[FFmpeg-devel] [PATCH 1/3] lavfi: add haldclutsrc filter.
Clément Bœsch
ubitux at gmail.com
Sun May 26 23:15:04 CEST 2013
On Sun, May 26, 2013 at 08:03:18PM +0000, Paul B Mahol wrote:
[...]
> > +static int haldclutsrc_config_props(AVFilterLink *outlink)
> > +{
> > + AVFilterContext *ctx = outlink->src;
> > + TestSourceContext *s = ctx->priv;
> > +
> > + if (outlink->w || outlink->h) {
> > + av_log(ctx, AV_LOG_ERROR, "output size must be configured with a
> > level option\n");
> > + return AVERROR(EINVAL);
> > + }
>
> I find this ugly. If filter does not allow user to change w/h, options for
> changing w/h should not be available.
>
I'll possibly change this to guess the level. Removing the size from the
common options will likely make things a bit ugly.
Also, after a talk with the author of the Hald CLUT, it was suggested to
add some padding at the bottom (or right) to add some extra information
ignored later (typically for a demonstration of the effect of the CLUT, so
users could watch the effect by opening the CLUT image himself)
> Also if something else sets w/h, it should be ignored/replaced.
>
> > + s->w = s->h = s->level * s->level * s->level;
> > + return config_props(outlink);
> > +}
> > +
> > +static const AVFilterPad haldclutsrc_outputs[] = {
> > + {
> > + .name = "default",
> > + .type = AVMEDIA_TYPE_VIDEO,
> > + .request_frame = request_frame,
> > + .config_props = haldclutsrc_config_props,
> > + },
> > + { NULL }
> > +};
> > +
> > +AVFilter avfilter_vsrc_haldclutsrc = {
> > + .name = "haldclutsrc",
> > + .description = NULL_IF_CONFIG_SMALL("Provide an identity Hald
> > CLUT."),
> > + .priv_class = &haldclutsrc_class,
> > + .priv_size = sizeof(TestSourceContext),
> > + .init = haldclutsrc_init,
> > + .uninit = uninit,
> > + .query_formats = haldclutsrc_query_formats,
> > + .inputs = NULL,
> > + .outputs = haldclutsrc_outputs,
> > +};
> > +#endif /* CONFIG_HALDCLUTSRC_FILTER */
> > +
> > #if CONFIG_NULLSRC_FILTER
> >
> > #define nullsrc_options options
> > --
> > 1.8.2.3
>
> Rest lgtm.
>
> I wonder if something similar could be done for yuvs.
>
I don't think so. Interpolating with R, G and B works because they have
the same weight. I doubt you can do that easily with YUVs.
--
Clément B.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 490 bytes
Desc: not available
URL: <http://ffmpeg.org/pipermail/ffmpeg-devel/attachments/20130526/559ac185/attachment.asc>
More information about the ffmpeg-devel
mailing list