[FFmpeg-devel] avfilter: add ANSNR filter

A new filter ANSNR is added. libavfilter/Makefile is changed.
Run 'ffmpeg -i input1 -i input2 -lavfi ansnr -f null -' to get an overall score while per-frame value is stored but not printed.
This implementation is for constructing the vmaf filter later since ANSNR is one of individual tools used in vmaf.

Signed-off-by: Betty Wu <lumosomul@gmail.com>
---
 libavfilter/Makefile     |   1 +
 libavfilter/allfilters.c |   1 +
 libavfilter/vf_ansnr.c   | 425 +++++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 427 insertions(+)
 create mode 100644 libavfilter/vf_ansnr.c

2017-04-03 23:00 GMT+08:00 Betty Wu <lumosomul@gmail.com>:

> A new filter ANSNR is added. libavfilter/Makefile is changed.
> Run 'ffmpeg -i input1 -i input2 -lavfi ansnr -f null -' to get an overall
> score while per-frame value is stored but not printed.
> This implementation is for constructing the vmaf filter later since ANSNR
> is one of individual tools used in vmaf.
>
> Signed-off-by: Betty Wu <lumosomul@gmail.com>
> ---
>  libavfilter/Makefile     |   1 +
>  libavfilter/allfilters.c |   1 +
>  libavfilter/vf_ansnr.c   | 425 ++++++++++++++++++++++++++++++
> +++++++++++++++++
>  3 files changed, 427 insertions(+)
>  create mode 100644 libavfilter/vf_ansnr.c
>
> diff --git a/libavfilter/Makefile b/libavfilter/Makefile
> index 9c15ed62d2..5416e0f34f 100644
> --- a/libavfilter/Makefile
> +++ b/libavfilter/Makefile
> @@ -123,6 +123,7 @@ OBJS-$(CONFIG_ANULLSINK_FILTER)              +=
> asink_anullsink.o
>  # video filters
>  OBJS-$(CONFIG_ALPHAEXTRACT_FILTER)           += vf_extractplanes.o
>  OBJS-$(CONFIG_ALPHAMERGE_FILTER)             += vf_alphamerge.o
> +OBJS-$(CONFIG_ANSNR_FILTER)                  += vf_ansnr.o dualinput.o
> framesync.o
>  OBJS-$(CONFIG_ASS_FILTER)                    += vf_subtitles.o
>  OBJS-$(CONFIG_ATADENOISE_FILTER)             += vf_atadenoise.o
>  OBJS-$(CONFIG_AVGBLUR_FILTER)                += vf_avgblur.o
> diff --git a/libavfilter/allfilters.c b/libavfilter/allfilters.c
> index 64b634e8f3..fbd9ec026e 100644
> --- a/libavfilter/allfilters.c
> +++ b/libavfilter/allfilters.c
> @@ -134,6 +134,7 @@ static void register_all(void)
>
>      REGISTER_FILTER(ALPHAEXTRACT,   alphaextract,   vf);
>      REGISTER_FILTER(ALPHAMERGE,     alphamerge,     vf);
> +    REGISTER_FILTER(ANSNR,          ansnr,          vf);
>      REGISTER_FILTER(ASS,            ass,            vf);
>      REGISTER_FILTER(ATADENOISE,     atadenoise,     vf);
>      REGISTER_FILTER(AVGBLUR,        avgblur,        vf);
> diff --git a/libavfilter/vf_ansnr.c b/libavfilter/vf_ansnr.c
> new file mode 100644
> index 0000000000..cfea6efbd4
> --- /dev/null
> +++ b/libavfilter/vf_ansnr.c
> @@ -0,0 +1,425 @@
> +/*
> + * Copyright (c) 2017 Betty Wu
> + *
> + * This file is part of FFmpeg.
> + *
> + * FFmpeg is free software; you can redistribute it and/or
> + * modify it under the terms of the GNU Lesser General Public
> + * License as published by the Free Software Foundation; either
> + * version 2.1 of the License, or (at your option) any later version.
> + *
> + * FFmpeg is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
> + * Lesser General Public License for more details.
> + *
> + * You should have received a copy of the GNU Lesser General Public
> + * License along with FFmpeg; if not, write to the Free Software
> + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
> 02110-1301 USA
> + */
> +
> +/*
> + * @file
> + * Caculate the ANSNR between two input videos.
> + * @author Betty Wu
> + */
> +
> +#include "libavutil/avstring.h"
> +#include "libavutil/opt.h"
> +#include "libavutil/pixdesc.h"
> +#include "avfilter.h"
> +#include "dualinput.h"
> +#include "drawutils.h"
> +#include "formats.h"
> +#include "internal.h"
> +#include "video.h"
> +
> +#define OPT_RANGE_PIXEL_OFFSET -128
> +
> +typedef double number_t;
> +
> +typedef struct ANSNRContext {
> +    const AVClass *class;
> +    FFDualInputContext dinput;
> +    uint64_t nb_frames;
> +    FILE *stats_file;
> +    char *stats_file_str;
> +    int stats_version;
> +    int stats_header_written;
> +    int stats_add_max;
> +    int is_rgb;
> +    uint8_t rgba_map[4];
> +    double score;
> +    double score_total;
> +    char comps[4];
> +    int nb_components;
> +    int planewidth[4];
> +    int planeheight[4];
> +
> +} ANSNRContext;
> +#define OFFSET(x) offsetof(ANSNRContext, x)
> +#define FLAGS AV_OPT_FLAG_FILTERING_PARAM|AV_OPT_FLAG_VIDEO_PARAM
> +
> +static const AVOption ansnr_options[] = {
> +    {"stats_file", "Set file where to store per-frame difference
> information", OFFSET(stats_file_str), AV_OPT_TYPE_STRING, {.str=NULL}, 0,
> 0, FLAGS },
> +    {"f",          "Set file where to store per-frame difference
> information", OFFSET(stats_file_str), AV_OPT_TYPE_STRING, {.str=NULL}, 0,
> 0, FLAGS },
> +    { NULL }
> +};
>
Document please.

> +AVFILTER_DEFINE_CLASS(ansnr);
> +
> +const int ans_filter3_stride = 3;
> +const int ans_filter5_stride = 5;
> +
> +const double ans_filter3[ans_filter3_stride*ans_filter3_stride] = {
> +    1.0 / 16.0, 2.0 / 16.0, 1.0 / 16.0,
> +    2.0 / 16.0, 4.0 / 16.0, 2.0 / 16.0,
> +    1.0 / 16.0, 2.0 / 16.0, 1.0 / 16.0
> +};
> +
> +const double ans_filter5[ans_filter5_stride*ans_filter5_stride] = {
> +    2.0 / 571.0,  7.0 / 571.0,  12.0 / 571.0,  7.0 / 571.0,  2.0 / 571.0,
> +    7.0 / 571.0,  31.0 / 571.0, 52.0 / 571.0,  31.0 / 571.0, 7.0 / 571.0,
> +    12.0 / 571.0, 52.0 / 571.0, 127.0 / 571.0, 52.0 / 571.0, 12.0 / 571.0,
> +    7.0 / 571.0,  31.0 / 571.0, 52.0 / 571.0,  31.0 / 571.0, 7.0 / 571.0,
> +    2.0 / 571.0,  7.0 / 571.0,  12.0 / 571.0,  7.0 / 571.0,  2.0 / 571.0
> +};
> +
> +static inline number_t pow_2(number_t base)
> +{
> +    return base*base;
> +}
> +
> +static inline double get_ansnr(double score)
> +{
> +    return 10.0 * log10(score);
> +}
> +
> +static inline double get_pow(double score)
> +{
> +    return pow(10, score/10.0);
> +}
> +
> +static void ansnr_mse(const number_t *ref, const number_t *dis, number_t
> *img_sig, number_t *img_noise, int row, int col)
> +{
> +    int ref_stride = col;
> +    int dis_stride = col;
> +
> +    *img_sig = 0;
> +    *img_noise = 0;
> +
> +    for (int i = 0; i < row; ++i) {
> +        for (int j = 0; j < col; ++j) {
> +            int ref_index = i * ref_stride + j;
> +            int dis_index = i * dis_stride + j;
> +
> +            *img_sig   += pow_2(ref[ref_index]);
> +            *img_noise += pow_2(ref[ref_index] - dis[dis_index]);
> +        }
> +    }
> +}
> +
> +static int ansnr_filter(const uint8_t* src_image, number_t* dst_image,
> int img_row, int img_col, const double *filter, int stride)
> +{
> +    int filter_row = stride;
> +    int filter_col = stride;
> +    int start_row = filter_row / 2;
> +    int start_col = filter_col / 2;
> +    int imme_row = img_row + 2 * start_row;
> +    int imme_col = img_col + 2 * start_col;
> +
> +    number_t **imme_image;
> +    if(!(imme_image = (number_t **)av_malloc((size_t)imme_row *
> sizeof(number_t *))))
> +        return AVERROR(ENOMEM);
> +
> +    for(int i = 0; i < imme_row; i++) {
> +        if(!(imme_image[i] = (number_t *)av_malloc((size_t)imme_col *
> sizeof(number_t))))
> +            return AVERROR(ENOMEM);
>
memleak

> +    }
> +
> +    for (int i = 0; i < imme_row; i++) {
> +        for (int j = 0; j < imme_col; j++) {
> +            int src_i = i - start_row;
> +            int src_j = j - start_col;
> +
> +            src_i = abs(src_i) >= img_row? 2 * img_row - abs(src_i) - 1 :
> abs(src_i);
> +            src_j = abs(src_j) >= img_col? 2 * img_col - abs(src_j) - 1 :
> abs(src_j);
> +
> +            imme_image[i][j] = (number_t)src_image[src_i * img_col +
> src_j] + OPT_RANGE_PIXEL_OFFSET;
> +        }
> +    }
> +
> +    int end_row = imme_row -1 - start_row ;
> +    int end_col = imme_col -1 - start_col ;
> +
> +    for (int i = start_row; i <= end_row; i++) {
> +        for (int j = start_col; j <= end_col; j++) {
> +            int dst_i = i - start_row  ;
> +            int dst_j = j - start_col  ;
> +            int dst_index = dst_i * img_col + dst_j;
> +            dst_image[dst_index] = 0;
> +
> +            for (int m = 0; m < filter_row; m++) {
> +                for (int n = 0; n < filter_col; n++) {
> +                    dst_image[dst_index] += (imme_image[dst_i + m][dst_j
> + n] * filter[m * filter_col + n]);
> +                }
> +            }
> +        }
> +    }
> +
> +    for(int i = 0; i < imme_row; i++)
> +        av_free(imme_image[i]);
> +    av_free(imme_image);
> +
> +    return 0;
> +}
> +
> +static inline
> +int compute_ansnr(const uint8_t *anc, const uint8_t *dis, int col, int
> row, double psnr_max, double *score, double *out)
> +{
> +    *score = 0.0;
> +    *out = 0.0;
> +
> +    number_t *anc_after;
> +    if(!(anc_after = (number_t *)av_malloc((size_t)row * col *
> sizeof(number_t)))) {
> +        av_free(anc_after);
> +        return AVERROR(ENOMEM);
> +    }
> +
> +    number_t *dis_after;
> +    if(!(dis_after = (number_t *)av_malloc((size_t)row * col *
> sizeof(number_t)))) {
> +        av_free(dis_after);
>
memleak; add av_free(anc_after);

> +        return AVERROR(ENOMEM);
> +    }
> +
> +    int ret;
> +    ret = ansnr_filter(anc, anc_after, row, col, ans_filter3,
> ans_filter3_stride);
> +    if(ret < 0)
> +        return ret;
>
memleak!

> +    ret = ansnr_filter(dis, dis_after, row, col, ans_filter5,
> ans_filter5_stride);
> +    if(ret < 0)
> +        return ret;
>
memleak!

> +
> +    number_t sig, noise;
> +
> +    ansnr_mse(anc_after, dis_after, &sig, &noise, row, col);
> +    *score = noise==0 ? psnr_max : get_ansnr((double)sig / noise);
>
*score = !noise ? get_ansnr((double)sig / noise) : psnr_max;

> +    *out = get_pow(*score);
> +
> +    av_free(anc_after);
> +    av_free(dis_after);
> +
> +    return 0;
> +}
> +
> +static void set_meta(AVDictionary **metadata, const char *key, char comp,
> float d)
> +{
> +    char value[128];
> +    snprintf(value, sizeof(value), "%0.2f", d);
> +    if (comp) {
> +        char key2[128];
> +        snprintf(key2, sizeof(key2), "%s%c", key, comp);
> +        av_dict_set(metadata, key2, value, 0);
> +    } else {
> +        av_dict_set(metadata, key, value, 0);
> +    }
> +}
> +
> +static int query_formats(AVFilterContext *ctx)
> +{
> +    static const enum AVPixelFormat pix_fmts[] = {
> +        AV_PIX_FMT_GRAY8, AV_PIX_FMT_GRAY16,
> +#define PF_NOALPHA(suf) AV_PIX_FMT_YUV420##suf,  AV_PIX_FMT_YUV422##suf,
> AV_PIX_FMT_YUV444##suf
> +#define PF_ALPHA(suf)   AV_PIX_FMT_YUVA420##suf, AV_PIX_FMT_YUVA422##suf,
> AV_PIX_FMT_YUVA444##suf
> +#define PF(suf)         PF_NOALPHA(suf), PF_ALPHA(suf)
> +        PF(P), PF(P9), PF(P10), PF_NOALPHA(P12), PF_NOALPHA(P14), PF(P16),
> +        AV_PIX_FMT_YUV440P, AV_PIX_FMT_YUV411P, AV_PIX_FMT_YUV410P,
> +        AV_PIX_FMT_YUVJ411P, AV_PIX_FMT_YUVJ420P, AV_PIX_FMT_YUVJ422P,
> +        AV_PIX_FMT_YUVJ440P, AV_PIX_FMT_YUVJ444P,
> +        AV_PIX_FMT_GBRP, AV_PIX_FMT_GBRP9, AV_PIX_FMT_GBRP10,
> +        AV_PIX_FMT_GBRP12, AV_PIX_FMT_GBRP14, AV_PIX_FMT_GBRP16,
> +        AV_PIX_FMT_GBRAP, AV_PIX_FMT_GBRAP16,
> +        AV_PIX_FMT_NONE
> +    };
> +
> +    AVFilterFormats *fmts_list = ff_make_format_list(pix_fmts);
> +    if (!fmts_list)
> +        return AVERROR(ENOMEM);
> +    return ff_set_common_formats(ctx, fmts_list);
> +}
> +
> +static AVFrame *do_ansnr(AVFilterContext *ctx, AVFrame *main,
> +                        const AVFrame *ref)
> +{
> +    ANSNRContext *s = ctx->priv;
> +    AVDictionary **metadata = avpriv_frame_get_metadatap(main);
> +    double psnr_max;
> +    char *frame_format = (char *)av_get_pix_fmt_name(main->format);
> +
> +    if( !strcmp(frame_format,"yuv420p") || !strcmp(frame_format,"yuv422p")
> || !strcmp(frame_format,"yuv444p"))
> +        psnr_max = 60;
> +    if( !strcmp(frame_format,"yuv420p10le") || !strcmp(frame_format,"yuv422p10le")
> || !strcmp(frame_format,"yuv444p10le"))
> +        psnr_max = 72;
>
use av_strcasecmp;

> +
> +    double score = 0.0;
> +    double out = 0.0;
> +
> +    compute_ansnr(ref->data[0], main->data[0], s->planewidth[0],
> s->planeheight[0], psnr_max, &score, &out);
> +    s->nb_frames++;
> +    s->score = score;
> +    s->score_total += out;
> +
> +    set_meta(metadata, "lavfi.ansnr.All", 0, score);
> +
> +    if (s->stats_file) {
> +        fprintf(s->stats_file, "n:%"PRId64" ", s->nb_frames);
> +        fprintf(s->stats_file, "All:%f\n", score);
> +    }
> +
> +    return main;
> +}
> +
> +static av_cold int init(AVFilterContext *ctx)
> +{
> +    ANSNRContext *s = ctx->priv;
> +    s->score_total = 0;
> +    s->score = 0;
> +
> +    if (s->stats_file_str) {
> +        if (s->stats_version < 2 && s->stats_add_max) {
> +            av_log(ctx, AV_LOG_ERROR,
> +                "stats_add_max was specified but stats_version < 2.\n" );
> +            return AVERROR(EINVAL);
> +        }
> +
> +        if (!strcmp(s->stats_file_str, "-")) {
>
use av_strcasecmp

> +            s->stats_file = stdout;
> +        } else {
> +            s->stats_file = fopen(s->stats_file_str, "w");
> +            if (!s->stats_file) {
> +                int err = AVERROR(errno);
> +                char buf[128];
> +                av_strerror(err, buf, sizeof(buf));
> +                av_log(ctx, AV_LOG_ERROR, "Could not open stats file %s:
> %s\n",
> +                       s->stats_file_str, buf);
> +                return err;
> +            }
> +        }
> +    }
> +
> +    s->dinput.process = do_ansnr;
> +    return 0;
> +}
> +
> +static int config_input_ref(AVFilterLink *inlink)
> +{
> +    const AVPixFmtDescriptor *desc = av_pix_fmt_desc_get(inlink->format);
> +    AVFilterContext *ctx  = inlink->dst;
> +    ANSNRContext *s = ctx->priv;
> +
> +    s->nb_components = desc->nb_components;
> +
> +    if (ctx->inputs[0]->w != ctx->inputs[1]->w ||
> +        ctx->inputs[0]->h != ctx->inputs[1]->h) {
> +        av_log(ctx, AV_LOG_ERROR, "Width and height of input videos must
> be same.\n");
> +        return AVERROR(EINVAL);
> +    }
> +    if (ctx->inputs[0]->format != ctx->inputs[1]->format) {
> +        av_log(ctx, AV_LOG_ERROR, "Inputs must be of same pixel
> format.\n");
> +        return AVERROR(EINVAL);
> +    }
> +    s->is_rgb = ff_fill_rgba_map(s->rgba_map, inlink->format) >= 0;
> +    s->comps[0] = s->is_rgb ? 'r' : 'y' ;
> +    s->comps[1] = s->is_rgb ? 'g' : 'u' ;
> +    s->comps[2] = s->is_rgb ? 'b' : 'v' ;
> +    s->comps[3] = 'a';
> +
> +    s->planeheight[1] = s->planeheight[2] = AV_CEIL_RSHIFT(inlink->h,
> desc->log2_chroma_h);
> +    s->planeheight[0] = s->planeheight[3] = inlink->h;
> +    s->planewidth[1]  = s->planewidth[2]  = AV_CEIL_RSHIFT(inlink->w,
> desc->log2_chroma_w);
> +    s->planewidth[0]  = s->planewidth[3]  = inlink->w;
> +
> +    return 0;
> +}
> +
> +static int config_output(AVFilterLink *outlink)
> +{
> +    AVFilterContext *ctx = outlink->src;
> +    ANSNRContext *s = ctx->priv;
> +    AVFilterLink *mainlink = ctx->inputs[0];
> +    int ret;
> +
> +    outlink->w = mainlink->w;
> +    outlink->h = mainlink->h;
> +    outlink->time_base = mainlink->time_base;
> +    outlink->sample_aspect_ratio = mainlink->sample_aspect_ratio;
> +    outlink->frame_rate = mainlink->frame_rate;
> +
> +    if ((ret = ff_dualinput_init(ctx, &s->dinput)) < 0)
> +        return ret;
> +
> +    return 0;
> +}
> +
> +static int filter_frame(AVFilterLink *inlink, AVFrame *inpicref)
> +{
> +    ANSNRContext *s = inlink->dst->priv;
> +    return ff_dualinput_filter_frame(&s->dinput, inlink, inpicref);
> +}
> +
> +static int request_frame(AVFilterLink *outlink)
> +{
> +    ANSNRContext *s = outlink->src->priv;
> +    return ff_dualinput_request_frame(&s->dinput, outlink);
> +}
> +
> +static av_cold void uninit(AVFilterContext *ctx)
> +{
> +    ANSNRContext *s = ctx->priv;
> +
> +    if (s->nb_frames > 0) {
> +        char buf[256];
> +        buf[0] = 0;
> +        av_log(ctx, AV_LOG_INFO, "ANSNR%s All:%f\n", buf,
> get_ansnr(s->score_total));
> +    }
> +
> +    ff_dualinput_uninit(&s->dinput);
> +
> +    if (s->stats_file && s->stats_file != stdout)
> +        fclose(s->stats_file);
> +}
> +
> +static const AVFilterPad ansnr_inputs[] = {
> +    {
> +        .name         = "main",
> +        .type         = AVMEDIA_TYPE_VIDEO,
> +        .filter_frame = filter_frame,
> +    },{
> +        .name         = "reference",
> +        .type         = AVMEDIA_TYPE_VIDEO,
> +        .filter_frame = filter_frame,
> +        .config_props = config_input_ref,
> +    },
> +    { NULL }
> +};
> +
> +static const AVFilterPad ansnr_outputs[] = {
> +    {
> +        .name          = "default",
> +        .type          = AVMEDIA_TYPE_VIDEO,
> +        .config_props  = config_output,
> +        .request_frame = request_frame,
> +    },
> +    { NULL }
> +};
> +
> +AVFilter ff_vf_ansnr = {
> +    .name          = "ansnr",
> +    .description   = NULL_IF_CONFIG_SMALL("Calculate the ANSNR between
> two video streams."),
> +    .init          = init,
> +    .uninit        = uninit,
> +    .query_formats = query_formats,
> +    .priv_size     = sizeof(ANSNRContext),
> +    .priv_class    = &ansnr_class,
> +    .inputs        = ansnr_inputs,
> +    .outputs       = ansnr_outputs,
> +};
> --
> 2.12.1
>
> _______________________________________________
> ffmpeg-devel mailing list
> ffmpeg-devel@ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
>

Hi Betty,

On Mon, Apr 3, 2017 at 11:00 AM, Betty Wu <lumosomul@gmail.com> wrote:

> +typedef double number_t;
>

Why?

+static int ansnr_filter(const uint8_t* src_image, number_t* dst_image, int
> img_row, int img_col, const double *filter, int stride)
>
[..]

> +    number_t **imme_image;
> +    if(!(imme_image = (number_t **)av_malloc((size_t)imme_row *
> sizeof(number_t *))))
> +        return AVERROR(ENOMEM);
> +
> +    for(int i = 0; i < imme_row; i++) {
> +        if(!(imme_image[i] = (number_t *)av_malloc((size_t)imme_col *
> sizeof(number_t))))
> +            return AVERROR(ENOMEM);
> +    }
>

In ffmpeg, we typically don't allocate lines and cols in separate arrays.
You can just use a 1D array and use y*stride+x for indexing, which is what
we do elsewhere.

These arrays should also be initialized once in the init function and then
reuse, instead of being re-allocated for each frame.

+    number_t *anc_after;
> +    if(!(anc_after = (number_t *)av_malloc((size_t)row * col *
> sizeof(number_t)))) {
> +        av_free(anc_after);
> +        return AVERROR(ENOMEM);
> +    }
>

Why free if allocation failed? And this also should be allocated once
during init, not re-allocated for each frame.


> +static int query_formats(AVFilterContext *ctx)
> +{
> +    static const enum AVPixelFormat pix_fmts[] = {
> +        AV_PIX_FMT_GRAY8, AV_PIX_FMT_GRAY16,
> +#define PF_NOALPHA(suf) AV_PIX_FMT_YUV420##suf,  AV_PIX_FMT_YUV422##suf,
> AV_PIX_FMT_YUV444##suf
> +#define PF_ALPHA(suf)   AV_PIX_FMT_YUVA420##suf, AV_PIX_FMT_YUVA422##suf,
> AV_PIX_FMT_YUVA444##suf
> +#define PF(suf)         PF_NOALPHA(suf), PF_ALPHA(suf)
> +        PF(P), PF(P9), PF(P10), PF_NOALPHA(P12), PF_NOALPHA(P14), PF(P16),
> +        AV_PIX_FMT_YUV440P, AV_PIX_FMT_YUV411P, AV_PIX_FMT_YUV410P,
> +        AV_PIX_FMT_YUVJ411P, AV_PIX_FMT_YUVJ420P, AV_PIX_FMT_YUVJ422P,
> +        AV_PIX_FMT_YUVJ440P, AV_PIX_FMT_YUVJ444P,
> +        AV_PIX_FMT_GBRP, AV_PIX_FMT_GBRP9, AV_PIX_FMT_GBRP10,
> +        AV_PIX_FMT_GBRP12, AV_PIX_FMT_GBRP14, AV_PIX_FMT_GBRP16,
> +        AV_PIX_FMT_GBRAP, AV_PIX_FMT_GBRAP16,
> +        AV_PIX_FMT_NONE
> +    };
>

Have all of these been tested? Since ANSNR is Y-only, it doesn't work on
GBR. Likewise, I believe the code is 8-bit only so it shouldn't work on
gray16 or any of the P9/10/12/14/16 formats.

For SIMD purposes, I would probably encourage you to work in fixed-point
integer. Maybe for the qualification task we can skip that (I hadn't really
thought about it), but I think for the final implementation, fixed-point
will be much faster, and speed is a significant goal here. An interesting
idea for speed is to make the filter decomposable (i.e. split the
horizontal/vertical pass) if that's possible. At least for the 3-tap
filter, that should be trivial (it's just a 121 lowpass in both
directions), and should give some speed gains because you go from n^2 to 2n
+ an extra load/store pair in terms of complexity (whether it's actually
faster remains to be proven). The other filter should also be decomposable
but I don't see from the top of my head what it decomposes into and I don't
feel like writing a script to figure it out. :-).

Ronald