<<Another approach I have naively used when faced with this problem will
probably make the professionals here shudder, but here goes:
- normalize
- identify the peaks, and bring them down in volume manually by working
with selections
- repeat until it sounds right (or wrong, in which case back up one
step and stop :-))
Apart from the risk of getting jumps in the sound through choosing the
selections sloppily (which did not seem to happen) I guess that repeated
processing of material in this way (even just for volume) is going to
result in some loss of quality?>>
This is compression, done manually rather than with a plugin. And if you're=
going to go with that approach I would strongly recommend doing your gain =
reduction via writing volume automation, since the volume shifts will be th=
e result of a smooth & controllable volume ramp rather than discrete jumps.=
Automation will allow much more subtle & undetectable transitions into & o=
ut of gain reduction. But you can only pull volume so much with any method =
before it becomes audible. It's very dependent on the source, but 5 or 6 dB=
is about as far as you can go before dynamic artifacts start becoming obvi=
ous.
Scott Fraser
|