There are two primary reasons to use compression. One is to electronically ride the gain so as not to overload the electronics,causing distortion, on the way into your recording medium of choice, and two, to use the compression as an "effect".

In the old day's, there was a balancing act between the noise floor and the available dynamic range of a given piece of gear. A vocalist could have a dynamic range of 90db, a whisper to a scream. If the dynamic range of a particular piece of equipment in the vocal chain is 60db, they needed to reduce the vocal's dynamic range by a minimum of 30db, otherwise, the whisper would be lost in the noise floor and the scream would cause distortion in the electronics of said piece of gear. They used to "ride the faders", manually adjusting the volume. This was very difficult and repeatability was next to impossible.

Then some brainiac devised a method of electonically riding the gain, (don't remember who off hand), and compressors were born.

Unfortunately, there were some drawbacks. after compressing, make up gain raised the noise floor of the track and it colored the sound, but you compromised and found the "middle ground". Then some people liked the sound that the compressor imparted on some material and it started being used as an "effect".

Current recording technology has drastically raised the dynamic range. Mid level gear typically has a dynamic range of 96db or better. At 44.1 24bit you can record with no compression and still stay well above the noise floor with no "overs". This is my preferred methodology on tracks that I don't want a compression "effect" on. If I need to reduce the dynamic range of a given track to bring it to the front of the mix I lean towards "riding the faders" which is much easier now with track automation. Repeatability is now possible. This sounds much more natural to my ears.

I personally don't like the "effect" that budget compressors impart on tracks. To my ears they suck the life out of them.

High end compressors can sound cool for "effect" on certain tracks. Distessor, 1176, LA2A etc. Sometimes I'll mult a track and squash the heck out of it, then bring it in underneath the un-compressed track to taste.

To answer your specific questions,

When trying to use compression for reducing the dynamic range transparently, go pre fader. This will make all the equipment upstream have a more stable signal to work with. (mostly moot for modern gear)

Multiband compressors are usually used at the mastering stage to squeese every last bit of dynamic range out of your material, so it's louder then the next guy's (icky)

There are some cool "digital" compressors. Compression by it's very nature is not a "set and forget" kind of device. You need to "twiddle" the knobs on each individual track to achieve the results you are looking for. A setting that work's on a trombone during the "break" most likely won't work on the "chorus", where you don't want it to step on the vocal. If I may be so bold, I'd suggest this approach. (Assuming you don't have thousands of $ to throw at high end hardware compressors). Compress at mix down. For volume level decisions, listen to a particular track, in a particular section. Where is it to loud? Where is it to soft? Now load the most transparent compressor you have, we're not trying to change the sound here, just get the track to sit in the mix where we want. Now "twiddle" the knobs until it sit's where you like and move to the next section and repeat. Repeat for other tracks. You can also achieve this with "fader automation" as I mentioned earlier. This sounds more natural to my ears, and gives you note by note control.

If you're trying to change the sound of the track, ie, using compression for "effect", sky's the limit. Pre, post, whatever manipulates the sound to your wish's. Experimentation is the key here.

For a inexpensive hardware compressor that's relatively transparent, there is the RNC. (Real Nice Compressor) About $250.00. A google search will bring it up. Some people swear by some of dbx's offering's. I think they suck the life out of the sound. YMMV.

For compression "effect", the UAD-1 is a PCI card with emulations of the 1176, LA2A that are pretty good. They have a bunch of other efects to. I think they are running about $650.00 street.

Hope this helps!

Rob





------------------
http://www.songramp.com/homepage.ez?Who=RJC


I used to be indecisive, now I'm not sure... My Songs