Simplify fix15().
authorJustin Ruggles <justin.ruggles@gmail.com>
Tue, 14 Dec 2010 14:51:02 +0000 (14:51 +0000)
committerJustin Ruggles <justin.ruggles@gmail.com>
Tue, 14 Dec 2010 14:51:02 +0000 (14:51 +0000)
commit918cd2255c449409c88879ce8c12a842b13f7c01
tree555f82fb04eaf849802ce2cfd077f83790c7bb2a
parent228315bbd3f8b2e77cea5fe0a14e461716f0c466
Simplify fix15().
Turn it into 2 macros, and use av_clip_int16() and lrintf().
This matches the int16 to float sample conversion in audioconvert.c.
The regression test output is different due to lrintf() rounding.

Originally committed as revision 25956 to svn://svn.ffmpeg.org/ffmpeg/trunk
libavcodec/ac3enc.c
tests/ref/acodec/ac3