Re: [FFmpeg-devel] [FFmpeg-cvslog] avcodec: add QOA decoder

* Re: [FFmpeg-devel] [FFmpeg-cvslog] avcodec: add QOA decoder
       [not found] <20231126164217.65B274100B2@natalya.videolan.org>
@ 2023-11-26 18:02 ` James Almer
  2023-12-02 13:50 ` Anton Khirnov
  1 sibling, 0 replies; 7+ messages in thread
From: James Almer @ 2023-11-26 18:02 UTC (permalink / raw)
  To: ffmpeg-devel

> ffmpeg | branch: master | Paul B Mahol <onemda at gmail.com  <https://ffmpeg.org/mailman/listinfo/ffmpeg-cvslog>> | Sat Sep 23 16:49:25 2023 +0200| [3609d2b78340c06920973dc46b85b517c789782f] | committer: Paul B Mahol
>
> avcodec: add QOA decoder
>
> >/http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=3609d2b78340c06920973dc46b85b517c789782f 
> /---
>
>   Changelog               |   1 +
>   libavcodec/Makefile     |   1 +
>   libavcodec/allcodecs.c  |   1 +
>   libavcodec/codec_desc.c |   7 ++
>   libavcodec/codec_id.h   |   1 +
>   libavcodec/qoadec.c     | 175 ++++++++++++++++++++++++++++++++++++++++++++++++
>   libavcodec/version.h    |   2 +-
>   7 files changed, 187 insertions(+), 1 deletion(-)
>
> diff --git a/Changelog b/Changelog
> index 7d79be1c3e..d945904bc0 100644
> --- a/Changelog
> +++ b/Changelog
> @@ -5,6 +5,7 @@ version <next>:
>   - LEAD MCMP decoder
>   - EVC decoding using external library libxevd
>   - EVC encoding using external library libxeve
> +- QOA decoder
>   
>   version 6.1:
>   - libaribcaption decoder
> diff --git a/libavcodec/Makefile b/libavcodec/Makefile
> index a5941d1284..748806e702 100644
> --- a/libavcodec/Makefile
> +++ b/libavcodec/Makefile
> @@ -623,6 +623,7 @@ OBJS-$(CONFIG_QCELP_DECODER)           += qcelpdec.o                     \
>   OBJS-$(CONFIG_QDM2_DECODER)            += qdm2.o
>   OBJS-$(CONFIG_QDMC_DECODER)            += qdmc.o
>   OBJS-$(CONFIG_QDRAW_DECODER)           += qdrw.o
> +OBJS-$(CONFIG_QOA_DECODER)             += qoadec.o
>   OBJS-$(CONFIG_QOI_DECODER)             += qoidec.o
>   OBJS-$(CONFIG_QOI_ENCODER)             += qoienc.o
>   OBJS-$(CONFIG_QPEG_DECODER)            += qpeg.o
> diff --git a/libavcodec/allcodecs.c b/libavcodec/allcodecs.c
> index 87595683f9..b0f004e15c 100644
> --- a/libavcodec/allcodecs.c
> +++ b/libavcodec/allcodecs.c
> @@ -522,6 +522,7 @@ extern const FFCodec ff_paf_audio_decoder;
>   extern const FFCodec ff_qcelp_decoder;
>   extern const FFCodec ff_qdm2_decoder;
>   extern const FFCodec ff_qdmc_decoder;
> +extern const FFCodec ff_qoa_decoder;
>   extern const FFCodec ff_ra_144_encoder;
>   extern const FFCodec ff_ra_144_decoder;
>   extern const FFCodec ff_ra_288_decoder;
> diff --git a/libavcodec/codec_desc.c b/libavcodec/codec_desc.c
> index 432a9c9ea6..033344304c 100644
> --- a/libavcodec/codec_desc.c
> +++ b/libavcodec/codec_desc.c
> @@ -3427,6 +3427,13 @@ static const AVCodecDescriptor codec_descriptors[] = {
>           .long_name = NULL_IF_CONFIG_SMALL("OSQ (Original Sound Quality)"),
>           .props     = AV_CODEC_PROP_INTRA_ONLY | AV_CODEC_PROP_LOSSLESS,
>       },
> +    {
> +        .id        = AV_CODEC_ID_QOA,
> +        .type      = AVMEDIA_TYPE_AUDIO,
> +        .name      = "qoa",
> +        .long_name = NULL_IF_CONFIG_SMALL("QOA (Quite OK Audio)"),
> +        .props     = AV_CODEC_PROP_INTRA_ONLY | AV_CODEC_PROP_LOSSY,
> +    },
>   
>       /* subtitle codecs */
>       {
> diff --git a/libavcodec/codec_id.h b/libavcodec/codec_id.h
> index 84ce8e6aa9..d96e49430e 100644
> --- a/libavcodec/codec_id.h
> +++ b/libavcodec/codec_id.h
> @@ -545,6 +545,7 @@ enum AVCodecID {
>       AV_CODEC_ID_RKA,
>       AV_CODEC_ID_AC4,
>       AV_CODEC_ID_OSQ,
> +    AV_CODEC_ID_QOA,
>   
>       /* subtitle codecs */
>       AV_CODEC_ID_FIRST_SUBTITLE = 0x17000,          ///< A dummy ID pointing at the start of subtitle codecs.
> diff --git a/libavcodec/qoadec.c b/libavcodec/qoadec.c
> new file mode 100644
> index 0000000000..9b2abae833
> --- /dev/null
> +++ b/libavcodec/qoadec.c
> @@ -0,0 +1,175 @@
> +/*
> + * QOA decoder
> + *
> + * 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
> + */
> +
> +#include "avcodec.h"
> +#include "codec_internal.h"
> +#include "decode.h"
> +#include "get_bits.h"
> +#include "bytestream.h"
> +#include "mathops.h"
> +
> +#define QOA_SLICE_LEN 20
> +#define QOA_LMS_LEN 4
> +
> +typedef struct QOAChannel {
> +    int history[QOA_LMS_LEN];
> +    int weights[QOA_LMS_LEN];
> +} QOAChannel;
> +
> +typedef struct QOAContext {
> +    QOAChannel *ch;
> +} QOAContext;
> +
> +static const int16_t qoa_dequant_tab[16][8] = {
> +    {   1,    -1,    3,    -3,    5,    -5,     7,     -7},
> +    {   5,    -5,   18,   -18,   32,   -32,    49,    -49},
> +    {  16,   -16,   53,   -53,   95,   -95,   147,   -147},
> +    {  34,   -34,  113,  -113,  203,  -203,   315,   -315},
> +    {  63,   -63,  210,  -210,  378,  -378,   588,   -588},
> +    { 104,  -104,  345,  -345,  621,  -621,   966,   -966},
> +    { 158,  -158,  528,  -528,  950,  -950,  1477,  -1477},
> +    { 228,  -228,  760,  -760, 1368, -1368,  2128,  -2128},
> +    { 316,  -316, 1053, -1053, 1895, -1895,  2947,  -2947},
> +    { 422,  -422, 1405, -1405, 2529, -2529,  3934,  -3934},
> +    { 548,  -548, 1828, -1828, 3290, -3290,  5117,  -5117},
> +    { 696,  -696, 2320, -2320, 4176, -4176,  6496,  -6496},
> +    { 868,  -868, 2893, -2893, 5207, -5207,  8099,  -8099},
> +    {1064, -1064, 3548, -3548, 6386, -6386,  9933,  -9933},
> +    {1286, -1286, 4288, -4288, 7718, -7718, 12005, -12005},
> +    {1536, -1536, 5120, -5120, 9216, -9216, 14336, -14336},
> +};
> +
> +static av_cold int qoa_decode_init(AVCodecContext *avctx)
> +{
> +    QOAContext *s = avctx->priv_data;
> +
> +    avctx->sample_fmt = AV_SAMPLE_FMT_S16;
> +
> +    s->ch = av_calloc(avctx->ch_layout.nb_channels, sizeof(*s->ch));
> +    if (!s->ch)
> +        return AVERROR(ENOMEM);
> +
> +    return 0;
> +}
> +
> +static int qoa_lms_predict(QOAChannel *lms)
> +{
> +    int prediction = 0;
> +    for (int i = 0; i < QOA_LMS_LEN; i++)
> +        prediction += lms->weights[i] * lms->history[i];
> +    return prediction >> 13;
> +}
> +
> +static void qoa_lms_update(QOAChannel *lms, int sample, int residual)
> +{
> +    int delta = residual >> 4;
> +    for (int i = 0; i < QOA_LMS_LEN; i++)
> +        lms->weights[i] += lms->history[i] < 0 ? -delta : delta;
> +    for (int i = 0; i < QOA_LMS_LEN-1; i++)
> +        lms->history[i] = lms->history[i+1];
> +    lms->history[QOA_LMS_LEN-1] = sample;
> +}
> +
> +static int qoa_decode_frame(AVCodecContext *avctx, AVFrame *frame,
> +                            int *got_frame_ptr, AVPacket *avpkt)
> +{
> +    QOAContext *s = avctx->priv_data;
> +    int ret, frame_size, nb_channels;
> +    GetByteContext gb;
> +    int16_t *samples;
> +
> +    bytestream2_init(&gb, avpkt->data, avpkt->size);
> +
> +    nb_channels = bytestream2_get_byte(&gb);
> +    if (avctx->ch_layout.nb_channels != nb_channels)
> +        return AVERROR_INVALIDDATA;
> +
> +    avctx->sample_rate = bytestream2_get_be24(&gb);
> +    frame->nb_samples = bytestream2_get_be16(&gb);
> +    frame_size = bytestream2_get_be16(&gb);
> +    if (frame_size > avpkt->size)
> +        return AVERROR_INVALIDDATA;
> +
> +    if (frame_size < 8 + QOA_LMS_LEN * 4 * nb_channels +
> +        8LL * frame->nb_samples * nb_channels / QOA_SLICE_LEN)
> +        return AVERROR_INVALIDDATA;
> +
> +    if ((ret = ff_get_buffer(avctx, frame, 0)) < 0)
> +        return ret;
> +    samples = (int16_t *)frame->data[0];
> +
> +    for (int ch = 0; ch < nb_channels; ch++) {
> +        QOAChannel *qch = &s->ch[ch];
> +
> +        for (int n = 0; n < QOA_LMS_LEN; n++)
> +            qch->history[n] = sign_extend(bytestream2_get_be16u(&gb), 16);
> +        for (int n = 0; n < QOA_LMS_LEN; n++)
> +            qch->weights[n] = sign_extend(bytestream2_get_be16u(&gb), 16);
> +    }
> +
> +    for (int sample_index = 0; sample_index < frame->nb_samples * nb_channels;
> +         sample_index += QOA_SLICE_LEN) {
> +        for (int ch = 0; ch < nb_channels; ch++) {
> +            QOAChannel *lms = &s->ch[ch];
> +            uint64_t slice = bytestream2_get_be64u(&gb);
> +            int scalefactor = (slice >> 60) & 0xf;
> +            int slice_start = sample_index * nb_channels + ch;
> +            int slice_end = av_clip(sample_index + QOA_SLICE_LEN, 0, frame->nb_samples) * nb_channels + ch;
> +
> +            for (int si = slice_start; si < slice_end; si += nb_channels) {
> +                int predicted = qoa_lms_predict(lms);
> +                int quantized = (slice >> 57) & 0x7;
> +                int dequantized = qoa_dequant_tab[scalefactor][quantized];
> +                int reconstructed = av_clip_int16(predicted + dequantized);
> +
> +                samples[si] = reconstructed;
> +                slice <<= 3;
> +
> +                qoa_lms_update(lms, reconstructed, dequantized);
> +            }
> +        }
> +    }
> +
> +    *got_frame_ptr = 1;
> +
> +    return avpkt->size;
> +}
> +
> +static av_cold int qoa_decode_end(AVCodecContext *avctx)
> +{
> +    QOAContext *s = avctx->priv_data;
> +    av_freep(&s->ch);
> +    return 0;
> +}
> +
> +const FFCodec ff_qoa_decoder = {
> +    .p.name         = "qoa",
> +    CODEC_LONG_NAME("QOA (Quite OK Audio)"),
> +    .p.type         = AVMEDIA_TYPE_AUDIO,
> +    .p.id           = AV_CODEC_ID_QOA,
> +    .priv_data_size = sizeof(QOAContext),
> +    .init           = qoa_decode_init,
> +    FF_CODEC_DECODE_CB(qoa_decode_frame),
> +    .close          = qoa_decode_end,
> +    .p.capabilities = AV_CODEC_CAP_CHANNEL_CONF |

The decoder is not setting channel info, it's using the one set by the 
caller and ensuring it doesn't change between frames. So this cap is not 
correct.

> +                      AV_CODEC_CAP_DR1,
> +    .p.sample_fmts  = (const enum AVSampleFormat[]) { AV_SAMPLE_FMT_S16,
> +                                                      AV_SAMPLE_FMT_NONE },
> +};
> diff --git a/libavcodec/version.h b/libavcodec/version.h
> index 0ef6c991f3..1008fead27 100644
> --- a/libavcodec/version.h
> +++ b/libavcodec/version.h
> @@ -29,7 +29,7 @@
>   
>   #include "version_major.h"
>   
> -#define LIBAVCODEC_VERSION_MINOR  34
> +#define LIBAVCODEC_VERSION_MINOR  35
>   #define LIBAVCODEC_VERSION_MICRO 100
>   
>   #define LIBAVCODEC_VERSION_INT  AV_VERSION_INT(LIBAVCODEC_VERSION_MAJOR, \
>

_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

^ permalink raw reply	[flat|nested] 7+ messages in thread