Stereo Tool

The analog peak compressor never settles on any level but instead seemingly goes to infinity unless a very high attack/release shape is chosen.

I just tested this with the SB compressor, default settings, Feedback to 0%, Ratio to 1000:1. Also with Feedback at 100% and FB^2 enabled. With a beep tone, a difference of 32 dB in drive does not have a visible impact on the output level meter. Tried both with fast and slow attack/release times and with attack slower than release.

Thanks for testing.

I re-did the measurements with a simpler setup to rule out error sources:

1. Generate 1 kHz sinewaves testone files at -8 dB, -3 dB and 0 dB.
2. Apply ST processing by using the SB2 of the CMD version. The settings are threshold = -20 dB, ratio = 2, attack time = 20 ms, release time = 60 ms, knee = 0 dB, feedback = 0 % and auto fix shape = on. (The settings don't really matter, though.)
3. For every input and output file data to get rid of 0 samples combine every 48 samples (samplerate = 48 kHz) by taking the mean of their absolute values.
4. Plot the quotient output_data/input_data.

What am I doing wrong? How to properly measure such things?

What do you see when you compare the input and output files in a wave editor?

_________________
Imaging producer & Stereo Tool audio processing consultant

The waves verify the results above. Since the input is only a sine wave and the output only processed by ST and it works fine for RMS or high shapes I don't see what I could've done wrong.

I tried a signal of 1s with alternating signs only. That works fine, too.

Problem identified! All my fault, sorry for beeing so stupid. Peak detection settles just fine.

I assumed that using the same attack time = 20 ms for both peak and RMS would roughly make them settle at about the same time, but of course peak detection is 25x slower (yes, every audio engineer would've known that, but I didn't).

From the left plot below, because RMS (blue) settles early I concluded that the same should've happened for peaks (red). However if you look at the right plot which shows the complete picture (waiting 30x longer), you see that peak detection settles just fine some ages later.

This can be fixed by using (much) higher shapes than 1, so everything is fine and file closed. So sorry again for the trouble!