[FFmpeg-devel] [PATCH] avfilter/af_loudnorm: Don't mix dB and linear values when calculating linear offset with inputs <3s

Sebastian Dröge slomo at coaxion.net
Sun Apr 26 22:48:55 EEST 2020


From: Sebastian Dröge <sebastian at centricular.com>

s->target_i and global are in dB but s->target_tp and true_peak are
linear. Instead of mixing these in the calculations, convert the former
first to have all following calculations in the same unit.
---
 libavfilter/af_loudnorm.c | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

diff --git a/libavfilter/af_loudnorm.c b/libavfilter/af_loudnorm.c
index 314b25fa39..8e3cdc36db 100644
--- a/libavfilter/af_loudnorm.c
+++ b/libavfilter/af_loudnorm.c
@@ -453,10 +453,9 @@ static int filter_frame(AVFilterLink *inlink, AVFrame *in)
                 true_peak = tmp;
         }
 
-        offset    = s->target_i - global;
-        offset_tp = true_peak + offset;
+        offset    = pow(10., (s->target_i - global) / 20.);
+        offset_tp = true_peak * offset;
         s->offset = offset_tp < s->target_tp ? offset : s->target_tp - true_peak;
-        s->offset = pow(10., s->offset / 20.);
         s->frame_type = LINEAR_MODE;
     }
 
-- 
2.26.2



More information about the ffmpeg-devel mailing list