pre-release 2006-07-02  alankila

    * Christopher O'Neill (Deltafire from ircnet's #amigaexotic, and one of the
      authors of milkytracker) tried to implement the blep approach. He made an
      excellent point that convinced me to separate the aging function from the
      function that adds new bleps. The function is now known as input_sample, and
      the aging function is called clock().

    * Christopher also found out the hard way a silly bug of mine: sampleclip
      routine had the comparison operators the wrong way. :-(


pre-release 2006-06-27  alankila

    * LED filter cutoff is at 3.2, not at 1.6. Duh. This was due to a bug in Butterworth
      filter computation routine, noticing as I reimplemented it from BLT.

    * LED filter on genuine Amiga appears to perform slightly suboptimally with about
      0.3 dB lowered resonance from theoretical. Perhaps component tolerances are at
      issue? Additionally, the filtering slope appears slightly different from
      expected.
