either copper, zinc, tin, etc. and it is still called sterling silver. We know from the tuning fork website linked above that the particular metal used to alloy 92.5% silver would effect the resonance.  As we have seen with copper alloys, an alloy with zinc produces brass and using tin produces bronze, each with distinct resonance and only one called "bell metal."  

Apparently Mr. Morgan didn't know about the large variety of alloys that are called sterling silver.  Sterling is uniform only in the percentage of silver but, because of the various alloys, it is not uniform in the variety of frequencies at which it will resonate (when formed into a shape that can resonate and struck with a hard object).  So much for the sterling silver mouthpiece nonsense. Darn science (and logic and experience and common sense).  

So what have we learned from all of this?  Material matters depending on what type of mouthpiece you are selling.  If you are marketing metal pieces, they ring like a bell.  If you are marketing rubber mouthpieces, your special hard rubber recipe makes them resonate like the wing beats of an angel.  If you are marketing wooden mouthpieces, you will have to make something up.  On that, Mr. Link, Mr. Morgan, and I agree.  

Mr. Morgan may be on to something when he states that not the material, but the average thickness of material, can effect the way a mouthpiece plays.  I'm thinking about writing a future blog about that concept and will provide a link if I get around to it.

Here is the link to the "material thickness" blog.

*  Silver as a magical material for bells became famous after the Christmas song "Silver Bells" was recorded by Bing Crosby in 1950.  Although the title has a nice "ring" to it, silver bells are not common except as charm bracelet bangles.  Despite the perceived "richness" of sterling silver, bells from silver are not popular because of the relatively high pitched "tinkle" that they make in comparison to brass and bronze.  A strange coincidence as the song's author first called it "Tinkle Bells" until his wife explained her understanding of what the word "tinkle" means to women. 

Wednesday, March 29, 2017

Does Mouthpiece Material Matter? Part 1.

Yes.  Absolutely.  But probably not in the way you are thinking.  It makes a difference in how a mouthpiece is manufactured.  It makes a difference in the weight of the mouthpiece.  It makes a difference on the look and feel of the mouthpiece.  It makes a difference in the production cost of the mouthpiece.  It makes a difference in the flammability of the mouthpiece.  It makes a difference in the color of the mouthpiece.  It might even make a difference in the smell of the mouthpiece.  

I know what you are thinking.  Does it make any difference in the sound produced by the mouthpiece?  Not enough to detect or worry about.  But given that most players are looking for that special advantage provided by some mystical property of their instrumentation, we should look more closely at whether mouthpiece material matters "acoustically."  For that examination, we can use expert testimony, existing data, test results, and common sense.  If you don't have those things, don't worry, you can use mine.

With expert testimony, we need to be careful because there are many unsubstantiated claims out there.  Assertions like "plastic mouthpieces make you sound like a duck because duck calls are made of plastic."  We would have to resort to our other sources (existing data, test results, and common sense) to determine whether to believe that expert's statement.

I'm going to rely on two people as my material experts.  Mr. Otto Link (who I have used before as an expert) and Mr. Ralph Morgan.  I'm going to start with an article authored by Ralph Morgan and printed in the Saxophone Journal many years ago (Does the material used make any difference in how mouthpieces play?).  Right on point and the article is a wealth of assertions.  Let's see if we can answer that question with the evidence presented in the article.

The article begins by talking about wood mouthpieces, but the main take away is that wood mouthpieces are difficult to manufacture because of shrinking, cracking, etc.  Wood was abandoned when other materials became available.  So, materials that are more durable than wood and more uniform to machine than wood are preferable.  Material matters to the manufacturing process.  The article hints at the problem of there being hundreds of types of wood and that any claims of the effect of the vast variety of wooden mouthpieces would be silly.  Hint: don't put a mouthpiece made of dogwood in your case with a mouthpiece made of pussy willow; they will fight.  Things like that. 

Most modern players are not concerned with wooden mouthpieces, so we will begin with the article's discussion of hard rubber mouthpieces.  Mr. Morgan states: "Hard rubber became “the thing to use” after Harvey Firestone discovered how to vulcanize, or harden, natural gum rubber.  This happened none too soon since the need for clarinets and saxophones grew rapidly in the late 1800s."

Let's look at that statement a moment.   The U.S. Patent Office issued patent 3633 for the vulcanization of rubber on June 15, 1844.  Harvey Firestone was born on December 20, 1868.  Do you see a problem?  Harvey Firestone missed the first several decades of ebonite development because he wasn't born yet.  Harvey Firestone did get a patent related to vulcanized rubber.  His patent was for tires on horse drawn buggies, and later, automobiles.  Firestone was a contemporary of Henry Ford and became a millionaire as a result of pneumatic tires.  But a claim that Firestone "discovered how to vulcanize?"  It was Charles Goodyear who patented the vulcanization of latex rubber.  Ummmm, I'm going to be skeptical on all further testimony by this materials expert, okay?

Mr. Morgan then talks about the hardness of rubber measured on the Shore D scale.  Albert Shore developed a method to measure hardness of vulcanized rubber in the 1920s (long after ebonite had been successfully used for woodwind mouthpieces).  It is basically a spring loaded pin and the measurement is how far the pin deforms the test material without penetration.  Two different scales exist, Shore A and Shore D.  Shore A is for softer rubber and Shore D for harder.  
You can see the little pin at the bottom of this Shore A durometer.  The gauge is pressed flat against the material to be tested.

Ebonite is usually tested with a Shore D meter, which uses a sharper pin and 5 times more force than a Shore A durometer.   Ebonite is defined as vulcanized rubber above 70 on the Shore D scale.  Some of this is defined in ASTM D2240 if you are interested in the minute details.  It should be noted that ASTM D2240 states "No simple relationship exists between indentation hardness determined by this test method and any fundamental property of the material tested."  In other words, according to the American Society for Testing and Materials, rubber hardness means diddly about its acoustical properties.  I'm going to start from that point of view.

There are problems with Shore test accuracy, of course.  In effect, you are trying to measure a particular rubber's "squishiness," which (as you would imagine) is a rather "elastic" concept.   Manufacturers of the Shore meters usually state that because of test sample temperatures, operator error, etc., the meters are only accurate within + or - 5.  So a sample that reads 90 could later test as either 86 or 93 the next moment or the next day.  Here is a YouTube video of using a durometer.  The pin tends to slowly extend if the durometer is held in place, so even the quickness of the reading effects the result.  

Another influence on the test results is the size of the test "puck."  ASTM D 2240 defines the size and thickness required of a test piece (the test must be conducted on a flat surface greater than 6mm thick and further than 12mm from the edge).  Because of this size requirement, a Shore durometer can't be used on an actual woodwind mouthpiece.  In fact, it can't be used on a cylinder.  Here is a video supposedly showing how to get a correct reading.  You can see that the reading varies because you can't get an accurate reading on a cylinder.

Durometer measurements are important for commercial and industrial uses of rubber and plastic, although whether it tests as Shore D 85 one day and Shore D 82 the next doesn't matter.  The test would be performed on a sample "puck" and the material is deemed fit for it's intended purpose when properly molded, cooked and cured.  So the manufacturing process also effects the Shore D number.  If the part manufactured is small and cylindrical, you have to simply guess that you have come close to the intended Shore D number based on the rubber formula and manufacturing process.  It clearly is not an exact science.

But wait a minute.  What if we are not interested in the the durometer reading for industrial use?  What if we don't want to acknowledge the limitations of Shore D measurements?  What if we are only interested in the mystical musical properties of vulcanized rubber?  What if we have a theory that a Shore 85 ebonite mouthpiece produces a luscious harmonic melody and a Shore 82 mouthpiece makes a dull flatulent moan?

A "perfect Shore D mouthpiece" is basically what is claimed in the Saxophone Journal article.  How do we examine our expert's testimony when he makes claims beyond known material science?  Super accurate Shore D numbers aren't available, direct testing of a mouthpiece isn't available, yet it is alleged to make a huge difference in the sound of a mouthpiece.  

There are inherent problems with claiming that a specific Shore D test number is required for an ebonite mouthpiece.   At approximately 80C, ebonite undergoes a thermoplastic transition, i.e., it approaches it's "yield" temperature and becomes plastic again.  (At approximately 200C, ebonite becomes liquid).  The Shore D number for all rubber products drops with an increase in temperature.  It is the nature of the material.  Race car driver's understand this and use a specific Shore D tire based on track conditions, knowing that the tires will soften when warm.  Even hip skate boarders know this one.  It is why the garden hose feels stiff on a cold day and limp on a warm day.

We can't predict exactly how much a hard rubber mouthpiece softens with warmth (and we can't measure a mouthpiece), but we know that it drops.  ASTM D2240 says that all Shore D testing must be performed at temperatures below 30C.  My durometer states that it can't be used at temperatures above 30C.  Based on the stated accuracy of my meter, one can assume that temperatures above 30 would cause inaccuracies even beyond the meter's claimed accuracy of + or - 5.

My durometer instructions and the American Society for Testing and Materials also state that relative humidity must be below 80%.  High relative humidity causes rubber, including ebonite, to become more flexible.  Maybe not enough so that you or I can tell, but enough for an inaccurate reading on a Shore D meter.  Again, it could result in an error outside of the meter's + or - 5 accuracy.

Commercially available tables show that 10 degree changes in temperature would change the Shore reading of some rubber products by about 5.  I couldn't find any tables on the direct effect of humidity.  The initial Shore D rating, as well as different additives, tend to effect the amount that heat and humidity will reduce the Shore D number.  Basically, there is no way of knowing how much the heat and humidity changes will "soften" a particular mouthpiece that we have guesstimated to be a "perfect" Shore D 85. 

Since we can't actually perform the Shore test on a mouthpiece, let's do some hypothetical testing.  We will start with an ebonite mouthpiece that we believe is a perfect Shore 85 (based on a sample of the material prior to molding, cooking, and curing).  First, we put the ebonite mouthpiece in our mouth.  Then we blow moist, warm air through it for a few minutes.  We will have now raised the temperature of the ebonite to 32C (90F).  It has condensation on the inside, indicating 100% humidity.  Our hypothetical test has now created the conditions scientifically known to increase the softness of the rubber by greater than 5 on the Shore D scale.  In fact, both temperature and humidity are now in excess of the requirements for an accurate Shore test.  In case you haven't noticed, we create these conditions every time we play a hard rubber mouthpiece.

We don't need to know exactly how much blowing through an ebonite mouthpiece would change the Shore number.  We do know that our durometer (which was only accurate within 5 begin with) is no longer accurate.  Maybe the Shore number has changed by 6?  We can't tell because the ASTM D2240 test can't be used on a warm, moist mouthpiece.  So what is the Shore D number after playing an ebonite mouthpiece that started out as the "perfect Shore 85"?  82?  Less?  We wouldn't know because even if it dropped to 74 it is still ebonite, i.e., it is still hard rubber that holds it's shape.  It appears that it is the mouthpiece shape, not some vague and changing Shore D number, that determines how ebonite plays.

And we don't need to know the Shore D number of a warm, moist ebonite mouthpiece.  At this point, we can rely on our experience and common sense.  Have you ever noticed that getting the mouthpiece warm and moist makes a perfect hard rubber mouthpiece unplayable after a few minutes?  No.  Have you ever heard the difference when an ebonite mouthpieces undergoes this dramatic change every time it is played?  No.  Nobody has ever noticed this because a Shore D fluctuation of >5 doesn't matter.  Conclusion: the range of Shore D fluctuation in an ebonite mouthpiece does not effect acoustics.

Even Mr. Morgan agrees that Shore D doesn't matter for mouthpieces when he goes on to talk about metal mouthpieces.  Shore D doesn't matter on metal pieces "since the average thickness of material used can do nothing but act with a damping effect on the reeds (sic?) vibration."  I have to admit, I have no idea what that statement means.  At least we can agree that Shore D claims are nonsense.

Okay, enough, enough.  Let's go on to our second expert, Mr. Otto Link.  I have used Mr. Link as an expert in another blog.  Since this blog is getting long, I'll start another one about Otto Link agreeing with Mr. Morgan that material doesn't matter.  I'll "Link" to it when it is written.