aacenc: make sure to encode enough frames to cover all input samples.
authorJustin Ruggles <justin.ruggles@gmail.com>
Sat, 28 Jan 2012 22:51:22 +0000 (17:51 -0500)
committerJustin Ruggles <justin.ruggles@gmail.com>
Mon, 30 Jan 2012 19:20:24 +0000 (14:20 -0500)
commit89eea6df281f535849055f9ce0e49b36eeef767a
treedf8f3ec229ffd25a7bef58f93402e8ec62d81db7
parentf44005b610abc5b1aca5b3191b64a0a0ba65a03f
aacenc: make sure to encode enough frames to cover all input samples.

Currently, any samples in the final frame are not decoded because they are
only represented by one frame instead of two. So we encode two final frames to
cover both the analysis delay and the MDCT delay.
libavcodec/aacenc.c