i kind of invented that somehow (or a bit more than somehow.)

one method how to build that and the first one which i was making in max is actually really simple:

- you build a dynamic gate with a variable breakpoint/treshold.

- it will have two outputs, one releases the part above the treshold and one the part below the treshold.

- now you cascade a few of them in a series. the "above treshold" part of the first one goes into the second one and so on.

- now you have split the signal onto several signals according to their power. when you mix those signals together again you get the original back. when you change the gain of the individual channels, you can do any kind of nonlinear, non- "normal compressor/gate" dynamic curve to your input.

there are other methods but this one is working well enough for most things.

-110

-

Hi Roman! Many thanks for your tips, I will dig them out. You did your dynamic gate in amplitude domain or in fft domain? I was thinking about these two techniques and I think it's possible to do them both ways but I didn't test the results.

Best Regards

no fast fouriers in 110 land.

simply an average/rms of the input which controls the weight of output channel 1 via rampsmooth.

you can use the same time values for average and rampsmooth in order to get nice round curves, but you dont have to.

the output channel 2 is made by (input minus channel 1).

-110

Hi Roman! I figured out how to do it in frequency domain, but in amplitude is missing me something, could you send me an example of what you describe in the other post.

many thanks, DC

You should also look into the kompressor example in the examples folder a different aproach, but I'd go that way...