[FFmpeg-devel] [PATCH 2/2] avfilter/vf_scale: allow overriding in/out yuv colorspace type
Paul B Mahol
onemda at gmail.com
Sun Jul 14 20:32:51 CEST 2013
On 7/14/13, Michael Niedermayer <michaelni at gmx.at> wrote:
> On Sun, Jul 14, 2013 at 05:20:57PM +0000, Paul B Mahol wrote:
>> On 7/14/13, Michael Niedermayer <michaelni at gmx.at> wrote:
>> > This also uses the decoders yuy type if "auto" is specified
>> >
>> > Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
>> > ---
>> > doc/filters.texi | 30 ++++++++++++++++++++++++
>> > libavfilter/vf_scale.c | 61
>> > ++++++++++++++++++++++++++++++++++++++++++++++++
>> > 2 files changed, 91 insertions(+)
>> >
>> > diff --git a/doc/filters.texi b/doc/filters.texi
>> > index 6f3994a..579f28b 100644
>> > --- a/doc/filters.texi
>> > +++ b/doc/filters.texi
>> > @@ -6210,6 +6210,36 @@ applies a bilinear scaling algorithm.
>> > @item size, s
>> > Set the video size, the value must be a valid abbreviation or in the
>> > form @var{width}x at var{height}.
>> > +
>> > + at item inyuvtype
>> > + at item outyuvtype
>> > +Set in/output YCbCr colorspace type.
>> > +This allows the autodetected value to be overridden as well as allows
>> > forcing
>> > +a specific value used for the output and encoder.
>> > +If not specified, the colorspace type depends on the pixel format.
>> > + at table @samp
>> > + at item auto
>> > +Choose automatically
>> > +
>> > + at item jpeg
>> > +Full 0-255 jpeg style
>> > +
>> > + at item bt709
>> > +ITU Rec BT709
>> > +
>> > + at item fcc
>> > +United States Federal Communications Commission Title 47 Code of
>> > +Federal Regulations (2003) 73.682 (a)
>> > +
>> > + at item bt601
>> > +ITU Rec BT601
>> > +ITU-R Rec. BT.470-6 System B, G
>> > +Society of Motion Picture and Television Engineers 170M (2004)
>> > +
>> > + at item smpte240m
>> > +Society of Motion Picture and Television Engineers 240M
>> > + at end table
>> > +
>> > @end table
>> >
>> > The values of the @var{w} and @var{h} options are expressions
>> > diff --git a/libavfilter/vf_scale.c b/libavfilter/vf_scale.c
>> > index b1246fe..4625167 100644
>> > --- a/libavfilter/vf_scale.c
>> > +++ b/libavfilter/vf_scale.c
>> > @@ -90,6 +90,9 @@ typedef struct {
>> > char *w_expr; ///< width expression string
>> > char *h_expr; ///< height expression string
>> > char *flags_str;
>> > +
>> > + char *inyuvtype;
>> > + char *outyuvtype;
>> > } ScaleContext;
>> >
>> > static av_cold int init(AVFilterContext *ctx)
>> > @@ -182,6 +185,40 @@ static int query_formats(AVFilterContext *ctx)
>> > return 0;
>> > }
>> >
>> > +static const int *parse_yuv_type(const char *s, int *full, enum
>> > AVColorSpace colorspace)
>> > +{
>> > + const static int32_t yuv2rgb_coeffs[8][4] = {
>> > + {},
>> > + { 117504, 138453, 13954, 34903 }, /* ITU-R Rec. 709 (1990) */
>> > + { 104597, 132201, 25675, 53279 }, /* unspecified */
>> > + { 104597, 132201, 25675, 53279 }, /* reserved */
>> > + { 104448, 132798, 24759, 53109 }, /* FCC */
>> > + { 104597, 132201, 25675, 53279 }, /* ITU-R Rec. 624-4 System B,
>> > G
>> > */
>> > + { 104597, 132201, 25675, 53279 }, /* SMPTE 170M */
>> > + { 117579, 136230, 16907, 35559 } /* SMPTE 240M (1987) */
>> > + };
>> > + if (!s)
>> > + s = "bt601";
>> > +
>> > + if (s && (strstr(s, "full") || strstr(s, "jpeg")))
>> > + *full = 1;
>> > + if (s && strstr(s, "bt709")) {
>> > + colorspace = AVCOL_SPC_BT709;
>> > + } else if (s && strstr(s, "fcc")) {
>> > + colorspace = AVCOL_SPC_FCC;
>> > + } else if (s && strstr(s, "smpte240m")) {
>> > + colorspace = AVCOL_SPC_SMPTE240M;
>> > + } else if (s && (strstr(s, "bt601") || strstr(s, "bt470") ||
>> > strstr(s,
>> > "smpte170m"))) {
>> > + colorspace = AVCOL_SPC_BT470BG;
>> > + }
>> > +
>> > + if (colorspace < 1 || colorspace > 7) {
>> > + colorspace = AVCOL_SPC_BT470BG;
>> > + }
>> > +
>> > + return yuv2rgb_coeffs[colorspace];
>> > +}
>> > +
>> > static int config_props(AVFilterLink *outlink)
>> > {
>> > AVFilterContext *ctx = outlink->src;
>> > @@ -370,6 +407,28 @@ static int filter_frame(AVFilterLink *link,
>> > AVFrame
>> > *in)
>> > if(scale->output_is_pal)
>> > avpriv_set_systematic_pal2((uint32_t*)out->data[1],
>> > outlink->format
>> > == AV_PIX_FMT_PAL8 ? AV_PIX_FMT_BGR8 : outlink->format);
>> >
>> > + if (scale->inyuvtype || scale->outyuvtype) {
>> > + int in_full, out_full, brightness, contrast, saturation;
>> > + const int *inv_table, *table;
>> > +
>> > + sws_getColorspaceDetails(scale->sws, (int **)&inv_table,
>> > &in_full,
>> > + (int **)&table, &out_full,
>> > + &brightness, &contrast, &saturation);
>> > +
>> > + inv_table = parse_yuv_type(scale->inyuvtype, &in_full,
>> > av_frame_get_colorspace(in));
>> > + table = parse_yuv_type(scale->outyuvtype, &out_full,
>> > AVCOL_SPC_UNSPECIFIED);
>> > +
>> > + sws_setColorspaceDetails(scale->sws, inv_table, in_full,
>> > + table, out_full,
>> > + brightness, contrast, saturation);
>> > + sws_setColorspaceDetails(scale->isws[0], inv_table, in_full,
>> > + table, out_full,
>> > + brightness, contrast, saturation);
>> > + sws_setColorspaceDetails(scale->isws[1], inv_table, in_full,
>> > + table, out_full,
>> > + brightness, contrast, saturation);
>> > + }
>> > +
>> > av_reduce(&out->sample_aspect_ratio.num,
>> > &out->sample_aspect_ratio.den,
>> > (int64_t)in->sample_aspect_ratio.num * outlink->h *
>> > link->w,
>> > (int64_t)in->sample_aspect_ratio.den * outlink->w *
>> > link->h,
>> > @@ -398,6 +457,8 @@ static const AVOption scale_options[] = {
>> > { "interl", "set interlacing", OFFSET(interlaced),
>> > AV_OPT_TYPE_INT,
>> > {.i64 = 0 }, -1, 1, FLAGS },
>> > { "size", "set video size", OFFSET(size_str),
>> > AV_OPT_TYPE_STRING, {.str = NULL}, 0, FLAGS },
>> > { "s", "set video size", OFFSET(size_str),
>> > AV_OPT_TYPE_STRING, {.str = NULL}, 0, FLAGS },
>> > + { "inyuvtype", "set input YCbCr type", OFFSET(inyuvtype),
>> > AV_OPT_TYPE_STRING, { .str = NULL }, .flags = FLAGS },
>> > + { "outyuvtype", "set output YCbCr type", OFFSET(outyuvtype),
>> > AV_OPT_TYPE_STRING, { .str = NULL }, .flags = FLAGS },
>>
>> As stefano already pointed out use named constans so user can see what can
>> be
>> set for in/out yuv type.
>
> as already said this prevents other matrixes from being
> specified.
> that is like "0.299, 0.123, 0.555"
Why you think that? Other matrixes are other matrixes, and you should have
special option to specify them.
>
>
>>
>> As colorspaces are already ints, you avoid useless use of strstr in code.
>
> without strstr or equivalent we need 4 instead of 2 parameters
> that change can be done of course if people prefer it.
I'm confused, what 4 parameters?
>
> [...]
>
> --
> Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB
>
> Freedom in capitalist society always remains about the same as it was in
> ancient Greek republics: Freedom for slave owners. -- Vladimir Lenin
>
More information about the ffmpeg-devel
mailing list