Well,
It depends what sort of testing you want to do. The most basic test of a compressor just requires the test oscillator in a console and using a channel with the compressor as an insert, or in a buss send, and then monitoring the input and output levels with the consoles metering. This way, you can set the no-compression gain of the compressor, and then start adjusting the threshold and ratio controls. By increasing the level of the signal from the oscillator, and monitoring the output level, you can check both the threshold and ratio levels for the compressor.
Armed with an AF signal generator and an audio millivoltmeter, you can get a (generally) more accurate result. At the same time you can also start to determine the frequency response of the compressor, and find its upper and lower -3dB points.
To discover any distortions due to the compressor, either when the signal is "straight thru" (not all compressors are True Bypass- most often the threshold is just taken way up, or the compressing-element is removed from the signal path- so the signal still passes through the input and output circuitry) or during compression, a scope and signal generator are best used.
Obviously these tests are best observed using a sinewave, a squarewave will show up LF and HF response problems, but will also look a bit strange even with a fully working compressor- the rise-time of a squarewave from a test-oscillator is quite demanding when seen by a compressor, and may not give accurate compression-ratio results. Bear this in mind when adjusting Attack and Release- the shape of the output waveform will be determined by these settings- what you see may NOT be distortion in the compression circuitry, but the "shaping" of the wave by the Attack-Release timing.
But the oscillator/scope method allows you to see crossover distortion and asymmetrical clipping, which is handy to allow a check of the basic functioning of the box.
Then there's always an AP....
:thumb:
Mark