libavcodec/libopenh264enc.c
8a3d9ca6
 /*
  * OpenH264 video encoder
  * Copyright (C) 2014 Martin Storsjo
  *
c09eeca5
  * This file is part of FFmpeg.
8a3d9ca6
  *
c09eeca5
  * FFmpeg is free software; you can redistribute it and/or
8a3d9ca6
  * 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.
  *
c09eeca5
  * FFmpeg is distributed in the hope that it will be useful,
8a3d9ca6
  * 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
c09eeca5
  * License along with FFmpeg; if not, write to the Free Software
8a3d9ca6
  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
  */
 
 #include <wels/codec_api.h>
 #include <wels/codec_ver.h>
 
 #include "libavutil/attributes.h"
 #include "libavutil/common.h"
 #include "libavutil/opt.h"
81c95eb8
 #include "libavutil/internal.h"
8a3d9ca6
 #include "libavutil/intreadwrite.h"
 #include "libavutil/mathematics.h"
 
 #include "avcodec.h"
 #include "internal.h"
c5d326f5
 #include "libopenh264.h"
8a3d9ca6
 
293676c4
 #if !OPENH264_VER_AT_LEAST(1, 6)
 #define SM_SIZELIMITED_SLICE SM_DYN_SLICE
 #endif
 
8a3d9ca6
 typedef struct SVCContext {
     const AVClass *av_class;
     ISVCEncoder *encoder;
     int slice_mode;
     int loopfilter;
     char *profile;
c3e5c47b
     int max_nal_size;
9e14a992
     int skip_frames;
8edaf625
     int skipped;
be00ec83
     int cabac;
8a3d9ca6
 } SVCContext;
 
 #define OFFSET(x) offsetof(SVCContext, x)
 #define VE AV_OPT_FLAG_VIDEO_PARAM | AV_OPT_FLAG_ENCODING_PARAM
 static const AVOption options[] = {
293676c4
 #if OPENH264_VER_AT_LEAST(1, 6)
     { "slice_mode", "set slice mode", OFFSET(slice_mode), AV_OPT_TYPE_INT, { .i64 = SM_FIXEDSLCNUM_SLICE }, SM_SINGLE_SLICE, SM_RESERVED, VE, "slice_mode" },
 #else
ae72b575
     { "slice_mode", "set slice mode", OFFSET(slice_mode), AV_OPT_TYPE_INT, { .i64 = SM_AUTO_SLICE }, SM_SINGLE_SLICE, SM_RESERVED, VE, "slice_mode" },
293676c4
 #endif
ae72b575
         { "fixed", "a fixed number of slices", 0, AV_OPT_TYPE_CONST, { .i64 = SM_FIXEDSLCNUM_SLICE }, 0, 0, VE, "slice_mode" },
293676c4
 #if OPENH264_VER_AT_LEAST(1, 6)
         { "dyn", "Size limited (compatibility name)", 0, AV_OPT_TYPE_CONST, { .i64 = SM_SIZELIMITED_SLICE }, 0, 0, VE, "slice_mode" },
         { "sizelimited", "Size limited", 0, AV_OPT_TYPE_CONST, { .i64 = SM_SIZELIMITED_SLICE }, 0, 0, VE, "slice_mode" },
 #else
ae72b575
         { "rowmb", "one slice per row of macroblocks", 0, AV_OPT_TYPE_CONST, { .i64 = SM_ROWMB_SLICE }, 0, 0, VE, "slice_mode" },
         { "auto", "automatic number of slices according to number of threads", 0, AV_OPT_TYPE_CONST, { .i64 = SM_AUTO_SLICE }, 0, 0, VE, "slice_mode" },
c4e23ca8
         { "dyn", "Dynamic slicing", 0, AV_OPT_TYPE_CONST, { .i64 = SM_DYN_SLICE }, 0, 0, VE, "slice_mode" },
293676c4
 #endif
ae72b575
     { "loopfilter", "enable loop filter", OFFSET(loopfilter), AV_OPT_TYPE_INT, { .i64 = 1 }, 0, 1, VE },
81d3986d
     { "profile", "set profile restrictions", OFFSET(profile), AV_OPT_TYPE_STRING, { .str = NULL }, 0, 0, VE },
6e891d51
     { "max_nal_size", "set maximum NAL size in bytes", OFFSET(max_nal_size), AV_OPT_TYPE_INT, { .i64 = 0 }, 0, INT_MAX, VE },
     { "allow_skip_frames", "allow skipping frames to hit the target bitrate", OFFSET(skip_frames), AV_OPT_TYPE_BOOL, { .i64 = 0 }, 0, 1, VE },
be00ec83
     { "cabac", "Enable cabac", OFFSET(cabac), AV_OPT_TYPE_INT, { .i64 = 0 }, 0, 1, VE },
8a3d9ca6
     { NULL }
 };
 
 static const AVClass class = {
e621b1ca
     .class_name = "libvo_amrwbenc",
97cfe1d8
     .item_name  = av_default_item_name,
     .option     = options,
     .version    = LIBAVUTIL_VERSION_INT,
8a3d9ca6
 };
 
 static av_cold int svc_encode_close(AVCodecContext *avctx)
 {
     SVCContext *s = avctx->priv_data;
 
     if (s->encoder)
         WelsDestroySVCEncoder(s->encoder);
8edaf625
     if (s->skipped > 0)
         av_log(avctx, AV_LOG_WARNING, "%d frames skipped\n", s->skipped);
8a3d9ca6
     return 0;
 }
 
 static av_cold int svc_encode_init(AVCodecContext *avctx)
 {
     SVCContext *s = avctx->priv_data;
     SEncParamExt param = { 0 };
c5d326f5
     int err;
1a4c5fe5
     int log_level;
e8c45b96
     WelsTraceCallback callback_function;
11c9bd63
     AVCPBProperties *props;
8a3d9ca6
 
c5d326f5
     if ((err = ff_libopenh264_check_version(avctx)) < 0)
         return err;
8a3d9ca6
 
     if (WelsCreateSVCEncoder(&s->encoder)) {
         av_log(avctx, AV_LOG_ERROR, "Unable to create encoder\n");
         return AVERROR_UNKNOWN;
     }
 
1a4c5fe5
     // Pass all libopenh264 messages to our callback, to allow ourselves to filter them.
     log_level = WELS_LOG_DETAIL;
     (*s->encoder)->SetOption(s->encoder, ENCODER_OPTION_TRACE_LEVEL, &log_level);
e8c45b96
 
     // Set the logging callback function to one that uses av_log() (see implementation above).
c5d326f5
     callback_function = (WelsTraceCallback) ff_libopenh264_trace_callback;
800d91d3
     (*s->encoder)->SetOption(s->encoder, ENCODER_OPTION_TRACE_CALLBACK, &callback_function);
e8c45b96
 
     // Set the AVCodecContext as the libopenh264 callback context so that it can be passed to av_log().
800d91d3
     (*s->encoder)->SetOption(s->encoder, ENCODER_OPTION_TRACE_CALLBACK_CONTEXT, &avctx);
e8c45b96
 
8a3d9ca6
     (*s->encoder)->GetDefaultParams(s->encoder, &param);
 
be00ec83
 #if FF_API_CODER_TYPE
 FF_DISABLE_DEPRECATION_WARNINGS
     if (!s->cabac)
         s->cabac = avctx->coder_type == FF_CODER_TYPE_AC;
 FF_ENABLE_DEPRECATION_WARNINGS
 #endif
 
7b404c94
     param.fMaxFrameRate              = 1/av_q2d(avctx->time_base);
8a3d9ca6
     param.iPicWidth                  = avctx->width;
     param.iPicHeight                 = avctx->height;
     param.iTargetBitrate             = avctx->bit_rate;
     param.iMaxBitrate                = FFMAX(avctx->rc_max_rate, avctx->bit_rate);
     param.iRCMode                    = RC_QUALITY_MODE;
     param.iTemporalLayerNum          = 1;
     param.iSpatialLayerNum           = 1;
     param.bEnableDenoise             = 0;
     param.bEnableBackgroundDetection = 1;
     param.bEnableAdaptiveQuant       = 1;
9e14a992
     param.bEnableFrameSkip           = s->skip_frames;
8a3d9ca6
     param.bEnableLongTermReference   = 0;
     param.iLtrMarkPeriod             = 30;
     param.uiIntraPeriod              = avctx->gop_size;
25c29d32
 #if OPENH264_VER_AT_LEAST(1, 4)
     param.eSpsPpsIdStrategy          = CONSTANT_ID;
 #else
8a3d9ca6
     param.bEnableSpsPpsIdAddition    = 0;
25c29d32
 #endif
8a3d9ca6
     param.bPrefixNalAddingCtrl       = 0;
     param.iLoopFilterDisableIdc      = !s->loopfilter;
     param.iEntropyCodingModeFlag     = 0;
     param.iMultipleThreadIdc         = avctx->thread_count;
     if (s->profile && !strcmp(s->profile, "main"))
         param.iEntropyCodingModeFlag = 1;
be00ec83
     else if (!s->profile && s->cabac)
8a3d9ca6
         param.iEntropyCodingModeFlag = 1;
 
     param.sSpatialLayers[0].iVideoWidth         = param.iPicWidth;
     param.sSpatialLayers[0].iVideoHeight        = param.iPicHeight;
     param.sSpatialLayers[0].fFrameRate          = param.fMaxFrameRate;
     param.sSpatialLayers[0].iSpatialBitrate     = param.iTargetBitrate;
     param.sSpatialLayers[0].iMaxSpatialBitrate  = param.iMaxBitrate;
 
31aa5335
     if ((avctx->slices > 1) && (s->max_nal_size)) {
         av_log(avctx, AV_LOG_ERROR,
                "Invalid combination -slices %d and -max_nal_size %d.\n",
                avctx->slices, s->max_nal_size);
7a763714
         return AVERROR(EINVAL);
c3e5c47b
     }
 
8a3d9ca6
     if (avctx->slices > 1)
         s->slice_mode = SM_FIXEDSLCNUM_SLICE;
c3e5c47b
 
     if (s->max_nal_size)
293676c4
         s->slice_mode = SM_SIZELIMITED_SLICE;
c3e5c47b
 
293676c4
 #if OPENH264_VER_AT_LEAST(1, 6)
     param.sSpatialLayers[0].sSliceArgument.uiSliceMode = s->slice_mode;
     param.sSpatialLayers[0].sSliceArgument.uiSliceNum  = avctx->slices;
 #else
8a3d9ca6
     param.sSpatialLayers[0].sSliceCfg.uiSliceMode               = s->slice_mode;
     param.sSpatialLayers[0].sSliceCfg.sSliceArgument.uiSliceNum = avctx->slices;
293676c4
 #endif
8a3d9ca6
 
293676c4
     if (s->slice_mode == SM_SIZELIMITED_SLICE) {
c3e5c47b
         if (s->max_nal_size){
             param.uiMaxNalSize = s->max_nal_size;
293676c4
 #if OPENH264_VER_AT_LEAST(1, 6)
             param.sSpatialLayers[0].sSliceArgument.uiSliceSizeConstraint = s->max_nal_size;
 #else
c3e5c47b
             param.sSpatialLayers[0].sSliceCfg.sSliceArgument.uiSliceSizeConstraint = s->max_nal_size;
293676c4
 #endif
c3e5c47b
         } else {
936f0d98
             av_log(avctx, AV_LOG_ERROR, "Invalid -max_nal_size, "
                    "specify a valid max_nal_size to use -slice_mode dyn\n");
7a763714
             return AVERROR(EINVAL);
c3e5c47b
         }
     }
 
8a3d9ca6
     if ((*s->encoder)->InitializeExt(s->encoder, &param) != cmResultSuccess) {
         av_log(avctx, AV_LOG_ERROR, "Initialize failed\n");
7a763714
         return AVERROR_UNKNOWN;
8a3d9ca6
     }
 
7c6eb0a1
     if (avctx->flags & AV_CODEC_FLAG_GLOBAL_HEADER) {
8a3d9ca6
         SFrameBSInfo fbi = { 0 };
         int i, size = 0;
         (*s->encoder)->EncodeParameterSets(s->encoder, &fbi);
         for (i = 0; i < fbi.sLayerInfo[0].iNalCount; i++)
             size += fbi.sLayerInfo[0].pNalLengthInByte[i];
059a9348
         avctx->extradata = av_mallocz(size + AV_INPUT_BUFFER_PADDING_SIZE);
7a763714
         if (!avctx->extradata)
             return AVERROR(ENOMEM);
8a3d9ca6
         avctx->extradata_size = size;
         memcpy(avctx->extradata, fbi.sLayerInfo[0].pBsBuf, size);
     }
 
11c9bd63
     props = ff_add_cpb_side_data(avctx);
7a763714
     if (!props)
         return AVERROR(ENOMEM);
11c9bd63
     props->max_bitrate = param.iMaxBitrate;
     props->avg_bitrate = param.iTargetBitrate;
 
8a3d9ca6
     return 0;
 }
 
 static int svc_encode_frame(AVCodecContext *avctx, AVPacket *avpkt,
                             const AVFrame *frame, int *got_packet)
 {
     SVCContext *s = avctx->priv_data;
     SFrameBSInfo fbi = { 0 };
     int i, ret;
     int encoded;
     SSourcePicture sp = { 0 };
     int size = 0, layer, first_layer = 0;
     int layer_size[MAX_LAYER_NUM_OF_FRAME] = { 0 };
 
     sp.iColorFormat = videoFormatI420;
     for (i = 0; i < 3; i++) {
         sp.iStride[i] = frame->linesize[i];
         sp.pData[i]   = frame->data[i];
     }
     sp.iPicWidth  = avctx->width;
     sp.iPicHeight = avctx->height;
 
     encoded = (*s->encoder)->EncodeFrame(s->encoder, &sp, &fbi);
     if (encoded != cmResultSuccess) {
         av_log(avctx, AV_LOG_ERROR, "EncodeFrame failed\n");
         return AVERROR_UNKNOWN;
     }
     if (fbi.eFrameType == videoFrameTypeSkip) {
8edaf625
         s->skipped++;
8a3d9ca6
         av_log(avctx, AV_LOG_DEBUG, "frame skipped\n");
         return 0;
     }
     first_layer = 0;
3852e2c9
     // Normal frames are returned with one single layer, while IDR
8a3d9ca6
     // frames have two layers, where the first layer contains the SPS/PPS.
     // If using global headers, don't include the SPS/PPS in the returned
     // packet - thus, only return one layer.
7c6eb0a1
     if (avctx->flags & AV_CODEC_FLAG_GLOBAL_HEADER)
8a3d9ca6
         first_layer = fbi.iLayerNum - 1;
 
     for (layer = first_layer; layer < fbi.iLayerNum; layer++) {
         for (i = 0; i < fbi.sLayerInfo[layer].iNalCount; i++)
             layer_size[layer] += fbi.sLayerInfo[layer].pNalLengthInByte[i];
         size += layer_size[layer];
     }
6996fd20
     av_log(avctx, AV_LOG_DEBUG, "%d slices\n", fbi.sLayerInfo[fbi.iLayerNum - 1].iNalCount);
8a3d9ca6
 
005d0069
     if ((ret = ff_alloc_packet2(avctx, avpkt, size, size))) {
8a3d9ca6
         av_log(avctx, AV_LOG_ERROR, "Error getting output packet\n");
         return ret;
     }
     size = 0;
     for (layer = first_layer; layer < fbi.iLayerNum; layer++) {
         memcpy(avpkt->data + size, fbi.sLayerInfo[layer].pBsBuf, layer_size[layer]);
         size += layer_size[layer];
     }
     avpkt->pts = frame->pts;
     if (fbi.eFrameType == videoFrameTypeIDR)
         avpkt->flags |= AV_PKT_FLAG_KEY;
     *got_packet = 1;
     return 0;
 }
 
 AVCodec ff_libopenh264_encoder = {
     .name           = "libopenh264",
3852e2c9
     .long_name      = NULL_IF_CONFIG_SMALL("OpenH264 H.264 / AVC / MPEG-4 AVC / MPEG-4 part 10"),
8a3d9ca6
     .type           = AVMEDIA_TYPE_VIDEO,
     .id             = AV_CODEC_ID_H264,
     .priv_data_size = sizeof(SVCContext),
     .init           = svc_encode_init,
     .encode2        = svc_encode_frame,
     .close          = svc_encode_close,
def97856
     .capabilities   = AV_CODEC_CAP_AUTO_THREADS,
7a763714
     .caps_internal  = FF_CODEC_CAP_INIT_THREADSAFE | FF_CODEC_CAP_INIT_CLEANUP,
58a840e2
     .pix_fmts       = (const enum AVPixelFormat[]){ AV_PIX_FMT_YUV420P,
                                                     AV_PIX_FMT_NONE },
8a3d9ca6
     .priv_class     = &class,
 };