Every recording is different, there really isnt a one size fits all setting for compression.
We have to rely on our ears as we add compression, which depends on recording levels and how dynamic the recording is. When it comes to attack, you have to listen to how the transients are affected by the compression as our ears need to hear the transients on vocals to know what words are being said and if the attack is too fast then we lose them. Release can be fairly quick on vocals, less than 100ms.
As a starting point, a ratio of 4:1:1 is a good starting point. 100ms attack is a good starting point as well as 100ms release. Then tune it until it sounds how you want it. You want to listen to the transients and when you can hear them disappearing then you pull back on the attack.
The threshold is entirely dependent on the level going in to the compressor, the ratio is dependent on how dynamic the recording is, and the gain is entirely dependent on how much compression is taking place to bring the level that comes out up to the original level going in.
It does take a while to be able to hear compression, and to understand what compression is, and why its used and when to use it. Its great on vocals as they can be pretty dynamic, and using compression can prevent some parts disappearing into the mix and keeping it at a consistent level.
The key with compression is that once you bring up the makeup gain IT SOUNDS LOUDER EVEN THOUGH THE ACTUAL LEVEL ISNT LOUDER, ITS THE SAME. It tricks our ears into thinking it sounds louder, when in reality its the same and often slightly less, but it shouldnt be more.
There are no real rules when it comes to creativity, only guidelines, they are flexible, what matters at the end is does it sound good.
I hope this can be helpful.
