The main change here is to output one frame at a time. There are a few issues currently:
1) If the output buffer is not large enough for all frames in the packet, it currently truncates the output (with no warning). 2) If there is no global header (e.g. FLV) there is no way to know the number of frames-per-packet without decoding. 3) We need to know how many samples will be output prior to decoding. This seemed to me to be the simplest solution to all 3. Another possibility could be to decode the first packet twice, but I like this patch better. Justin Ruggles (4): libspeexdec: decode one frame at a time. libspeexdec: cosmetics: reindent libspeexdec: return meaningful error codes libspeexdec: include system headers before local headers libavcodec/libspeexdec.c | 68 +++++++++++++++++++++++++++++----------------- 1 files changed, 43 insertions(+), 25 deletions(-) _______________________________________________ libav-devel mailing list [email protected] https://lists.libav.org/mailman/listinfo/libav-devel
