Re: [Harp-L] Sorry, it's combs again



  ----- Original Message ----- 
  From: Steve Baker 
  To: Vern Smith 
  Sent: Monday, September 22, 2008 1:55 PM
  Subject: Re: [Harp-L] Sorry, it's combs again


  > The test I'm suggesting is purely mechanical and involves no player. All that's necessary is to blow the same 
  > reed mechanically on various different combs and compare the resulting overtone and volume profiles. Surely 
  > you would agree that any measured difference under these circumstances is objective and has nothing to do 
  > with any preferences anyone may have?

  I thought that that everyone has agreed that listeners could not perceive differences in sound arising from different comb materials.  I conceded that maybe the player could perceive differences.  Your experiment cannot demonstrate that a player can perceive differences if there is no human player.

  I do not question that a spectrogram might possibly show differences among machine-blown harps.  However.....

  In 1997, I made spectrograms of machine-blown harps.  I found that the spectrograms of two harps having the same comb materials but different reedplates were just as different as two harps having different comb materials.  Every spectrogram was a little different from every other.  I could not correlate these differences to comb materials nor could others. I suppose that this could have been my experimental technique.

  If you plan to make spectrograms of machine-blown harps, based on my experience, I would suggest that you do the following:

  Make a very large number of spectrograms using permutations and combinations of th                          ese variables:
  - Several of each combination of materils and reedplates at different times.
  - Several of the same reedplates with different materials.
  - Several of the same materials with different reedplates.
  - Several combs of each material.

  Since the results will be statistical, they will not be valid without a large number of samples and without adequate controls.

  Identify the spectrograms only by code numbers.
  Ask your expert analyst to sort the unidentified, scrambled  set into groups by material....then name the material of each group.
  I predict that they cannot do it, but I could be wrong.

  The results should be interesting even if it does not demonstrate the ability of any player or listener to perceive material differences..

  Vern
  Visit my harmonica website www.Hands-Free-Chromatic.7p.com



This archive was generated by a fusion of Pipermail 0.09 (Mailman edition) and MHonArc 2.6.8.