here's my Analogy - I use this with my students and it seems to clarify compression for many of them:
imagine that the volume needle of an old style hardware VU meter is not simply
indicating volume but
controlling it. Take it out of the case and turn it so the bare needle is bobbing up and down.
if you leave the needle alone, the music behaves normally- no compression
if you grab it with your hand - you stop the volume from going any louder (or softer)
now instead of grabbing it with your hand, stretch a rubber band across its path to slow its rise.
where you put that rubber band is the
threshold
a low threshold grabs the volume needle and starts resisting its rise immediately. A high threshold means that the volume swings freely until it gets to a higher level and then the rubber band resists
how
thick or unstretchy the rubber band is is the
ratio.
a 2:1 ratio would be on of those skinny office rubberbands with a lot of give. It will push back on the rising needle, but not too much. A 10:1 ratio would be one of those fat produce rubberbands that wrap the broccoli. The sound has to get much louder to get the needle to move against the band.
then you could take a brick and stick it in the path of the VU needle and stop it cold and you have a brick wall limiter.
My students are also tested for being able to
hear various effects. In my experience, compression is usually the hardest one for them to hear. It doesn't really change anything but how the gain varies over time, so it is pretty subtle.
The way to get away from presets is to stop using presets. What I mean is that you have to do a lot of tweaking and listening in order to
make a connection between what you are doing and what you want it to sound like. Having a mental picture of what the controls are doing seems to help.