Compression Effects

[Michael Maloney]

>
>To get more on-topic, has anybody here built a multi-band compressor?
Yep - or more accurately, I cuurently build a two-way 24 dB/oct
comp/limiter/crossover.  Roughly the same thing, less an output mixer to
recombine the band outputs.  In my case, I use air as a summing amp.

Definitely a huge improvement over single band compressors for full mix
work - I also have a four-band model designed - I just haven't had the need
to build it yet [I did these for a speaker manufacturer to build
self-powered biamped monitors, and they haven't gone to production on the
four-way boxes yet.]

Simply side-chaining isn't the same thing - in side-chaining you are
manipulating the sensitivity of the detector circuit with respect to
spectral content, but you are still controlling the gain of the
full-bandwidth signal. As an example, a de-esser does not reduce high
frequency content.  It adjusts the overall gain of the signal, "leveling"
it in proportion to its high frequency content.

Beyond the simple ability to compress bands based on their individual power
levels, it is also advantageous to have differing attack/release times on
the different bands, especially when using high compression ratios.  As
someone else pointed out, with fast attack times you run a good chance of
creating undesirable IMD at low frequencies, and with slow attack times you
can miss the better part of high freq. peaks.  With multiband, you can
optimize a/r times based on the frequency range.

Mike Maloney

PS - What happened to the guy with the anagram software? With all the bad
vibes surrounding Barbour/Doepfer, this place could stand a bit of harmless
silliness.