posted 26 November 2000 17:07            

I just tried granulating a sample of very low frequency content (about 27 Hz) recorded to disk as a stereo file at 48 KHz sample rate. After granulation, I notice that the spectrum of the signal from 12 KHz on up to 24 KHz is a reversed replica of the entire spectrum from DC to 12 KHz. This indicates that the sample from the disk was decimated by a factor of two.

Is this because I was granulating a stereo sample? Or is it because I recorded a sample at 48 KHz sample rate and your software normally expects 44.1 KHz?

Now that I think about it, I'll bet your granlator expects mono samples, and it treats my 5 seconds of stereo as 10 seconds of mono. This would account for the decimation by 2...

posted 27 November 2000 00:52            

The manual even states that the granulator needs mono samples. This little test also proves to me (who knows nothing about the innards of .AIF files) that the stereo samples are interleaved, just like stereo .WAV files.

The pitch should have been halved as well, but at 27 Hz, with all its harmonics, who can tell?