I'm currently making some environmental loops for a game project, and I've run into what I believe may be a precision artifact.
I basically create a region and add audio material with a heavy low-end. To avoid clicks, I add fadeins and fadeouts which roughly correspond to the base frequency of the wave. I then render the loop, and study the output. It seems there lowest amplitude at the absolute last sample is about -44 dB, which would be conclusive if the quickfades were done at an 8-bit resolution.
As a comparison, I made a Volume Envelope with roughly the same curvature, and this time the resulting wave faded smoothly towards -inf dB, suggesting the Volume precision is in fact larger than 8 bits (I'm guessing 16 or even 32 bits)
Are my observations correct, and is this fixable? In exrtreme cases, this means we have to either use Volume Envelopes, which are slightly more cumbersome than quickfades, or post-edit the resulting waves, which is simply bit of a pain in the ass.
I basically create a region and add audio material with a heavy low-end. To avoid clicks, I add fadeins and fadeouts which roughly correspond to the base frequency of the wave. I then render the loop, and study the output. It seems there lowest amplitude at the absolute last sample is about -44 dB, which would be conclusive if the quickfades were done at an 8-bit resolution.
As a comparison, I made a Volume Envelope with roughly the same curvature, and this time the resulting wave faded smoothly towards -inf dB, suggesting the Volume precision is in fact larger than 8 bits (I'm guessing 16 or even 32 bits)
Are my observations correct, and is this fixable? In exrtreme cases, this means we have to either use Volume Envelopes, which are slightly more cumbersome than quickfades, or post-edit the resulting waves, which is simply bit of a pain in the ass.