[FONT="]Compression has 3 basic uses (the first two are very similar), from what I have seen:
1) Speaker protection
2) Signal leveling
3) "Squash" effect
All of this is accomplished through the basic function of a compressor- gain reduction. This is how they compress the dynamic range. Some compressors have an adjustable amplifier on the output that provides "make up gain". This is why some folks think that a compressor makes a signal louder. The amplifier made the signal louder
period.
You usually have a few controls on a compressor:
Ratio: the amount of reduction applied for a given increase in signal. The higher the ratio the higher the reduction. 1:1 equals no compression. 3:1 means that you get 1/3 of the actual increase in signal strength. Infinity means that no matter how big the input signal gets, the output signal gets no larger.
Threshold: What point the compression actually starts happening. If you set the threshold low the compressor starts working quickly. If you set it low with a high ratio you will get a very steady output (though low) no matter what the input signal is. This is how you get a super-squashed sound.
Knee: Hard Knee is where the compression comes on very quickly near the threshold. Soft Knee is where the compression come son in a gradual manner. Most soft-knee settings actually start compressing before the threshold setting.
Also, some compressors have a display for gain reduction, and sometimes it is a multimode display that can show other parameters. I really like to have an indication of the amount of gain reduction being applied. It makes it easier to set the compressor correctly.
In Use:
The least audible use is a soft knee with the threshold set high and the ratio set low. This will keep the big peaks down and leave most of the dynamic range intact. The actual setting for your application will be a function of the input signal level, the range of the input signal, and the output device sensitivity. Again, when you set the threshold the threshold is fixed at that setting, so the sound of the compressor changes with changes in input level.
For Example:
Using volts instead of dB, for clarity, say you set the threshold at 1.5V and the ratio at 4:1. As you approach 1.5V the gain reduction starts, and as you go from 1.5 to 1.9 the output level actually goes only to 1.6 (because 1/4 of the 0.4V increase was allowed). If you have a level that swings from 0.5 to 1.8V you will only get compression at the peaks with a range of 0.5 to ~1.6V. If you engage an effect that boosts the range to 1.0 to 2.3V your range is slashed to 1.0 to ~1.7V, and you hear compression earlier. You didn't change any settings but the compressor will be more of a factor in the final sound.
As for where to put a compressor... I wouldn't take any advice other than "try it a few ways and see what sounds best", because that is what you are going to have to do anyway.
)
