Original commit message from CVS:
* gst/audioconvert/gstchannelmix.c: (gst_audio_convert_mix):
more overwriting protection due to modifying channels one by one
instead of all at once
Original commit message from CVS:
* gst/audioconvert/gstchannelmix.c: (gst_audio_convert_mix):
walk the samples backwards if out_channels > in_channels so we don't
overwrite data