http://bugs.winehq.org/show_bug.cgi?id=14717
--- Comment #60 from Krzysztof Nikiel zzdz2@yahoo.pl 2010-12-22 06:18:48 CST --- (In reply to comment #59)
There is one bug, though, in their implementation for 24-bit samples: it doesn't take the sign bit into account.
Indeed.
Also, it would be nice to avoid filtering the input through the FIR altogether if no resampling is needed (e.g., 44.1k -> 44.1k).
I think there is no filtering when in/out rates are equal but i haven't tested the output. Look at the 'if (dsb->freq != dsb->outfreq)' (line 197). I should just copy the samples if they are equal.