What it does
- Reduces dynamic range, which the accepted answer summarised as making the loudest parts softer.
- Evens out an inconsistent performance, which is why it is reached for on voice more than on anything else.
- Softens the attack at the front of a sound, named particularly for percussive material.
- Lets a part sit under something else without disappearing in its quiet moments — useful for anything meant to stay in the background.
The effect nobody planned
One answer traced the sound of compression to constraints rather than to taste. Cutting a record required the dynamic range to be limited so the needle would not jump, so compression was applied as a physical necessity. Analogue tape compressed of its own accord as a side effect of how it saturates. The contributor described loving the sound of their first four-track machine despite its hiss, and realising later that what they liked was the compression it was applying without being asked.
Used deliberately as a sound
Because it changes shape as well as level, heavy compression is a recognisable effect rather than a correction. The example given was a snare drum, where a large transient and a wide dynamic range give a compressor a great deal to act on, and the result is audibly processed rather than merely consistent. Whether that is wanted is a decision about the material.
One point of confusion worth naming
- The word means two unrelated things in audio. This one is dynamic range compression, which changes how a recording sounds.
- The other is data compression, which is about storing a file in less space, and one answer in the thread went there before correcting itself.
- A compressed file and a compressed performance have nothing to do with each other, which is worth keeping straight when reading advice about either.