In my experience unless you're working on a calibrated setup (which isn't quite as quiet as you might think) then you really need to check the mix at various levels.  The ear certainly isn't "flat" in its frequency response and quiet levels are just as uneven as loud levels.

For example, your mix might sound good at low level but when you turn it up the rhythm section has no weight to it and the high mids/highs can become overbearing.  Conversely you can over process things when listening loud and when you go back to quiet then you often lose a lot of clarity.

Thank you for your reply. If you'd like to continue this discussion, assume the hypothetical "I" is deaf. I am not deaf, and do not mean to insult or disrespect the hearing impaired.

Also, please refrain from posting to refute the rationale of this topic. If that is your contribution, kindly discuss it elsewhere. I'm not asking for personal assistance. I'm making an (thus far, relatively poor) attempt to compile a library of industry technique pro bono (for the benefit of the members of TPF).

The best analogy for mixing I've found is a food chef.  They take a variety of ingredients and using their techniques and methods, create a tasteful dish which others can enjoy.   While the fundamental ingredients may be the same, it's their taste factor combined with their recipe which makes their dish special.

Going back to what seems to be your goal about providing a resource of basic techniques, there are so many tutorials available that tell you how to use a knife or what temperature things should be baked at that I'm not sure another one is needed.  It's far more interesting to find how how someone got to their end goal and realize that what works for them may not work for you but it can inspire you to forge your own path and discover a new method.  It bears mentioning again that unless you have that all-important "taste" factor then no amount of knowing basic technique will make you a master chef.

The attack is how much time the compressor will delay its activation... *snip*
The release is how fast the compressor level returns back to 0db... *snip*

This is simply not true.  Please see the "understanding compression" thread.

I think you might be starting to confuse yourself.  In your example, below the black lines (or threshold) is the boundary where no gain reduction will occur however as soon as that boundary has been crossed then attack/release phases come into play.

Let me try and explain it another way and we'll take an imaginary "perfect" compressor as our basis with some tone values.

We'll use a sine tone @ -12dB FS with a compressor set with a threshold of -18dB FS and a 2:1 ratio (we'll ignore attack/release values for the moment).  As soon as that sine wave crosses the threshold the detection circuit will "hear" that the signal is too loud and instruct the gain reduction circuit to start working.  How quickly it does this is related to our attack setting (and compressor design!).  Our channel output level (post compressor) should now read -15dB FS as we're applying 3dB of gain reduction (6dB over threshold with a 2:1 ratio) and it should be steady once the attack phase is complete (and it's only steady because the input level is not varying).

If we now increase the gain of the sine wave tone to -6dB FS the attack phase will begin again (remember at this point we have not dropped below the set compressor threshold).  Our output should now be -12dB FS as we're applying 6dB of gain reduction (we're now 12dB over threshold with a 2:1 ratio).

If we now decrease the sine wave tone back to -12dB FS the release phase will start, and if we drop it even further back it will continue to release until no gain reduction is taking place and we're below the threshold again.

If you extrapolate that to a musical signal which is constantly varying the amount of signal over threshold then you can see how as soon as you cross that boundary into gain reduction then the circuit is constantly attacking and releasing depending on detection circuit and attack/release settings.

Hope that makes things clearer for you.

There is some mis-information being spread here.

The easiest way to think about a compressor is (as has been mentioned) a robot on the fader.  It is simply adjusting the fader level depending on what it's listening to (the detection circuit) and the control parameters you're giving it.

Some things to take note of.  Once a signal crosses a boundary (depends on threshold as well as knee settings) the compressor is constantly attacking and releasing (how it does this depends on compressor design and attack/release settings - this is partly what gives each compressor its unique sound).   The attack time is how long it takes for the compressor to reach around .707 (or approx 2/3rds) of the targeted gain reduction.  Release is the opposite of this (how long it takes to restore that gain reduction that has been applied).  An easy way to think about the ratio is if you imagine a rubber band with the signal pushing over the threshold into it.   The tighter you make that rubber band, the harder the input signal will have to "push" against it to see changes on the output.

Hope that helps clear up some of the confusion which often surrounds this topic.