Another way-in-the-weeds analysis. I wanted to use the compressor for clean make-up gain after another effect cut the signal quite a bit. I figured that I could keep THRS=127, RAT=0, and MIX=127. Varying GAIN should just increase gain. But by how much?
I looked at it in two ways: I measured it with a null signal and then I started “poking” the DSP code. Here is what I found:
| GAIN | dB (ideal) | dB (measured) |
|---|---|---|
| 0 | 0.00 | 0.0 |
| 1 | 0.13 | 0.1 |
| 2 | 0.52 | 0.5 |
| 3 | 1.14 | 1.1 |
| 4 | 1.93 | 2.0 |
| 5 | 2.85 | 2.9 |
| 6 | 3.86 | 3.9 |
| 7 | 4.92 | 4.9 |
| 8 | 6.00 | |
| 9 | 7.08 | |
| 10 | 8.15 | 8.1 |
| 11 | 9.20 | |
| 12 | 10.21 | 10.2 |
| 13 | 11.20 | |
| 14 | 12.15 | |
| 15 | 13.07 | 13.0 |
| 16 | 13.95 | |
| 17 | 14.80 | |
| 18 | 15.62 | |
| 19 | 16.41 | |
| 20 | 17.18 | 17.2 |
| 21 | 17.91 | |
| 22 | 18.62 | |
| 23 | 19.31 | |
| 24 | 19.97 | |
| 25 | 20.61 | 20.6 |
| 26 | 21.23 | |
| 27 | 21.83 | |
| 28 | 22.41 | |
| 29 | 22.98 | |
| 30 | 23.53 | |
| 31 | 24.06 | 24.0 |
Could be useful if you know how much an earler step decreased gain. There is actually about 48 dB of gain on tap, but in my experience things start clipping after 24 dB. If you’re actually using the compressor to compress, you might need that kind of make-up gain.