The Elf wrote:The thing is, most people bandy around the term 'compressed' as if it is one specific effect. It very much isn't! Compression can reduce dynamic range, but it can also *increase* it
The gospel is above.
As an addition, it may be worth commenting that what has become (largely) irrelevant - or, better, a "matter of personal preference" as opposite to "need" - is applying compression when recording.
You once needed to compress in input it because the recording medium had limited dynamic range, and considering that you wanted a good amount of headroom, compressing was often a necessity. Now the recording medium (24-bit PCM data streams) has much more range than needed to record any sensible audio, so you can just skip compression when recording and applying afterwards if you so like.
That said, if you get a good sound and trust your monitoring, you can still compress in input just fine - all of my vocal recordings tend to have a LA2A somewhere in the chain, and recently the Pultec clone in bypass mode as well, because I would just do the same first thing anyways.