Hi Aleksey,

I was using Polysquasher last night (re-mastering) placing it after the revolutionary Soniformer2 (hehe my hype is true!) and really got Polysqusher to sound great using mode2, ratio 1.04.

By playing around with the threshold/knee I got such a transparant sound out of the vocals if I didn't know better I'd think I was doing a mix and had the vocal track all to myself - except this was a full mix.  I ended up with the threshold around -12dB and played with the knee I guess it was about 20dB or so.  The amount of compression varied from 0dB to 0.1 dB just enough to 'follow' the vocal mid peaks and soften them a bit.

I guess I'm wondering what attack/realease Polysquasher use ?  It seems very fast and natural - I think some of this is because of the knee also.

Polysquasher has no defined attack/release value.  Due to its design both attack and release times depend on the program material and the mode used.  At least, I cannot provide any exact attack/release figure.

Hi Aleksey,

One more question on knee.  Is Knee affected by the automatic attack/release algorithm in the same way Threshold is?

I don't know if this is a good way to look at it...but if I had a block diagram of Polysquasher - would the automatic attack/realease algorithm come first followed by Knee and Threshold?

Thanks!

PS You mentioned in another thread that Polysquasher has 1024 latency - is that because of the cool visual gain window?

Knee is directly related to peak detection only, it is not automatic.  Latency is introduced to better catch peaks.