Josh Coalson;196116 Wrote: > oversampling cannot add any more useful information to a signal, at best > it can add as little audible noise as possible. > > whatever is processing the signal afterwards is probably using a float > representation internally. by resampling you are just giving it a > noisier signal to start with. > > resampling with a higher sample rate will increase the size of the FLAC > file because it adds noise that is impossible to compress. > > resampling by increasing the sample resolution (e.g. 16 to 24 bits) > will not increase the size of the FLAC file only if all samples are > shifted to the left and filled in with zeroes; FLAC can detect this > condition. but this will add more audible noise than with the common > method of resolution conversion, which will dither the lower bits, > adding (less audible) noise that FLAC cannot compress and hence making > the file larger.
You don't add dither noise when you upres, you add it when you downres. So yes it would just be a shift and fill by zeros. This in itself does not add any information and therefore compresses about the same. My system will downres (after processing) (probably in floating point) back down to the same word size that came in. -- mswlogo ------------------------------------------------------------------------ mswlogo's Profile: http://forums.slimdevices.com/member.php?userid=9090 View this thread: http://forums.slimdevices.com/showthread.php?t=34597 _______________________________________________ ripping mailing list [email protected] http://lists.slimdevices.com/lists/listinfo/ripping
