libx264.c 30.4 KB
Newer Older
1 2
/*
 * H.264 encoding using the x264 library
3
 * Copyright (C) 2005  Mans Rullgard <mans@mansr.com>
4
 *
5
 * This file is part of Libav.
6
 *
7
 * Libav is free software; you can redistribute it and/or
8 9
 * modify it under the terms of the GNU Lesser General Public
 * License as published by the Free Software Foundation; either
10
 * version 2.1 of the License, or (at your option) any later version.
11
 *
12
 * Libav is distributed in the hope that it will be useful,
13 14 15 16 17
 * 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
18
 * License along with Libav; if not, write to the Free Software
19
 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
20 21
 */

22
#include "libavutil/internal.h"
23
#include "libavutil/opt.h"
24
#include "libavutil/mem.h"
25
#include "libavutil/pixdesc.h"
26
#include "libavutil/stereo3d.h"
27
#include "avcodec.h"
28
#include "internal.h"
29 30 31 32 33

#if defined(_MSC_VER)
#define X264_API_IMPORTS 1
#endif

34
#include <x264.h>
35
#include <float.h>
Michael Niedermayer's avatar
Michael Niedermayer committed
36
#include <math.h>
Robert Swain's avatar
Robert Swain committed
37 38 39
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
40 41

typedef struct X264Context {
42
    AVClass        *class;
43 44 45 46 47
    x264_param_t    params;
    x264_t         *enc;
    x264_picture_t  pic;
    uint8_t        *sei;
    int             sei_size;
48 49 50 51
    char *preset;
    char *tune;
    char *profile;
    int fastfirstpass;
52
    float crf;
53
    float crf_max;
54
    int cqp;
55
    int aq_mode;
56
    float aq_strength;
57
    char *psy_rd;
58
    int psy;
59
    int rc_lookahead;
60
    int weightp;
61
    int weightb;
62
    int ssim;
63
    int intra_refresh;
64
    int bluray_compat;
65
    int b_bias;
66
    int b_pyramid;
67
    int mixed_refs;
68
    int dct8x8;
69
    int fast_pskip;
70
    int aud;
71
    int mbtree;
72
    char *deblock;
73
    float cplxblur;
74
    char *partitions;
75
    int direct_pred;
76
    int slice_max_size;
77
    char *stats;
Anton Khirnov's avatar
Anton Khirnov committed
78
    int nal_hrd;
79
    int motion_est;
80
    char *x264_params;
81 82
} X264Context;

83
static void X264_log(void *p, int level, const char *fmt, va_list args)
84 85
{
    static const int level_map[] = {
86
        [X264_LOG_ERROR]   = AV_LOG_ERROR,
87
        [X264_LOG_WARNING] = AV_LOG_WARNING,
88 89
        [X264_LOG_INFO]    = AV_LOG_INFO,
        [X264_LOG_DEBUG]   = AV_LOG_DEBUG
90 91
    };

92
    if (level < 0 || level > X264_LOG_DEBUG)
93
        return;
94 95 96 97 98

    av_vlog(p, level_map[level], fmt, args);
}


99 100
static int encode_nals(AVCodecContext *ctx, AVPacket *pkt,
                       x264_nal_t *nals, int nnal)
101
{
102
    X264Context *x4 = ctx->priv_data;
103 104 105 106 107 108 109 110 111 112 113 114 115
    uint8_t *p;
    int i, size = x4->sei_size, ret;

    if (!nnal)
        return 0;

    for (i = 0; i < nnal; i++)
        size += nals[i].i_payload;

    if ((ret = ff_alloc_packet(pkt, size)) < 0)
        return ret;

    p = pkt->data;
116 117

    /* Write the SEI as part of the first frame. */
118
    if (x4->sei_size > 0 && nnal > 0) {
119 120 121 122
        memcpy(p, x4->sei, x4->sei_size);
        p += x4->sei_size;
        x4->sei_size = 0;
    }
123

124
    for (i = 0; i < nnal; i++){
125 126
        memcpy(p, nals[i].p_payload, nals[i].i_payload);
        p += nals[i].i_payload;
127 128
    }

129
    return 1;
130 131
}

132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216
static void reconfig_encoder(AVCodecContext *ctx, const AVFrame *frame)
{
    X264Context *x4 = ctx->priv_data;
    AVFrameSideData *side_data;


    if (x4->params.b_tff != frame->top_field_first) {
        x4->params.b_tff = frame->top_field_first;
        x264_encoder_reconfig(x4->enc, &x4->params);
    }
    if (x4->params.vui.i_sar_height != ctx->sample_aspect_ratio.den ||
        x4->params.vui.i_sar_width  != ctx->sample_aspect_ratio.num) {
        x4->params.vui.i_sar_height = ctx->sample_aspect_ratio.den;
        x4->params.vui.i_sar_width  = ctx->sample_aspect_ratio.num;
        x264_encoder_reconfig(x4->enc, &x4->params);
    }

    if (x4->params.rc.i_vbv_buffer_size != ctx->rc_buffer_size / 1000 ||
        x4->params.rc.i_vbv_max_bitrate != ctx->rc_max_rate    / 1000) {
        x4->params.rc.i_vbv_buffer_size = ctx->rc_buffer_size / 1000;
        x4->params.rc.i_vbv_max_bitrate = ctx->rc_max_rate    / 1000;
        x264_encoder_reconfig(x4->enc, &x4->params);
    }

    if (x4->params.rc.i_rc_method == X264_RC_ABR &&
        x4->params.rc.i_bitrate != ctx->bit_rate / 1000) {
        x4->params.rc.i_bitrate = ctx->bit_rate / 1000;
        x264_encoder_reconfig(x4->enc, &x4->params);
    }

    if (x4->crf >= 0 &&
        x4->params.rc.i_rc_method == X264_RC_CRF &&
        x4->params.rc.f_rf_constant != x4->crf) {
        x4->params.rc.f_rf_constant = x4->crf;
        x264_encoder_reconfig(x4->enc, &x4->params);
    }

    if (x4->params.rc.i_rc_method == X264_RC_CQP &&
        x4->params.rc.i_qp_constant != x4->cqp) {
        x4->params.rc.i_qp_constant = x4->cqp;
        x264_encoder_reconfig(x4->enc, &x4->params);
    }

    if (x4->crf_max >= 0 &&
        x4->params.rc.f_rf_constant_max != x4->crf_max) {
        x4->params.rc.f_rf_constant_max = x4->crf_max;
        x264_encoder_reconfig(x4->enc, &x4->params);
    }

    side_data = av_frame_get_side_data(frame, AV_FRAME_DATA_STEREO3D);
    if (side_data) {
        AVStereo3D *stereo = (AVStereo3D *)side_data->data;
        int fpa_type;

        switch (stereo->type) {
        case AV_STEREO3D_CHECKERBOARD:
            fpa_type = 0;
            break;
        case AV_STEREO3D_COLUMNS:
            fpa_type = 1;
            break;
        case AV_STEREO3D_LINES:
            fpa_type = 2;
            break;
        case AV_STEREO3D_SIDEBYSIDE:
            fpa_type = 3;
            break;
        case AV_STEREO3D_TOPBOTTOM:
            fpa_type = 4;
            break;
        case AV_STEREO3D_FRAMESEQUENCE:
            fpa_type = 5;
            break;
        default:
            fpa_type = -1;
            break;
        }

        if (fpa_type != x4->params.i_frame_packing) {
            x4->params.i_frame_packing = fpa_type;
            x264_encoder_reconfig(x4->enc, &x4->params);
        }
    }
}

217 218
static int X264_frame(AVCodecContext *ctx, AVPacket *pkt, const AVFrame *frame,
                      int *got_packet)
219 220 221
{
    X264Context *x4 = ctx->priv_data;
    x264_nal_t *nal;
222
    int nnal, i, ret;
223 224
    x264_picture_t pic_out;

225
    x264_picture_init( &x4->pic );
226
    x4->pic.img.i_csp   = x4->params.i_csp;
227 228
    if (x264_bit_depth > 8)
        x4->pic.img.i_csp |= X264_CSP_HIGH_DEPTH;
229 230
    x4->pic.img.i_plane = 3;

231 232 233
    if (frame) {
        for (i = 0; i < 3; i++) {
            x4->pic.img.plane[i]    = frame->data[i];
Måns Rullgård's avatar
Måns Rullgård committed
234 235
            x4->pic.img.i_stride[i] = frame->linesize[i];
        }
236

237
        x4->pic.i_pts  = frame->pts;
238
        x4->pic.i_type =
239 240 241
            frame->pict_type == AV_PICTURE_TYPE_I ? X264_TYPE_KEYFRAME :
            frame->pict_type == AV_PICTURE_TYPE_P ? X264_TYPE_P :
            frame->pict_type == AV_PICTURE_TYPE_B ? X264_TYPE_B :
242
                                            X264_TYPE_AUTO;
243
        reconfig_encoder(ctx, frame);
244
    }
245
    do {
246
        if (x264_encoder_encode(x4->enc, &nal, &nnal, frame? &x4->pic: NULL, &pic_out) < 0)
247
            return AVERROR_UNKNOWN;
248

249 250
        ret = encode_nals(ctx, pkt, nal, nnal);
        if (ret < 0)
251
            return ret;
252
    } while (!ret && !frame && x264_encoder_delayed_frames(x4->enc));
253

254 255
    pkt->pts = pic_out.i_pts;
    pkt->dts = pic_out.i_dts;
256

257 258
#if FF_API_CODED_FRAME
FF_DISABLE_DEPRECATION_WARNINGS
259
    switch (pic_out.i_type) {
260 261
    case X264_TYPE_IDR:
    case X264_TYPE_I:
262
        ctx->coded_frame->pict_type = AV_PICTURE_TYPE_I;
263 264
        break;
    case X264_TYPE_P:
265
        ctx->coded_frame->pict_type = AV_PICTURE_TYPE_P;
266 267 268
        break;
    case X264_TYPE_B:
    case X264_TYPE_BREF:
269
        ctx->coded_frame->pict_type = AV_PICTURE_TYPE_B;
270 271
        break;
    }
272 273
FF_ENABLE_DEPRECATION_WARNINGS
#endif
274

275
    pkt->flags |= AV_PKT_FLAG_KEY*pic_out.b_keyframe;
276 277 278 279 280 281 282
    if (ret) {
        uint8_t *sd = av_packet_new_side_data(pkt, AV_PKT_DATA_QUALITY_FACTOR,
                                              sizeof(int));
        if (!sd)
            return AVERROR(ENOMEM);
        *(int *)sd = (pic_out.i_qpplus1 - 1) * FF_QP2LAMBDA;

283 284
#if FF_API_CODED_FRAME
FF_DISABLE_DEPRECATION_WARNINGS
285
        ctx->coded_frame->quality = (pic_out.i_qpplus1 - 1) * FF_QP2LAMBDA;
286 287
FF_ENABLE_DEPRECATION_WARNINGS
#endif
288
    }
289

290 291
    *got_packet = ret;
    return 0;
292 293
}

294
static av_cold int X264_close(AVCodecContext *avctx)
295 296 297
{
    X264Context *x4 = avctx->priv_data;

Ralf Terdic's avatar
Ralf Terdic committed
298
    av_freep(&avctx->extradata);
299
    av_freep(&x4->sei);
Ralf Terdic's avatar
Ralf Terdic committed
300

301
    if (x4->enc) {
302
        x264_encoder_close(x4->enc);
303 304
        x4->enc = NULL;
    }
305 306 307 308

    return 0;
}

309
static int convert_pix_fmt(enum AVPixelFormat pix_fmt)
310 311
{
    switch (pix_fmt) {
312 313 314 315 316
    case AV_PIX_FMT_YUV420P:
    case AV_PIX_FMT_YUVJ420P:
    case AV_PIX_FMT_YUV420P9:
    case AV_PIX_FMT_YUV420P10: return X264_CSP_I420;
    case AV_PIX_FMT_YUV422P:
317
    case AV_PIX_FMT_YUVJ422P:
318 319
    case AV_PIX_FMT_YUV422P10: return X264_CSP_I422;
    case AV_PIX_FMT_YUV444P:
320
    case AV_PIX_FMT_YUVJ444P:
321 322
    case AV_PIX_FMT_YUV444P9:
    case AV_PIX_FMT_YUV444P10: return X264_CSP_I444;
323 324 325
    case AV_PIX_FMT_NV12:      return X264_CSP_NV12;
    case AV_PIX_FMT_NV16:
    case AV_PIX_FMT_NV20:      return X264_CSP_NV16;
326 327 328 329
    };
    return 0;
}

330 331 332 333 334 335
#define PARSE_X264_OPT(name, var)\
    if (x4->var && x264_param_parse(&x4->params, name, x4->var) < 0) {\
        av_log(avctx, AV_LOG_ERROR, "Error parsing option '%s' with value '%s'.\n", name, x4->var);\
        return AVERROR(EINVAL);\
    }

336
static av_cold int X264_init(AVCodecContext *avctx)
337 338 339
{
    X264Context *x4 = avctx->priv_data;

340 341 342 343 344 345
#if CONFIG_LIBX262_ENCODER
    if (avctx->codec_id == AV_CODEC_ID_MPEG2VIDEO) {
        x4->params.b_mpeg2 = 1;
        x264_param_default_mpeg2(&x4->params);
    } else
#else
346
    x264_param_default(&x4->params);
347
#endif
348

349
    x4->params.b_deblocking_filter         = avctx->flags & AV_CODEC_FLAG_LOOP_FILTER;
350

351 352 353 354 355 356
    if (x4->preset || x4->tune)
        if (x264_param_default_preset(&x4->params, x4->preset, x4->tune) < 0) {
            av_log(avctx, AV_LOG_ERROR, "Error setting preset/tune %s/%s.\n", x4->preset, x4->tune);
            return AVERROR(EINVAL);
        }

357 358 359
    if (avctx->level > 0)
        x4->params.i_level_idc = avctx->level;

360 361 362
    x4->params.pf_log               = X264_log;
    x4->params.p_log_private        = avctx;
    x4->params.i_log_level          = X264_LOG_DEBUG;
363
    x4->params.i_csp                = convert_pix_fmt(avctx->pix_fmt);
364

365 366 367 368
    if (avctx->bit_rate) {
        x4->params.rc.i_bitrate   = avctx->bit_rate / 1000;
        x4->params.rc.i_rc_method = X264_RC_ABR;
    }
369 370
    x4->params.rc.i_vbv_buffer_size = avctx->rc_buffer_size / 1000;
    x4->params.rc.i_vbv_max_bitrate = avctx->rc_max_rate    / 1000;
371 372
    x4->params.rc.b_stat_write      = avctx->flags & AV_CODEC_FLAG_PASS1;
    if (avctx->flags & AV_CODEC_FLAG_PASS2) {
373 374
        x4->params.rc.b_stat_read = 1;
    } else {
375 376 377
        if (x4->crf >= 0) {
            x4->params.rc.i_rc_method   = X264_RC_CRF;
            x4->params.rc.f_rf_constant = x4->crf;
378 379 380
        } else if (x4->cqp >= 0) {
            x4->params.rc.i_rc_method   = X264_RC_CQP;
            x4->params.rc.i_qp_constant = x4->cqp;
381
        }
382 383 384

        if (x4->crf_max >= 0)
            x4->params.rc.f_rf_constant_max = x4->crf_max;
385 386
    }

387
    if (avctx->rc_buffer_size && avctx->rc_initial_buffer_occupancy > 0 &&
388
        (avctx->rc_initial_buffer_occupancy <= avctx->rc_buffer_size)) {
Robert Swain's avatar
Robert Swain committed
389
        x4->params.rc.f_vbv_buffer_init =
390
            (float)avctx->rc_initial_buffer_occupancy / avctx->rc_buffer_size;
391
    }
Robert Swain's avatar
Robert Swain committed
392

393 394
    if (avctx->i_quant_factor > 0)
        x4->params.rc.f_ip_factor         = 1 / fabs(avctx->i_quant_factor);
395
    x4->params.rc.f_pb_factor             = avctx->b_quant_factor;
Robert Swain's avatar
Robert Swain committed
396 397
    x4->params.analyse.i_chroma_qp_offset = avctx->chromaoffset;

398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415
    if (avctx->gop_size >= 0)
        x4->params.i_keyint_max         = avctx->gop_size;
    if (avctx->max_b_frames >= 0)
        x4->params.i_bframe             = avctx->max_b_frames;
    if (avctx->scenechange_threshold >= 0)
        x4->params.i_scenecut_threshold = avctx->scenechange_threshold;
    if (avctx->qmin >= 0)
        x4->params.rc.i_qp_min          = avctx->qmin;
    if (avctx->qmax >= 0)
        x4->params.rc.i_qp_max          = avctx->qmax;
    if (avctx->max_qdiff >= 0)
        x4->params.rc.i_qp_step         = avctx->max_qdiff;
    if (avctx->qblur >= 0)
        x4->params.rc.f_qblur           = avctx->qblur;     /* temporally blur quants */
    if (avctx->qcompress >= 0)
        x4->params.rc.f_qcompress       = avctx->qcompress; /* 0.0 => cbr, 1.0 => constant qp */
    if (avctx->refs >= 0)
        x4->params.i_frame_reference    = avctx->refs;
416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431
    if (avctx->trellis >= 0)
        x4->params.analyse.i_trellis    = avctx->trellis;
    if (avctx->me_range >= 0)
        x4->params.analyse.i_me_range   = avctx->me_range;
    if (avctx->noise_reduction >= 0)
        x4->params.analyse.i_noise_reduction = avctx->noise_reduction;
    if (avctx->me_subpel_quality >= 0)
        x4->params.analyse.i_subpel_refine   = avctx->me_subpel_quality;
    if (avctx->b_frame_strategy >= 0)
        x4->params.i_bframe_adaptive = avctx->b_frame_strategy;
    if (avctx->keyint_min >= 0)
        x4->params.i_keyint_min = avctx->keyint_min;
    if (avctx->coder_type >= 0)
        x4->params.b_cabac = avctx->coder_type == FF_CODER_TYPE_AC;
    if (avctx->me_cmp >= 0)
        x4->params.analyse.b_chroma_me = avctx->me_cmp & FF_CMP_CHROMA;
432

433 434
    if (x4->aq_mode >= 0)
        x4->params.rc.i_aq_mode = x4->aq_mode;
435 436
    if (x4->aq_strength >= 0)
        x4->params.rc.f_aq_strength = x4->aq_strength;
437 438
    PARSE_X264_OPT("psy-rd", psy_rd);
    PARSE_X264_OPT("deblock", deblock);
439
    PARSE_X264_OPT("partitions", partitions);
440
    PARSE_X264_OPT("stats", stats);
441 442
    if (x4->psy >= 0)
        x4->params.analyse.b_psy  = x4->psy;
443 444
    if (x4->rc_lookahead >= 0)
        x4->params.rc.i_lookahead = x4->rc_lookahead;
445 446
    if (x4->weightp >= 0)
        x4->params.analyse.i_weighted_pred = x4->weightp;
447 448
    if (x4->weightb >= 0)
        x4->params.analyse.b_weighted_bipred = x4->weightb;
449 450
    if (x4->cplxblur >= 0)
        x4->params.rc.f_complexity_blur = x4->cplxblur;
451

452 453
    if (x4->ssim >= 0)
        x4->params.analyse.b_ssim = x4->ssim;
454 455
    if (x4->intra_refresh >= 0)
        x4->params.b_intra_refresh = x4->intra_refresh;
456 457 458 459
    if (x4->bluray_compat >= 0) {
        x4->params.b_bluray_compat = x4->bluray_compat;
        x4->params.b_vfr_input = 0;
    }
460 461
    if (x4->b_bias != INT_MIN)
        x4->params.i_bframe_bias              = x4->b_bias;
462 463
    if (x4->b_pyramid >= 0)
        x4->params.i_bframe_pyramid = x4->b_pyramid;
464 465
    if (x4->mixed_refs >= 0)
        x4->params.analyse.b_mixed_references = x4->mixed_refs;
466 467
    if (x4->dct8x8 >= 0)
        x4->params.analyse.b_transform_8x8    = x4->dct8x8;
468 469
    if (x4->fast_pskip >= 0)
        x4->params.analyse.b_fast_pskip       = x4->fast_pskip;
470 471
    if (x4->aud >= 0)
        x4->params.b_aud                      = x4->aud;
472 473
    if (x4->mbtree >= 0)
        x4->params.rc.b_mb_tree               = x4->mbtree;
474 475
    if (x4->direct_pred >= 0)
        x4->params.analyse.i_direct_mv_pred   = x4->direct_pred;
476

477 478 479
    if (x4->slice_max_size >= 0)
        x4->params.i_slice_max_size =  x4->slice_max_size;

480 481 482
    if (x4->fastfirstpass)
        x264_param_apply_fastfirstpass(&x4->params);

483 484 485
    if (x4->nal_hrd >= 0)
        x4->params.i_nal_hrd = x4->nal_hrd;

486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504
    if (x4->motion_est >= 0) {
        x4->params.analyse.i_me_method = x4->motion_est;
#if FF_API_MOTION_EST
FF_DISABLE_DEPRECATION_WARNINGS
    } else {
        if (avctx->me_method == ME_EPZS)
            x4->params.analyse.i_me_method = X264_ME_DIA;
        else if (avctx->me_method == ME_HEX)
            x4->params.analyse.i_me_method = X264_ME_HEX;
        else if (avctx->me_method == ME_UMH)
            x4->params.analyse.i_me_method = X264_ME_UMH;
        else if (avctx->me_method == ME_FULL)
            x4->params.analyse.i_me_method = X264_ME_ESA;
        else if (avctx->me_method == ME_TESA)
            x4->params.analyse.i_me_method = X264_ME_TESA;
FF_ENABLE_DEPRECATION_WARNINGS
#endif
    }

505 506 507 508 509 510 511 512 513 514 515 516 517
    if (x4->profile)
        if (x264_param_apply_profile(&x4->params, x4->profile) < 0) {
            av_log(avctx, AV_LOG_ERROR, "Error setting profile %s.\n", x4->profile);
            return AVERROR(EINVAL);
        }

    x4->params.i_width          = avctx->width;
    x4->params.i_height         = avctx->height;
    x4->params.vui.i_sar_width  = avctx->sample_aspect_ratio.num;
    x4->params.vui.i_sar_height = avctx->sample_aspect_ratio.den;
    x4->params.i_fps_num = x4->params.i_timebase_den = avctx->time_base.den;
    x4->params.i_fps_den = x4->params.i_timebase_num = avctx->time_base.num;

518
    x4->params.analyse.b_psnr = avctx->flags & AV_CODEC_FLAG_PSNR;
Robert Swain's avatar
Robert Swain committed
519

520
    x4->params.i_threads      = avctx->thread_count;
521 522
    if (avctx->thread_type)
        x4->params.b_sliced_threads = avctx->thread_type == FF_THREAD_SLICE;
523

524
    x4->params.b_interlaced   = avctx->flags & AV_CODEC_FLAG_INTERLACED_DCT;
525

526
    x4->params.b_open_gop     = !(avctx->flags & AV_CODEC_FLAG_CLOSED_GOP);
527

528 529
    x4->params.i_slice_count  = avctx->slices;

530
    x4->params.vui.b_fullrange = avctx->pix_fmt == AV_PIX_FMT_YUVJ420P ||
531 532
                                 avctx->pix_fmt == AV_PIX_FMT_YUVJ422P ||
                                 avctx->pix_fmt == AV_PIX_FMT_YUVJ444P ||
533
                                 avctx->color_range == AVCOL_RANGE_JPEG;
534

535 536 537 538 539
    // x264 validates the values internally
    x4->params.vui.i_colorprim = avctx->color_primaries;
    x4->params.vui.i_transfer  = avctx->color_trc;
    x4->params.vui.i_colmatrix = avctx->colorspace;

540
    if (avctx->flags & AV_CODEC_FLAG_GLOBAL_HEADER)
541 542
        x4->params.b_repeat_headers = 0;

543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558
    if (x4->x264_params) {
        AVDictionary *dict    = NULL;
        AVDictionaryEntry *en = NULL;

        if (!av_dict_parse_string(&dict, x4->x264_params, "=", ":", 0)) {
            while ((en = av_dict_get(dict, "", en, AV_DICT_IGNORE_SUFFIX))) {
                if (x264_param_parse(&x4->params, en->key, en->value) < 0)
                    av_log(avctx, AV_LOG_WARNING,
                           "Error parsing option '%s = %s'.\n",
                            en->key, en->value);
            }

            av_dict_free(&dict);
        }
    }

559 560 561
    // update AVCodecContext with x264 parameters
    avctx->has_b_frames = x4->params.i_bframe ?
        x4->params.i_bframe_pyramid ? 2 : 1 : 0;
562 563 564
    if (avctx->max_b_frames < 0)
        avctx->max_b_frames = 0;

565 566
    avctx->bit_rate = x4->params.rc.i_bitrate*1000;

567
    x4->enc = x264_encoder_open(&x4->params);
568
    if (!x4->enc)
569
        return AVERROR_UNKNOWN;
570

571
    if (avctx->flags & AV_CODEC_FLAG_GLOBAL_HEADER) {
572
        x264_nal_t *nal;
573
        uint8_t *p;
574
        int nnal, s, i;
575

576
        s = x264_encoder_headers(x4->enc, &nal, &nnal);
577
        avctx->extradata = p = av_malloc(s);
578 579
        if (!p)
            return AVERROR(ENOMEM);
580

581 582 583
        for (i = 0; i < nnal; i++) {
            /* Don't put the SEI in extradata. */
            if (nal[i].i_type == NAL_SEI) {
584
                av_log(avctx, AV_LOG_INFO, "%s\n", nal[i].p_payload+25);
585 586
                x4->sei_size = nal[i].i_payload;
                x4->sei      = av_malloc(x4->sei_size);
587 588
                if (!x4->sei)
                    return AVERROR(ENOMEM);
589 590 591 592 593 594 595
                memcpy(x4->sei, nal[i].p_payload, nal[i].i_payload);
                continue;
            }
            memcpy(p, nal[i].p_payload, nal[i].i_payload);
            p += nal[i].i_payload;
        }
        avctx->extradata_size = p - avctx->extradata;
596 597
    }

598 599 600
    return 0;
}

601 602 603 604
static const enum AVPixelFormat pix_fmts_8bit[] = {
    AV_PIX_FMT_YUV420P,
    AV_PIX_FMT_YUVJ420P,
    AV_PIX_FMT_YUV422P,
605
    AV_PIX_FMT_YUVJ422P,
606
    AV_PIX_FMT_YUV444P,
607
    AV_PIX_FMT_YUVJ444P,
608 609
    AV_PIX_FMT_NV12,
    AV_PIX_FMT_NV16,
610
    AV_PIX_FMT_NONE
611
};
612 613 614 615
static const enum AVPixelFormat pix_fmts_9bit[] = {
    AV_PIX_FMT_YUV420P9,
    AV_PIX_FMT_YUV444P9,
    AV_PIX_FMT_NONE
616
};
617 618 619 620
static const enum AVPixelFormat pix_fmts_10bit[] = {
    AV_PIX_FMT_YUV420P10,
    AV_PIX_FMT_YUV422P10,
    AV_PIX_FMT_YUV444P10,
621
    AV_PIX_FMT_NV20,
622
    AV_PIX_FMT_NONE
623 624 625 626 627 628 629 630 631 632 633 634
};

static av_cold void X264_init_static(AVCodec *codec)
{
    if (x264_bit_depth == 8)
        codec->pix_fmts = pix_fmts_8bit;
    else if (x264_bit_depth == 9)
        codec->pix_fmts = pix_fmts_9bit;
    else if (x264_bit_depth == 10)
        codec->pix_fmts = pix_fmts_10bit;
}

635 636 637
#define OFFSET(x) offsetof(X264Context, x)
#define VE AV_OPT_FLAG_VIDEO_PARAM | AV_OPT_FLAG_ENCODING_PARAM
static const AVOption options[] = {
638 639 640
    { "preset",        "Set the encoding preset (cf. x264 --fullhelp)",   OFFSET(preset),        AV_OPT_TYPE_STRING, { .str = "medium" }, 0, 0, VE},
    { "tune",          "Tune the encoding params (cf. x264 --fullhelp)",  OFFSET(tune),          AV_OPT_TYPE_STRING, { 0 }, 0, 0, VE},
    { "profile",       "Set profile restrictions (cf. x264 --fullhelp) ", OFFSET(profile),       AV_OPT_TYPE_STRING, { 0 }, 0, 0, VE},
641
    { "fastfirstpass", "Use fast settings when encoding first pass",      OFFSET(fastfirstpass), AV_OPT_TYPE_INT,    { .i64 = 1 }, 0, 1, VE},
642 643
    { "crf",           "Select the quality for constant quality mode",    OFFSET(crf),           AV_OPT_TYPE_FLOAT,  {.dbl = -1 }, -1, FLT_MAX, VE },
    { "crf_max",       "In CRF mode, prevents VBV from lowering quality beyond this point.",OFFSET(crf_max), AV_OPT_TYPE_FLOAT, {.dbl = -1 }, -1, FLT_MAX, VE },
644 645
    { "qp",            "Constant quantization parameter rate control method",OFFSET(cqp),        AV_OPT_TYPE_INT,    { .i64 = -1 }, -1, INT_MAX, VE },
    { "aq-mode",       "AQ method",                                       OFFSET(aq_mode),       AV_OPT_TYPE_INT,    { .i64 = -1 }, -1, INT_MAX, VE, "aq_mode"},
646 647 648
    { "none",          NULL,                              0, AV_OPT_TYPE_CONST, {.i64 = X264_AQ_NONE},         INT_MIN, INT_MAX, VE, "aq_mode" },
    { "variance",      "Variance AQ (complexity mask)",   0, AV_OPT_TYPE_CONST, {.i64 = X264_AQ_VARIANCE},     INT_MIN, INT_MAX, VE, "aq_mode" },
    { "autovariance",  "Auto-variance AQ (experimental)", 0, AV_OPT_TYPE_CONST, {.i64 = X264_AQ_AUTOVARIANCE}, INT_MIN, INT_MAX, VE, "aq_mode" },
649
    { "aq-strength",   "AQ strength. Reduces blocking and blurring in flat and textured areas.", OFFSET(aq_strength), AV_OPT_TYPE_FLOAT, {.dbl = -1}, -1, FLT_MAX, VE},
650
    { "psy",           "Use psychovisual optimizations.",                 OFFSET(psy),           AV_OPT_TYPE_INT,    { .i64 = -1 }, -1, 1, VE },
651
    { "psy-rd",        "Strength of psychovisual optimization, in <psy-rd>:<psy-trellis> format.", OFFSET(psy_rd), AV_OPT_TYPE_STRING,  {0 }, 0, 0, VE},
652 653 654
    { "rc-lookahead",  "Number of frames to look ahead for frametype and ratecontrol", OFFSET(rc_lookahead), AV_OPT_TYPE_INT, { .i64 = -1 }, -1, INT_MAX, VE },
    { "weightb",       "Weighted prediction for B-frames.",               OFFSET(weightb),       AV_OPT_TYPE_INT,    { .i64 = -1 }, -1, 1, VE },
    { "weightp",       "Weighted prediction analysis method.",            OFFSET(weightp),       AV_OPT_TYPE_INT,    { .i64 = -1 }, -1, INT_MAX, VE, "weightp" },
655 656 657
    { "none",          NULL, 0, AV_OPT_TYPE_CONST, {.i64 = X264_WEIGHTP_NONE},   INT_MIN, INT_MAX, VE, "weightp" },
    { "simple",        NULL, 0, AV_OPT_TYPE_CONST, {.i64 = X264_WEIGHTP_SIMPLE}, INT_MIN, INT_MAX, VE, "weightp" },
    { "smart",         NULL, 0, AV_OPT_TYPE_CONST, {.i64 = X264_WEIGHTP_SMART},  INT_MIN, INT_MAX, VE, "weightp" },
658 659
    { "ssim",          "Calculate and print SSIM stats.",                 OFFSET(ssim),          AV_OPT_TYPE_INT,    { .i64 = -1 }, -1, 1, VE },
    { "intra-refresh", "Use Periodic Intra Refresh instead of IDR frames.",OFFSET(intra_refresh),AV_OPT_TYPE_INT,    { .i64 = -1 }, -1, 1, VE },
660
    { "bluray-compat", "Bluray compatibility workarounds.",               OFFSET(bluray_compat) ,AV_OPT_TYPE_INT,    { .i64 = -1 }, -1, 1, VE },
661 662
    { "b-bias",        "Influences how often B-frames are used",          OFFSET(b_bias),        AV_OPT_TYPE_INT,    { .i64 = INT_MIN}, INT_MIN, INT_MAX, VE },
    { "b-pyramid",     "Keep some B-frames as references.",               OFFSET(b_pyramid),     AV_OPT_TYPE_INT,    { .i64 = -1 }, -1, INT_MAX, VE, "b_pyramid" },
663 664 665
    { "none",          NULL,                                  0, AV_OPT_TYPE_CONST, {.i64 = X264_B_PYRAMID_NONE},   INT_MIN, INT_MAX, VE, "b_pyramid" },
    { "strict",        "Strictly hierarchical pyramid",       0, AV_OPT_TYPE_CONST, {.i64 = X264_B_PYRAMID_STRICT}, INT_MIN, INT_MAX, VE, "b_pyramid" },
    { "normal",        "Non-strict (not Blu-ray compatible)", 0, AV_OPT_TYPE_CONST, {.i64 = X264_B_PYRAMID_NORMAL}, INT_MIN, INT_MAX, VE, "b_pyramid" },
666 667 668 669 670
    { "mixed-refs",    "One reference per partition, as opposed to one reference per macroblock", OFFSET(mixed_refs), AV_OPT_TYPE_INT, { .i64 = -1}, -1, 1, VE },
    { "8x8dct",        "High profile 8x8 transform.",                     OFFSET(dct8x8),        AV_OPT_TYPE_INT,    { .i64 = -1 }, -1, 1, VE},
    { "fast-pskip",    NULL,                                              OFFSET(fast_pskip),    AV_OPT_TYPE_INT,    { .i64 = -1 }, -1, 1, VE},
    { "aud",           "Use access unit delimiters.",                     OFFSET(aud),           AV_OPT_TYPE_INT,    { .i64 = -1 }, -1, 1, VE},
    { "mbtree",        "Use macroblock tree ratecontrol.",                OFFSET(mbtree),        AV_OPT_TYPE_INT,    { .i64 = -1 }, -1, 1, VE},
671
    { "deblock",       "Loop filter parameters, in <alpha:beta> form.",   OFFSET(deblock),       AV_OPT_TYPE_STRING, { 0 },  0, 0, VE},
672
    { "cplxblur",      "Reduce fluctuations in QP (before curve compression)", OFFSET(cplxblur), AV_OPT_TYPE_FLOAT,  {.dbl = -1 }, -1, FLT_MAX, VE},
673
    { "partitions",    "A comma-separated list of partitions to consider. "
674
                       "Possible values: p8x8, p4x4, b8x8, i8x8, i4x4, none, all", OFFSET(partitions), AV_OPT_TYPE_STRING, { 0 }, 0, 0, VE},
675
    { "direct-pred",   "Direct MV prediction mode",                       OFFSET(direct_pred),   AV_OPT_TYPE_INT,    { .i64 = -1 }, -1, INT_MAX, VE, "direct-pred" },
676 677 678 679
    { "none",          NULL,      0,    AV_OPT_TYPE_CONST, { .i64 = X264_DIRECT_PRED_NONE },     0, 0, VE, "direct-pred" },
    { "spatial",       NULL,      0,    AV_OPT_TYPE_CONST, { .i64 = X264_DIRECT_PRED_SPATIAL },  0, 0, VE, "direct-pred" },
    { "temporal",      NULL,      0,    AV_OPT_TYPE_CONST, { .i64 = X264_DIRECT_PRED_TEMPORAL }, 0, 0, VE, "direct-pred" },
    { "auto",          NULL,      0,    AV_OPT_TYPE_CONST, { .i64 = X264_DIRECT_PRED_AUTO },     0, 0, VE, "direct-pred" },
680
    { "slice-max-size","Limit the size of each slice in bytes",           OFFSET(slice_max_size),AV_OPT_TYPE_INT,    { .i64 = -1 }, -1, INT_MAX, VE },
681
    { "stats",         "Filename for 2 pass stats",                       OFFSET(stats),         AV_OPT_TYPE_STRING, { 0 },  0,       0, VE },
682 683 684 685 686
    { "nal-hrd",       "Signal HRD information (requires vbv-bufsize; "
                       "cbr not allowed in .mp4)",                        OFFSET(nal_hrd),       AV_OPT_TYPE_INT,    { .i64 = -1 }, -1, INT_MAX, VE, "nal-hrd" },
    { "none",          NULL, 0, AV_OPT_TYPE_CONST, {.i64 = X264_NAL_HRD_NONE}, INT_MIN, INT_MAX, VE, "nal-hrd" },
    { "vbr",           NULL, 0, AV_OPT_TYPE_CONST, {.i64 = X264_NAL_HRD_VBR},  INT_MIN, INT_MAX, VE, "nal-hrd" },
    { "cbr",           NULL, 0, AV_OPT_TYPE_CONST, {.i64 = X264_NAL_HRD_CBR},  INT_MIN, INT_MAX, VE, "nal-hrd" },
687 688 689 690 691 692
    { "motion-est",   "Set motion estimation method",                     OFFSET(motion_est),    AV_OPT_TYPE_INT,    { .i64 = -1 }, -1, X264_ME_TESA, VE, "motion-est"},
    { "dia",           NULL, 0, AV_OPT_TYPE_CONST, { .i64 = X264_ME_DIA },  INT_MIN, INT_MAX, VE, "motion-est" },
    { "hex",           NULL, 0, AV_OPT_TYPE_CONST, { .i64 = X264_ME_HEX },  INT_MIN, INT_MAX, VE, "motion-est" },
    { "umh",           NULL, 0, AV_OPT_TYPE_CONST, { .i64 = X264_ME_UMH },  INT_MIN, INT_MAX, VE, "motion-est" },
    { "esa",           NULL, 0, AV_OPT_TYPE_CONST, { .i64 = X264_ME_ESA },  INT_MIN, INT_MAX, VE, "motion-est" },
    { "tesa",          NULL, 0, AV_OPT_TYPE_CONST, { .i64 = X264_ME_TESA }, INT_MIN, INT_MAX, VE, "motion-est" },
693
    { "x264-params",  "Override the x264 configuration using a :-separated list of key=value parameters", OFFSET(x264_params), AV_OPT_TYPE_STRING, { 0 }, 0, 0, VE },
694 695 696
    { NULL },
};

697 698
static const AVCodecDefault x264_defaults[] = {
    { "b",                "0" },
699 700
    { "bf",               "-1" },
    { "g",                "-1" },
701
    { "i_qfactor",        "-1" },
702 703 704 705 706 707 708
    { "qmin",             "-1" },
    { "qmax",             "-1" },
    { "qdiff",            "-1" },
    { "qblur",            "-1" },
    { "qcomp",            "-1" },
    { "refs",             "-1" },
    { "sc_threshold",     "-1" },
709 710 711
    { "trellis",          "-1" },
    { "nr",               "-1" },
    { "me_range",         "-1" },
712
#if FF_API_MOTION_EST
713
    { "me_method",        "-1" },
714
#endif
715 716 717 718 719
    { "subq",             "-1" },
    { "b_strategy",       "-1" },
    { "keyint_min",       "-1" },
    { "coder",            "-1" },
    { "cmp",              "-1" },
720
    { "threads",          AV_STRINGIFY(X264_THREADS_AUTO) },
721
    { "thread_type",      "0" },
722
    { "flags",            "+cgop" },
723
    { "rc_init_occupancy","-1" },
724 725 726
    { NULL },
};

727 728 729 730 731 732 733 734
#if CONFIG_LIBX264_ENCODER
static const AVClass class = {
    .class_name = "libx264",
    .item_name  = av_default_item_name,
    .option     = options,
    .version    = LIBAVUTIL_VERSION_INT,
};

735
AVCodec ff_libx264_encoder = {
736
    .name             = "libx264",
737
    .long_name        = NULL_IF_CONFIG_SMALL("libx264 H.264 / AVC / MPEG-4 AVC / MPEG-4 part 10"),
738
    .type             = AVMEDIA_TYPE_VIDEO,
739
    .id               = AV_CODEC_ID_H264,
740 741 742 743
    .priv_data_size   = sizeof(X264Context),
    .init             = X264_init,
    .encode2          = X264_frame,
    .close            = X264_close,
744
    .capabilities     = AV_CODEC_CAP_DELAY | AV_CODEC_CAP_AUTO_THREADS,
745 746
    .priv_class       = &class,
    .defaults         = x264_defaults,
747
    .init_static_data = X264_init_static,
748 749
    .caps_internal    = FF_CODEC_CAP_INIT_THREADSAFE |
                        FF_CODEC_CAP_INIT_CLEANUP,
750
};
751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777
#endif

#if CONFIG_LIBX262_ENCODER
static const AVClass X262_class = {
    .class_name = "libx262",
    .item_name  = av_default_item_name,
    .option     = options,
    .version    = LIBAVUTIL_VERSION_INT,
};

AVCodec ff_libx262_encoder = {
    .name             = "libx262",
    .long_name        = NULL_IF_CONFIG_SMALL("libx262 MPEG2VIDEO"),
    .type             = AVMEDIA_TYPE_VIDEO,
    .id               = AV_CODEC_ID_MPEG2VIDEO,
    .priv_data_size   = sizeof(X264Context),
    .init             = X264_init,
    .encode2          = X264_frame,
    .close            = X264_close,
    .capabilities     = CODEC_CAP_DELAY | CODEC_CAP_AUTO_THREADS,
    .priv_class       = &X262_class,
    .defaults         = x264_defaults,
    .pix_fmts         = pix_fmts_8bit,
    .caps_internal    = FF_CODEC_CAP_INIT_THREADSAFE |
                        FF_CODEC_CAP_INIT_CLEANUP,
};
#endif