Chuckk Hubbard wrote:
> I know this should be the simplest thing, but it still befuddles me.
> Can anyone tell me why I get clicks between notes with this?  It 
> sounds to me like the clicks are at both the beginning and end.
> I apply the a-rate amplitude envelope after the filter, so that the 
> last filter input won't cut off suddenly when the note ends.  But 
> something still clicks.  Why?
>
> Is there any difference between basing the envelope on amplitude 1 and 
> the oscillators on iamp, vs. basing the envelope on amplitude iamp and 
> using 1 for the oscillators' amplitude?
> It seems that either way would cause a staircase effect, when 
> something of the range 0-1 is amplified to the range 0-32767, but I 
> don't completely understand floating point arithmetic.
>
> Or, perhaps I should use the envelope as the kamp argument of the 
> oscillators, and have a separate linear envelope that only tapers 
> sharply at the very end, to be applied after the filter?
> Any suggestions?  

This may be a bug in the rezzy opcode. I've experienced similar problems 
with it, but forgot to mention it. I've attached a csd file that 
illustrates the problem. By changing the value for ialgorithm, say 203, 
the click is gone. I can't say when the problem is there and when not, I 
didn't look into it enough.

As you can see in the comment, I solved the problem by adding a 
millisecond of silence at the begin (which I undid here to show the 
problem).

-- 
  Mark
  _________________________________________
  When you get lemons, you make lemonade.
  When you get hardware, you make software.

Mark Van Peteghem wrote:
> This may be a bug in the rezzy opcode. I've experienced similar 
> problems with it, but forgot to mention it. I've attached a csd file 
> that illustrates the problem. By changing the value for ialgorithm, 
> say 203, the click is gone. I can't say when the problem is there and 
> when not, I didn't look into it enough.
>
> As you can see in the comment, I solved the problem by adding a 
> millisecond of silence at the begin (which I undid here to show the 
> problem).

I should add that when I rendered this csd-file to a wavefile and opened 
it in a soundeditor, I saw that the second sample had the maximum value 
(so it was possibly clipped).

But I just tried it without the line with the balance opcode, and the 
click is gone. So the problem may have to do with balance, not with rezzy.

-- 
  Mark
  _________________________________________
  When you get lemons, you make lemonade.
  When you get hardware, you make software.