[Stk] Bug in ADSR.setAllTimes decay rate?

[Zacko Belsch]

Howdy, Gary,

It was an interesting issue.  I'm pretty new to STK, having just taken the
chuck mooc.  I doubt I have the ear you folks do.  My statement about the
effect this would have on the sound of the instruments may not have been
warranted.  The effect does change the waveforms, though possibly
imperceptibly.

Further, my earlier statement that ADSR.setAllTimes calls would have to
have their decay times doubled to compensate for a fix isn't quite correct.
 This is true for Rhodey ("true" only being an appropriate term if the
exact waveform is desired).  But for other calls it depends on what the
sustain argument is.  Grepping for setAllTimes, I see Rhodey, Sitar,
TubeBell and Wurley would need doubled decay.  BandedWG, BeeThree,
BlowBotl, Bowed, Brass, FMVoices, Flute, HevyMetl, Moog and PercFlut, would
need more careful analysis, though it is be fairly simple to compute the
effective decay time the older code produces.  Some decays are longer than
expected and few are shorter.  In the case of BeeThree and HevyMetl no
decay would be expected on components 0,1 and 2, but there is decay. Maybe
only perceptible on long notes with no other effects(?).

It may sound like I'm making a mountain of a molehile.  Not my intent.  I
just want to document the technical effect of the fix, so it's known.

Bob Harris
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://ccrma-mail.stanford.edu/pipermail/stk/attachments/20140120/85dec990/attachment.html