Can r8brain (or r8brain pro) do time-domain dithering to reduce cyclic repeated errors during rate conversion?

Do you think there might be a perceived benefit from doing this?  Intuitively, it seems so.  For example, a very good argument for normal dithering involves a digitally-produced sine wave.  Without dithering, the exact same quantization errors will appear in the exact same locations of the waveform, causing significant harmonic distortion. (Recording an analog sine wave or more complex signal, the effect is less severe, but there nonetheless).

It seems to me that the same kind of argument should apply to downsamping. (Probably upsampling too, but I suspect it's not as simple a case.) If so, applying dithering in the time domain (essentially, phase dithering) should make the results sound better.

I'm interested in your opinions.

Thanks!

Jeff

I'm not sure what are you talking about here (i.e. what's really different than what is already available). r8brain PRO applies dithering when you create 16 and 24 bit output files.  Otherwise, indeed, quantization could give a lot of problems.  Also dithering has nothing to do with upsaming and downsampling alone - these are performed in high resolution.