diff mbox series

[FFmpeg-devel,2/6] ac3enc: do not clip coefficients after transforms

Message ID MQc_nSR--3-2@lynne.ee
State Superseded
Headers show
Series [FFmpeg-devel,1/6] ac3enc_fixed: convert to 32-bit sample format | expand

Checks

Context Check Description
andriy/x86_make success Make finished
andriy/x86_make_fate fail Make fate failed
andriy/PPC64_make success Make finished
andriy/PPC64_make_fate warning Make fate failed

Commit Message

Lynne Jan. 9, 2021, 7:19 p.m. UTC 
In either encoder, its impossible for the coefficients to go past 25 bits 
right after the MDCT. Our MDCT is numerically stable.
For the floating point encoder, in case a NaN is contained, lrintf() will
raise a floating point exception during the conversion.

Patch attached.
Subject: [PATCH 2/6] ac3enc: do not clip coefficients after transforms

In either encoder, its impossible for the coefficients to go past 25 bits
right after the MDCT. Our MDCT is numerically stable.
For the floating point encoder, in case a NaN is contained, lrintf() will
raise a floating point exception during the conversion.
---
 libavcodec/ac3enc_template.c | 3 ---
 1 file changed, 3 deletions(-)
diff mbox series

Patch

diff --git a/libavcodec/ac3enc_template.c b/libavcodec/ac3enc_template.c
index de6eba71d8..4f1e181e0b 100644
--- a/libavcodec/ac3enc_template.c
+++ b/libavcodec/ac3enc_template.c
@@ -383,9 +383,6 @@  int AC3_NAME(encode_frame)(AVCodecContext *avctx, AVPacket *avpkt,
 
     apply_mdct(s);
 
-    clip_coefficients(&s->adsp, s->blocks[0].mdct_coef[1],
-                      AC3_MAX_COEFS * s->num_blocks * s->channels);
-
     s->cpl_on = s->cpl_enabled;
     ff_ac3_compute_coupling_strategy(s);