A recent article explored the problem of parallel compression combined with automation, but what does parallel compression actually sound like on vocals?
How to automate tracks that have parallel compression is an article I wrote about the problem of changing fader level, either manually or through automation, while you are parallel compressing. Of course the article includes a simple and effective solution, so I recommend reading it if this is an issue for you.
What I didn’t include in the article because it would have drawn attention away from the actual issue, is what parallel compression sounds like, specifically on vocals.
Here’s a vocal. You might just need to listen to a few lines, or further into the song as you please. Listen out for breaths, small mouth noises, background ambience and reverberation, and later on another singer warming up in the studio’s rest area – We did of course retake this but the intrusion forms another interesting demonstration of what happens in compression.
The microphone by the way is a Neumann U47 from Abbey Road Studios’ famous collection, and the singer is the excellent Chilli Gold.
And now the parallel compressed version. Here’s a pic that includes all of the settings (click the pic for a larger version)…
And the audio…
The effect of parallel compression is subtle, and is heard on the lower-level sounds. The higher level sounds are almost unaffected. This is the entire point of parallel compression. Regular compression squashes down the high levels; parallel compression brings up the low levels.
If it isn’t clear to you, then I could say listen again. But I have an interesting test that will make it crystal clear to you that there is a big difference between uncompressed and parallel compressed.
What I have done is to put the uncompressed version in the left channel and the parallel compressed version in the right. You should listen on headphones.
What you will hear is that when the level is high, the image is central, showing that the level of both uncompressed and parallel compressed is the same. But when the level is low, the image will swing to the right, showing that the parallel compressed version is louder.
Of course you would never do this in a real-world project, but it is a very clear demonstration (if you are listening on headphones as I suggested) and you can now go back to the previous uncompressed and parallel compressed examples and listen with newly-focussed ears.