On Sun, Dec 04, 2011 at 03:56:41PM +0100, Luca Barbato wrote:
> ---
> libavfilter/vf_yadif.c | 7 +++++--
> 1 files changed, 5 insertions(+), 2 deletions(-)
>
> diff --git a/libavfilter/vf_yadif.c b/libavfilter/vf_yadif.c
> index 64b2561..e1f6b42 100644
> --- a/libavfilter/vf_yadif.c
> +++ b/libavfilter/vf_yadif.c
> @@ -145,7 +145,7 @@ static void filter(AVFilterContext *ctx,
> AVFilterBufferRef *dstpic,
> int w = dstpic->video->w;
> int h = dstpic->video->h;
> int refs = yadif->cur->linesize[i];
> - int df = (yadif->csp->comp[i].depth_minus1+1) / 8;
> + int df = (yadif->csp->comp[i].depth_minus1+8) / 8;
>
> if (i == 1 || i == 2) {
> /* Why is this not part of the per-plane description thing? */
> @@ -212,7 +212,7 @@ static void return_frame(AVFilterContext *ctx, int
> is_second)
>
> if (!yadif->csp)
> yadif->csp = &av_pix_fmt_descriptors[link->format];
> - if (yadif->csp->comp[0].depth_minus1 == 15)
> + if (yadif->csp->comp[0].depth_minus1 > 7)
if (yadif->csp->comp[0].depth_minus1 / 8 == 1)
instead maybe. not that's likely that we'll support more than 16 bit
per component soon. I don't like that this is done for every output
frame but avfilter initialization happens before the output format
is known.
> yadif->filter_line = filter_line_c_16bit;
>
> filter(ctx, yadif->out, tff ^ !is_second, tff);
> @@ -354,6 +354,9 @@ static int query_formats(AVFilterContext *ctx)
> AV_NE( PIX_FMT_GRAY16BE, PIX_FMT_GRAY16LE ),
> PIX_FMT_YUV440P,
> PIX_FMT_YUVJ440P,
> + AV_NE( PIX_FMT_YUV420P10BE, PIX_FMT_YUV420P10LE ),
> + AV_NE( PIX_FMT_YUV422P10BE, PIX_FMT_YUV422P10LE ),
> + AV_NE( PIX_FMT_YUV444P10BE, PIX_FMT_YUV444P10LE ),
> AV_NE( PIX_FMT_YUV420P16BE, PIX_FMT_YUV420P16LE ),
> AV_NE( PIX_FMT_YUV422P16BE, PIX_FMT_YUV422P16LE ),
> AV_NE( PIX_FMT_YUV444P16BE, PIX_FMT_YUV444P16LE ),
ok otherwise
Janne
_______________________________________________
libav-devel mailing list
[email protected]
https://lists.libav.org/mailman/listinfo/libav-devel
