Juan Carlos

 great tutorial of Josh Reiss on dynamic range compression and playing in gen~.

In the gain computer (slide 11) I’m trying both options for the soft knee: linear (marked as “bad” in the pdf) and second order interpolation (the “good” one).

The thing is that I’m getting the expected results in my comp prototype only with the linear one and a huge amount of clicks with the second order implementation. Hard knee in both works fine (W = 0).

Is there something I am missing for the second order version? I'd like to compare both formulas and see if there are audible differences.

return (T-(W/2)) + ((((xg-T)+(W/2))*(1+(1/R)))/2);

out1 = softkneeLinear(in1, in2, in3, in4);

return xg + (((1/(R-1)) * (pow((xg-T+(W/2)),2)))/(2*W));

out1 = softkneeSecondOrder(in1, in2, in3, in4);

I’m reading this great tutorial of Josh Reiss on dynamic range compression and playing in gen~.

Hi, 
I’m reading this great tutorial of Josh Reiss on dynamic range compression and playing in gen~.
In the gain computer (slide 11) I’m try

Soft knee second order interpolation