提交 33021583 编写于 作者: P Paul B Mahol

avfilter/vf_mix: clip output pixel values

Signed-off-by: NPaul B Mahol <onemda@gmail.com>
上级 a5172dca
......@@ -44,6 +44,7 @@ typedef struct MixContext {
int nb_frames;
int depth;
int max;
int nb_planes;
int linesize[4];
int height[4];
......@@ -137,7 +138,7 @@ static void mix_frames(MixContext *s, AVFrame **in, AVFrame *out)
val += src * s->weights[i];
}
dst[x] = val * s->wfactor;
dst[x] = av_clip_uint8(val * s->wfactor);
}
dst += out->linesize[p];
......@@ -157,7 +158,7 @@ static void mix_frames(MixContext *s, AVFrame **in, AVFrame *out)
val += src * s->weights[i];
}
dst[x] = val * s->wfactor;
dst[x] = av_clip(val * s->wfactor, 0, s->max);
}
dst += out->linesize[p] / 2;
......@@ -216,6 +217,7 @@ static int config_output(AVFilterLink *outlink)
return AVERROR_BUG;
s->nb_planes = av_pix_fmt_count_planes(outlink->format);
s->depth = s->desc->comp[0].depth;
s->max = (1 << s->depth) - 1;
if ((ret = av_image_fill_linesizes(s->linesize, inlink->format, inlink->w)) < 0)
return ret;
......
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册