 The Clarinet BBoard
|
Author: L. Omar Henderson
Date: 2000-04-12 16:44
Is there a difference between reeds - we would all say yes but how to quantitate this difference??? How important is "the" reed to our tone??
At Mark's urging I will share some preliminary data and ask for your critique and help to refine the process further.
I was a little awe-struck by Marks authoratative information on all the "partials" that make up the various notes but with the help of some knowledgable friends and some neat computer hardware and software decided to try to define some relationships in the spectrum of frequency and amplitude of clarinet sounds. It "appears" (maybe real or merely virtual on my part) from the spectral analysis that I have done so far that it should not be all that difficult to quantitate the differences that can be attributed to individual reeds within a box, box to box, and manufacturer to manufacturer. (Some interesting possibilities and quality control factors here). There are of course many variables in addition to the reed, some of which are: player, mouthpiece, ligature, barrel, clarinet, environmental conditions, etc, etc. There are also differences within a reed due to hydration, reed placement and ligature placement. (Too much data to present - and bore you with - here today. ** You can help me by elucidating some more areas to study! ** Anyway, in the upper register (with the octive key depressed), the repeatable sequence of frequencies in my patterns are: primary tone, octive, 1/5, 1/4, 1/3 and several multiples of 1/3's or 1/12's. There are only 4-6 major peaks (above 2db sampling at 44K and 22K Hz with 14 Hz averaging - twice audible decernable range - and some other sophisticated manipulations and gating). Increasing the sensitivity of course increases the numbers of partials but with moderate sensitivity and amplitude gates I was able to quantitate some interesting aspects of tone differences due to various factors. In the lower register the sequence is: primary tone, no (or only slight) octive, 1/5, no (or only slight) 1/4, 1/3, and several multiples of 1/3's or 1/12's. There are up to 10 major peaks that make up these notes.
Using the same clarinet:
The data so far collected indicate consistent peaks day to day with some minor differences in amplitude - (my statistician can help me here on variability measures), some minor (differences in amplitudes but no new major peaks some partials tranpositions) with changing mouthpieces (custom rubber, tip opening, commercial rubber, hard plastic) or several ligatures. Significant differences between reeds in the numbers and amplitudes of the major peaks (harmonic, 1/5, 1/4, and 1/3-1/12's). Differences between reed manufacturer's reeds, primarily in the numbers and amplitudes of peaks. Have not yet looked at differences within a reed over playing time or different reed strengths (this is a study in progress!) Differences between players on same clarinet - primarily amplitude of peaks, some in partials (can we get a spectral signature of our favorite player --???). Differences between clarinets - yet to be done (could we characterize the tonal qualities of clarinets we are thinking of buying --??). Altering reed has some significant and interesting changes (more to come). This is just a preliminary report and I would appreciate your input (good, bad or ulgly will do), and if you are interested enough there will be more forthcoming. The human ear is a exquisite spectral analysis instrument and I may of course be missing some very important factors affecting tonal qualities. Also, I can not at this point give you many particulars or specifications because of legal and patent activities in progress but I thought that it might be a ray of sunshine in the "REED FOG" we all live with daily.
|
|
Reply To Message
|
|
Author: paul
Date: 2000-04-12 18:49
I used to do detailed aeronautical testing for a living. It was a long time ago in a land far, far away, but I remember some of the basic things that made the tests valuable to my customers. Control your variables, run against a tightly written script, do it the same way each and every time, control the human element as much as possible. Record absolutely everything you can. Save the raw data and separate it from any analysis or calculations you want to do in the data reduction stage. Record your findings. Then, retest all interesting findings to remove all possible mistakes. With those things in mind, I have a few comments for you.
You have way too many variables to handle at one time, IMHO. You need to develop a test plan. Have your test plan reviewed by peers and pros alike. Once the test plan is approved, use it and stick to the instructions in the written plan absolutely to the letter. If your want to find out about reed behavior, then focus in on the reed and only the reed. The skill of the player makes a huge difference, so use one middle of the road player and don't permit any significant increase or decrease of playing skills during the tests. A novice won't be good enough and a pro may be too good. Perhaps a good amateur or a good college level player may be sufficiently standard enough to use for such a test. Use a temperature and humidity controlled environment for the recordings. Use a sound proofed room, too. Make sure you keep the mouthpiece, ligature, and the instrument exactly the same throughout your tests. Write down each and every batch number from your boxes of reeds. Test every reed in that box to establish as much of a statistical norm for that entire box as possible. Don't doctor up any of the reeds. Make sure the reeds are the same vintage, or grade reeds within say a couple of month's vintage against each other and not against the entire pool of reeds. Ditto for different brands of reeds and different classes and strengths (i.e. Vandoren regular cut #3 versus V-12 #3). Break in each reed in exactly the same way. Warm up and play each reed exactly the same amount of time. Use exactly the same warm-up drill for each reed. Swab out for a reasonably dry horn for each test. Don't let the horn (or mouthpiece) cool off too much in between tests. Check for exact reed placement on the mouthpiece. Have the player be well warmed up and well rested so embouchure and air support are not problems. Play exactly the same notes for exactly the same amount of time for each reed. Record the player both for the spectra test and on video to ensure the player does exactly the same things for each reed. Check for the angle of the horn the player uses for each note and for each situation, the player's standing or sitting stance for all situations, the distance to the spectra machine's microphone from a preselected part of the clarinet (bell, midjoint, barrel, mouthpiece), and more. Standardize, standardize, standardize.
|
|
Reply To Message
|
|
Author: L. Omar Henderson
Date: 2000-04-12 19:52
Thanks! I am sure there will be many other good suggestions and just wanted y'll (from here in Atlanta) to know that I appreciate your help. As my "day job" I run a scientific quality control program for a large Government agency so your suggestions are well aligned with my daily mantra. You are right that there are too many uncontrolled variables to look at in a given experiment even using an ANOVA (analysis of variance analysis) approach, but in these preliminary experiments I wanted to test(Q&D) just how much variability there might be associated with some common "improve your tone" myths that we hear every day. I hope that no one takes these as etched in stone facts quite yet, but raise the possibility of bringing some science to the wonderful art of clarinet playing. I am excited by the mere possibility of testing out some of these factors in a quantitative way.
|
|
Reply To Message
|
|
Author: Meri
Date: 2000-04-13 00:17
There are indeed differences between good/bad reeds. A study was done a few years ago; it's in one of the back issues of the ICA publication; not sure which one though.
Meri
|
|
Reply To Message
|
|
Author: Bob Gardner
Date: 2000-04-13 03:32
I have bdeen playing with a Rico V2 1/2 and tonight i put in a Vandoren #2 and i swear I have a new clarinet. It's sounds better, plays easier and takes less air. Maybe i just got lucky and picked a great reed. Anyhow I'm going to stick with the Vandoren for a while.
This entire process was done with no scientific research or grants. However i would be willing to accept any grant money the goverment is willing to put up.
|
|
Reply To Message
|
|
The Clarinet Pages
|
 |
.jpg){kind=small}
.jpg){kind=small}
.jpg){kind=small}
.jpg){kind=small}
.jpg){kind=small}
.jpg){kind=small}
