Question about 'parallel compression'

From the Vintage64 docs:

The "Parallel Compression" setup can be used for upward compression, lifting up quiet passages of music.

How exactly does that work? They don't explain it other than to say it can be done. I can do it with the expander in Ozone, but what specifically about the compression being done in parallel lends itself to this function?

Parallel compression is using compression as a send effect rather than an insert. You combine the original with the compressed version to bring up the quiet sounds yet still keep the original dynamics. I don't believe expansion is capable of this. Remember, there is upward and downward compression and expansion. Parallel compression can be used as upward compression.

downward compression: make sound above threshold quieter (smaller dynamic range)
upward compression: make sound below threshold louder (smaller dynamic range)
downward expansion: make sound below threshold quieter (larger dynamic range)
upward expansion: make sound above threshold louder (larger dynamic range)

Well, if you look at routing in the "parallel room lift" preset in V64 there really are 2 compressors in parallel. One performing light downward compression and one with heavy compression and volume gain performing the upwards compression. It has a little bit more flexibility in that regard.

ORIGINAL: Jonny M
I'm guessing putting compression in an effects send would give the track a more natural sound then? So if I'm going to compress something quite heavily it's probably a good way of not making the track sound to machinesed.

What's the 'standard' for applying compression? Direct to track or in an effects send?

I'm inclined to say that most people use compression as an insert (downward compression) since it helps out with clipping issues better than parallel compression does. Parallel compression retains the full transients of the original track which may or may not be what you want. It would be best to try both to see which one works better.

When I heard the term, I thought of a technique that's used in expensive active hearing protectors for shooters--and probably also construction workers and pilots and other folks in loud environments.

Passive hearing protectors are just mechanical earmuffs that attenuate sound, especially high-frequency sound. Standard active hearing protectors have microphones on the outsides of the earmuffs and headphone drivers on the inside with a compressor between, so that when somebody speaks normally to you you can hear them pretty much normally, but if you or someone near you fires a gun, the compressor shuts down and the BLAM! becomes a pop! while you lose a second or so of conversation. (You'd certainly lose at least that much if you didn't have the protectors on at all!)

But there are hearing protectors that split the signal from the microphone into several frequency bands and run each band through its own compressor, then re-sum them after compression. With those, only the parts of the incoming signal that are dangerously loud are reduced. If you were listening to a conversation while somebody emptied the magazine of a Thompson 1927 a foot away, you wouldn't miss a word. They're also good for repetitive or continuous noises, because they don't shut your entire spectrum down the way single-compressor protectors do: that can be very fatiguing.

Anyway, that's what I thought of when I heard "parallel compression."