diff options
| author | Paul B Mahol <onemda@gmail.com> | 2015-07-17 11:23:45 +0000 |
|---|---|---|
| committer | Paul B Mahol <onemda@gmail.com> | 2015-07-17 11:29:29 +0000 |
| commit | 297df52f303529fd3c326bc8f398a5b3cbf71c0b (patch) | |
| tree | 771543456b09743687706da1a4e860382c11b184 | |
| parent | 21436b95dc96e9cb2ae3f583f219349976ec1b7e (diff) | |
avfilter/af_compand: fix clipping with default options
Decays and attacks where by default set only for first channel
which caused poor defaults to be used which produced clipping
on any higher channel.
Reported-by: lachs0r
Signed-off-by: Paul B Mahol <onemda@gmail.com>
| -rw-r--r-- | doc/filters.texi | 2 | ||||
| -rw-r--r-- | libavfilter/af_compand.c | 5 |
2 files changed, 7 insertions, 0 deletions
diff --git a/doc/filters.texi b/doc/filters.texi index 518aef8f22..1bef8363b6 100644 --- a/doc/filters.texi +++ b/doc/filters.texi @@ -1462,6 +1462,8 @@ situations, the attack time (response to the audio getting louder) should be shorter than the decay time, because the human ear is more sensitive to sudden loud audio than sudden soft audio. A typical value for attack is 0.3 seconds and a typical value for decay is 0.8 seconds. +If specified number of attacks & decays is lower than number of channels, the last +set attack/decay will be used for all remaining channels. @item points A list of points for the transfer function, specified in dB relative to the diff --git a/libavfilter/af_compand.c b/libavfilter/af_compand.c index 55ad98a15d..610787f54e 100644 --- a/libavfilter/af_compand.c +++ b/libavfilter/af_compand.c @@ -398,6 +398,11 @@ static int config_output(AVFilterLink *outlink) return AVERROR(EINVAL); } + for (i = nb_decays; i < channels; i++) { + s->channels[i].attack = s->channels[nb_decays - 1].attack; + s->channels[i].decay = s->channels[nb_decays - 1].decay; + } + #define S(x) s->segments[2 * ((x) + 1)] p = s->points; for (i = 0, new_nb_items = 0; i < nb_points; i++) { |