"Rise from your grave..."
 
I just came across this thread, and I find it interesting. However, I am not finding the particular faults, as described by the OP, in any of my compressor plugins.  Maybe I am, but I need a good, bad example to know what to listen for.

I tried the Sonitus Multiband compressor, but the only thing I really heard was that it really didn't do much.   On the other hand, I tried the single-band Sonitus Comp, and it sounded very good and clean with no defects.  Martin Eastwood CompressivePro was about as USELESS (capitals for a reason) as it can get! There was no real change in sound or defects, but the controls didn't do anything! FXpansion DCAMFreeComp...just a "tinny" sound that was always present, probably caused by the plugin's delay and a dry signal (i.e. some parallel compression) always present in the mix. Variety of Sound's DensitymkIII performed well, with relatively light compression and a decent attack--the sound seemed to ebb and flow well. This one is actually regarded as a good, freebie compressor plugin.
 
Fast release times and a high ratio on the PC76 introduced "sputtering noise." This however, could be a characteristic of the real-deal Holyfield, so perhaps it is OK in terms of the quality being true to form. The PC4K, on the other hand, sounded clear and clean, but not very aggressive.
 
Melda's MCompressor was very "sputtery" when set to fast attacks and moderate release times. It only sounded good when the knee was set to super soft.
 
Fish Filet's Blockfish put a lot of bass into the difference mix. It also added some "swirling" type noise into the mix that could be heard if low cut was enabled. It also wasn't very tweakable, as the only setting that sounded decent was to have the compression knob remain at the halfway mark.

A demo of Fabfilter's Pro-C was very clean and all controls worked as they should, yielding very smooth and flowing results. Obviously, this is a high-end plugin with much praise. Overall, it was very tweakable and predictable.

This test was also very useful to understand how the ProChannel Console Emulators work. I could definitely hear the saturation/distortion if the Drive knob is pushed way up. I guess distortion is easy to hear, but because it's subtle, it may not be obvious on the channel or in the mix.
 
So maybe I did the test correctly and heard the right things. What do you guys think?

Talk about rise from the dead! Yes you did well. The fact that you heard all those things means you did the test perfectly and it revealed much to you.
 
The problem with compressors is when the music is present you cannot actually hear what they are doing! (or easily I mean) This test reveals how the compressor in question is actually doing its thing. Note how the better ones behave in this test.
 
And as you have mentioned too just because a good quality compressor behaves badly in this test it does not mean it is bad either. It means it is not adjusted well that is all. This test allows you to arrive at smoother more bouncy settings which is what you are after.

I would think that they're all different, and can be used for different applications? I like Fab... But I also use Alloy and it's compressor(s) and I like PSP Vintage Warmer for acoustic guitar? I'm sure there are others that are really sweet, these are my favs. 
 
I did enjoy this,
"If I understand the use of compression right, is it mean to be used to shorten the width of the sound in the stereo spectrum or sound of a song? I dunno, For some reason I can not hear any difference using compression. To me it's more or less been a glorified volume increasing utility."
 
That was pretty funny.

i find a lot of times, what i'm going for with a compressor, is what it DOES NOT do.
 
sometimes, a compressor is simply an effect for color.
 
other times, more of a level controller.
 
knowing how a compressor grabs a signal, and lets go, is a good thing to know....
but not necessarily how you will end up using it.

My best is Logic's own compressor. I did the above test and it confirmed that it behaves adequately and doesn't add anything nasty to the mix.
 
At this point, it's practically the only compressor I use, because it does it all - VariMU, Opto, VCA, FET, etc. and it does it extremely well.
 
Others may offer instant gratification and nicer graphics, but the truth is, the more I use this one, the better I know it and the closer I get to instant gratification. Even on the stereo bus - I prefer it to every bus compressor I've tried except maybe the Duende SSL. 
 
Good enough for me.

Why can you just cut to the chase and tells us what compressors we shouldn't use, because of artifacts that we never heard in our mixes, are now proved to be there?  ;-)

But on a serious note, a mention of the supplied Cake compressors would be interesting to read about.  That way I know for sure if their CA2A is junk and I should buy the competitor's version instead!   If I can prevent people giggling about my mixes because of incorrect plug choices, that would be at least some sort of advancement in this craft.

It is not so much about which plug-ins are bad and which are good. It is about setting up your compressor to give you the right result. Something many don't know how to do and this test goes a long way towards getting it right.
 
A really good plug-in can be set badly and it will perform badly in this test just as an another compressor (more average or less expensive) can be set to work very smoothly under this test. (eg Rain's Logic compressor which I agree with him is very good in fact)
 
If however you cannot get your plug-in (good or average) to perform well under this test then it is a matter of perhaps realising it is a not so good plug-in after all and you may be better off looking for something different.

I had a few minutes to play with this technique so I ran through a few compressors I have installed.
The best sounding were the Waves SSL compressor, Blue Cat Dynamics, and the NI Solid bus compressor.
All of the Waves compressors I have performed well as did the TRackS3 classic compressor.
The worst were the pro-channel compressor, CA-2A leveler, and the Nomad Factory CP2S. All of them had pretty nasty attack noise and adjusting the attack time did little to help.
It was very easy to get the null using makeup gain on the compressors, I had no issues with latency. YMMV
I was pleasantly surprised by Solid, I'm going to have to use it more often.
 
Thanks for pointing out this technique Jeff.
 
P.S. I didn't test the Sonitus compressor but it will always be my goto compressor for drop ducking, it's amazingly tweakable for that purpose.

Interesting how Tom picked some plug-ins that did not perform well.  That mixmkr is exactly what I mean.  If the attack part created some artifacts, adjusting the attack will usually fix the problem.  The fact they could not means there may be something seriously wrong in how these plug-ins are handling the attack part of the sound.  And these artifacts do get added into your signal source and are adding unwanted sounds.  You just don't notice them so much while the music is present but when you take it away suddenly it becomes quite clear.
 
BTW Tom keep persevering with various parameters until the attack nasties go away. Other factors can influence things such as ratio (were you using a very high ratio) and threshold as well. You might find these plugs ins will suddenly come good when you start backing some of these other parameters off a bit.
 
If you however you cannot do anything to reduce the attack artifacts then simply keep that in mind before using them.

Nice thread Jeff. I too have Stav's book on mixing secrets and it certainly gives some good advice. A little too expensive I thought, but when you don't have the cash to splash on a full blown engineering apprenticeship every bit helps.