Open darwingrosse opened 4 years ago
This was kind of odd. One mistake was on the output scaling to beap scale. It should be amplified to 5, but was attenuated .5. The outputs of the models varied dramatically depending on how many voices were used. In the simpler instances, I attenuated the output to be correct. In two instances, I completely reworked the voice as mc and used mc.mixdown~ @autogain 1 to sum the voices.
The output levels of the Risset Oscillator range from -130 to +137 (give or take), rather than the expected -5 to 5.