This is why it makes sense to use dam. It used an averaged power calculation to
decided if the volume has to be compressed or expanded. Also, by ensuring the
attack is not to fast w.r.t. your signal's content, you do not degrade its
harmonic content.

Marc.

Mike Berry wrote:

> It should be noted that doing this (waveshaping) does cause distortion in the
> signal.  If you run a full-range sine wave through this table, you do not get
> a sine wave out the other end.  The distortion may be pleasant (more pleasant
> than clipping!!!) and may be the desired result, but it is there.  So this is
> not the same as scaling your score to avoid clipping.
>

Marc Resibois wrote:
> This is why it makes sense to use dam. It used an averaged power calculation to
> decided if the volume has to be compressed or expanded. Also, by ensuring the
> attack is not to fast w.r.t. your signal's content, you do not degrade its
> harmonic content.

I want to throw in a caveat on this one too: if you run your whole mix
through a comp/limiter, it's going to sound like it's been through a
comp/limiter.  If you haven't used one before, your piece may sound
just fine until it's burned to CD, at which point you learn to hear
the artifacts.  They can be tricky to use.  (I've never had occasion to
try dam in particular.)

-- 
     Eli Brandt  |  eli+@cs.cmu.edu  |  http://www.cs.cmu.edu/~eli/