On 12/29/2011 06:08 PM, Mike Melanson wrote: >> The patch implements make fate-ADX . Link to sample files : >> http://www.google-melange.com/gci/work/download/google/gci2011/7130478?id=3001 > > The CRCs for samples 4 and 7 are all 0. This isn't useful for testing. > > Actually, I downloaded the samples and listened. They're all extremely > short and I didn't hear anything for any of them. What was your approach > for encoding these? > > Finally, I think this is the wrong approach for testing this codec. You'll > note that Libav has both an encoder and a decoder for CRI ADX. When this > is the case, and the encoder exercises enough of the codec's features, > it's best to use a combined test which encodes a contrived wave file > (generated via the internal audiogen tool) and decodes the compressed > result. > > Look at tests/codec-regression.sh for examples of how this encode-decode > round trip operates in testing. New samples should not be necessary for > this codec.
The encoder isn't great though. I do have a codec test pending for this, but I think it would be nice to also have at least a couple purely decoding tests with external samples as well. Oana, could you pick one or two of these samples that give good code coverage? -Justin _______________________________________________ libav-devel mailing list [email protected] https://lists.libav.org/mailman/listinfo/libav-devel
