You're totally correct, and just to clarify for anyone who's not into audio/DSP this isn't referring to "data compression" but rather dynamic range compression.

You're totally correct, and just to clarify for anyone who's not into audio/DSP this isn't referring to "data compression" but rather dynamic range compression. Basically quieter sounds get amplified so that the difference between a quiet sound and a loud sound is smaller, and then the whole thing gets amplified so that it's dancing right on the edge of clipping.