[x264-devel] about block_residual_write_cavlc function in cavlc.c

Jason Garrett-Glaser darkshikari at gmail.com
Fri May 28 19:55:50 CEST 2010


2010/5/28 紫梨 <sj_indark at qq.com>:
>     I'am working with x264 now ,but I find that in
> block_residual_write_cavlc function ( cavlc.c ) i_sign is calculated by
> right shifting 31 bits of runlevel.level[i] . The question is that in the
> struct of x264_run_level_t  (bs.h), level is an array of short value . May
> it be the same if I right shift 15 bits ?

Yes, it would be the same.  Fortunately, on most modern CPUs, shifts
are the same speed regardless of size ;)

Dark Shikari


More information about the x264-devel mailing list