PCRECORDING.COM - Compresson Basics
One of the most common and useful effects used on recordings is compression. Virtually any recording you hear will have been compressed to some degree. Essentially, compression is a dynamic range limiter. When a signal reaches a "threshold" volume level set by the user, it begins to cut back on the volume of the signal. The speed with which the signal is cut back is called the "attack" and the speed with which the signal is let go is called the "release." Lastly, the amount a signal is compressed relative to its input is called the "ratio."
I am generally a believer in the "less is more" school of compression. If you carefully record your sound initially, a great deal of compression should not be necessary. That is, think ahead about what your instruments frequency responses are and try to plan relative levels ahead of time. Compression is very effective at evening the field between signal types. There are times when it can provide just the right "punch" you are looking for in a signal. Unfortunately, it can also profoundly affect the sound quality, altering it in many different ways - - some good some bad.
Keep in mind that there are a number of different considerations with the use of compression. If your settings are too severe it will adversely affect your sound. For instance, a heavily compressed vocal will end up sounding like the voice is coming through a tube. Also, if you get alternating sharp attacks and releases you will get a pumping effect that will be all too audible. For some truly in-depth coverage of compression basics please visit Tech Notes at Alesis.