[FFmpeg-cvslog] opustab: macro constant as a double

Lynne git at videolan.org
Mon Apr 29 16:31:22 EEST 2024


ffmpeg | branch: master | Lynne <dev at lynne.ee> | Mon Apr 29 01:44:46 2024 +0200| [f492095bd3e22ecf5565a08437a0816910bac949] | committer: Lynne

opustab: macro constant as a double

May increase intermediate precision on some compilers.

> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=f492095bd3e22ecf5565a08437a0816910bac949
---

 libavcodec/opustab.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/libavcodec/opustab.c b/libavcodec/opustab.c
index 47624fe651..917375253e 100644
--- a/libavcodec/opustab.c
+++ b/libavcodec/opustab.c
@@ -1164,7 +1164,7 @@ const uint32_t * const ff_celt_pvq_u_row[15] = {
  * libopus uses a slighly rounded constant, set to 0.85 exactly,
  * to simplify its fixed-point version, but it's not significant to impact
  * compliance. */
-#define CELT_EMPH_COEFF 0.8500061035f
+#define CELT_EMPH_COEFF 0.8500061035
 
 DECLARE_ALIGNED(16, const float, ff_opus_deemph_weights)[] = {
     CELT_EMPH_COEFF,



More information about the ffmpeg-cvslog mailing list