smooth amp from fiddle~


    Nov 07 2006 | 11:29 pm
    I'm trying to modulate the amplitude of a sound from a file (using *~ and sfplay~) from the amplitude analysis of another input using fiddle~. The result is in dB, so I use jasch's db2a~ to change it to a linear value. Since the variation of the amplitude is a bit jerky, I tried to use qlim and a line~ to smooth the amplitude variation. It's still jerky... Does anybody has an idea of a more efficient way to do that?

    • Nov 08 2006 | 12:19 am
      try [slide~]
      Quote: fakegolfer wrote on Tue, 07 November 2006 15:29
      ----------------------------------------------------
      > I'm trying to modulate the amplitude of a sound from a file (using *~ and sfplay~) from the amplitude analysis of another input using fiddle~. The result is in dB, so I use jasch's db2a~ to change it to a linear value. Since the variation of the amplitude is a bit jerky, I tried to use qlim and a line~ to smooth the amplitude variation. It's still jerky... Does anybody has an idea of a more efficient way to do that?
      ----------------------------------------------------
    • Nov 08 2006 | 12:45 am
      The slide~ helps to smooth the variation, but it often clips... I'm gonna keep working on it.
    • Nov 08 2006 | 1:43 am
      I would use dbtoa~ and rampsmooth~
      Also, doesn't fiddle~ have another idea about dB values than standard
      Max?, I think it neds a subtraction of 96 or something
      But then again, why not use peakamp~ instead of fiddle~ for the
      amplitude analysis?
      /J
      07/11/06, kl. 0:29 +0100 , skrev Pm B:
      >
      >I'm trying to modulate the amplitude of a sound from a file (using *~
      >and sfplay~) from the amplitude analysis of another input using fiddle~.
      >The result is in dB, so I use jasch's db2a~ to change it to a linear
      >value. Since the variation of the amplitude is a bit jerky, I tried to
      >use qlim and a line~ to smooth the amplitude variation. It's still
      >jerky... Does anybody has an idea of a more efficient way to do that?
      >--
      >Pm B
      >