libx264.c 35.2 KB
Newer Older
1 2
/*
 * H.264 encoding using the x264 library
3
 * Copyright (C) 2005  Mans Rullgard <mans@mansr.com>
4
 *
5 6 7
 * This file is part of FFmpeg.
 *
 * FFmpeg 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
 * FFmpeg 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 FFmpeg; 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/eval.h"
23
#include "libavutil/internal.h"
24
#include "libavutil/opt.h"
25
#include "libavutil/mem.h"
26
#include "libavutil/pixdesc.h"
27
#include "libavutil/stereo3d.h"
28
#include "avcodec.h"
29
#include "internal.h"
30 31 32 33 34

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

35
#include <x264.h>
36
#include <float.h>
Michael Niedermayer's avatar
Michael Niedermayer committed
37
#include <math.h>
Robert Swain's avatar
Robert Swain committed
38 39 40
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
41 42

typedef struct X264Context {
43
    AVClass        *class;
44 45 46 47 48
    x264_param_t    params;
    x264_t         *enc;
    x264_picture_t  pic;
    uint8_t        *sei;
    int             sei_size;
49 50 51 52
    char *preset;
    char *tune;
    char *profile;
    char *level;
53
    int fastfirstpass;
54
    char *wpredp;
55
    char *x264opts;
56
    float crf;
57
    float crf_max;
58
    int cqp;
59
    int aq_mode;
60
    float aq_strength;
61
    char *psy_rd;
62
    int psy;
63
    int rc_lookahead;
64
    int weightp;
65
    int weightb;
66
    int ssim;
67
    int intra_refresh;
68
    int bluray_compat;
69
    int b_bias;
70
    int b_pyramid;
71
    int mixed_refs;
72
    int dct8x8;
73
    int fast_pskip;
74
    int aud;
75
    int mbtree;
76
    char *deblock;
77
    float cplxblur;
78
    char *partitions;
79
    int direct_pred;
80
    int slice_max_size;
81
    char *stats;
Anton Khirnov's avatar
Anton Khirnov committed
82
    int nal_hrd;
83
    int avcintra_class;
84
    char *x264_params;
85 86
} X264Context;

87
static void X264_log(void *p, int level, const char *fmt, va_list args)
88 89
{
    static const int level_map[] = {
90
        [X264_LOG_ERROR]   = AV_LOG_ERROR,
91
        [X264_LOG_WARNING] = AV_LOG_WARNING,
92
        [X264_LOG_INFO]    = AV_LOG_INFO,
93
        [X264_LOG_DEBUG]   = AV_LOG_DEBUG
94 95
    };

96
    if (level < 0 || level > X264_LOG_DEBUG)
97
        return;
98 99 100 101 102

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


103
static int encode_nals(AVCodecContext *ctx, AVPacket *pkt,
104
                       const x264_nal_t *nals, int nnal)
105
{
106
    X264Context *x4 = ctx->priv_data;
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;

116
    if ((ret = ff_alloc_packet2(ctx, pkt, size)) < 0)
117 118 119
        return ret;

    p = pkt->data;
120 121

    /* Write the SEI as part of the first frame. */
122
    if (x4->sei_size > 0 && nnal > 0) {
123 124 125 126
        if (x4->sei_size > size) {
            av_log(ctx, AV_LOG_ERROR, "Error: nal buffer is too small\n");
            return -1;
        }
127 128 129
        memcpy(p, x4->sei, x4->sei_size);
        p += x4->sei_size;
        x4->sei_size = 0;
130
        av_freep(&x4->sei);
131
    }
132

133
    for (i = 0; i < nnal; i++){
134 135
        memcpy(p, nals[i].p_payload, nals[i].i_payload);
        p += nals[i].i_payload;
136 137
    }

138
    return 1;
139 140
}

141 142 143
static int avfmt2_num_planes(int avfmt)
{
    switch (avfmt) {
144 145 146 147 148
    case AV_PIX_FMT_YUV420P:
    case AV_PIX_FMT_YUVJ420P:
    case AV_PIX_FMT_YUV420P9:
    case AV_PIX_FMT_YUV420P10:
    case AV_PIX_FMT_YUV444P:
149 150
        return 3;

151 152
    case AV_PIX_FMT_BGR24:
    case AV_PIX_FMT_RGB24:
153 154 155 156 157 158 159
        return 1;

    default:
        return 3;
    }
}

160 161
static int X264_frame(AVCodecContext *ctx, AVPacket *pkt, const AVFrame *frame,
                      int *got_packet)
162 163 164
{
    X264Context *x4 = ctx->priv_data;
    x264_nal_t *nal;
165
    int nnal, i, ret;
166
    x264_picture_t pic_out = {0};
167
    AVFrameSideData *side_data;
168

169
    x264_picture_init( &x4->pic );
170
    x4->pic.img.i_csp   = x4->params.i_csp;
171 172
    if (x264_bit_depth > 8)
        x4->pic.img.i_csp |= X264_CSP_HIGH_DEPTH;
173
    x4->pic.img.i_plane = avfmt2_num_planes(ctx->pix_fmt);
174

175
    if (frame) {
176
        for (i = 0; i < x4->pic.img.i_plane; i++) {
177
            x4->pic.img.plane[i]    = frame->data[i];
Måns Rullgård's avatar
Måns Rullgård committed
178 179
            x4->pic.img.i_stride[i] = frame->linesize[i];
        }
180

181
        x4->pic.i_pts  = frame->pts;
182
        x4->pic.i_type =
183 184 185
            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 :
186
                                            X264_TYPE_AUTO;
187 188

        if (x4->avcintra_class < 0) {
189
        if (x4->params.b_interlaced && x4->params.b_tff != frame->top_field_first) {
190 191 192
            x4->params.b_tff = frame->top_field_first;
            x264_encoder_reconfig(x4->enc, &x4->params);
        }
193 194
        if (x4->params.vui.i_sar_height != ctx->sample_aspect_ratio.den ||
            x4->params.vui.i_sar_width  != ctx->sample_aspect_ratio.num) {
195
            x4->params.vui.i_sar_height = ctx->sample_aspect_ratio.den;
196
            x4->params.vui.i_sar_width  = ctx->sample_aspect_ratio.num;
197
            x264_encoder_reconfig(x4->enc, &x4->params);
198
        }
199

200 201 202 203 204 205 206 207 208 209 210 211 212
        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);
        }

213 214
        if (x4->crf >= 0 &&
            x4->params.rc.i_rc_method == X264_RC_CRF &&
215 216 217 218 219 220
            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 &&
221
            x4->cqp >= 0 &&
222 223 224 225 226
            x4->params.rc.i_qp_constant != x4->cqp) {
            x4->params.rc.i_qp_constant = x4->cqp;
            x264_encoder_reconfig(x4->enc, &x4->params);
        }

227
        if (x4->crf_max >= 0 &&
228 229 230
            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);
231
        }
232
        }
233

234 235 236 237 238 239 240 241 242
        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;
243
            case AV_STEREO3D_COLUMNS:
244 245
                fpa_type = 1;
                break;
246
            case AV_STEREO3D_LINES:
247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268
                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);
            }
        }
    }
269
    do {
270 271
        if (x264_encoder_encode(x4->enc, &nal, &nnal, frame? &x4->pic: NULL, &pic_out) < 0)
            return -1;
272

273 274
        ret = encode_nals(ctx, pkt, nal, nnal);
        if (ret < 0)
275
            return -1;
276
    } while (!ret && !frame && x264_encoder_delayed_frames(x4->enc));
277

278 279
    pkt->pts = pic_out.i_pts;
    pkt->dts = pic_out.i_dts;
280

281
    switch (pic_out.i_type) {
282 283
    case X264_TYPE_IDR:
    case X264_TYPE_I:
284
        ctx->coded_frame->pict_type = AV_PICTURE_TYPE_I;
285 286
        break;
    case X264_TYPE_P:
287
        ctx->coded_frame->pict_type = AV_PICTURE_TYPE_P;
288 289 290
        break;
    case X264_TYPE_B:
    case X264_TYPE_BREF:
291
        ctx->coded_frame->pict_type = AV_PICTURE_TYPE_B;
292 293 294
        break;
    }

295 296
    pkt->flags |= AV_PKT_FLAG_KEY*pic_out.b_keyframe;
    if (ret)
297
        ctx->coded_frame->quality = (pic_out.i_qpplus1 - 1) * FF_QP2LAMBDA;
298

299 300
    *got_packet = ret;
    return 0;
301 302
}

303
static av_cold int X264_close(AVCodecContext *avctx)
304 305 306
{
    X264Context *x4 = avctx->priv_data;

Ralf Terdic's avatar
Ralf Terdic committed
307
    av_freep(&avctx->extradata);
308
    av_freep(&x4->sei);
Ralf Terdic's avatar
Ralf Terdic committed
309

310
    if (x4->enc)
311
        x264_encoder_close(x4->enc);
312

313 314
    av_frame_free(&avctx->coded_frame);

315 316 317
    return 0;
}

318 319 320
#define OPT_STR(opt, param)                                                   \
    do {                                                                      \
        int ret;                                                              \
321
        if (param && (ret = x264_param_parse(&x4->params, opt, param)) < 0) { \
322 323 324 325 326 327 328 329
            if(ret == X264_PARAM_BAD_NAME)                                    \
                av_log(avctx, AV_LOG_ERROR,                                   \
                        "bad option '%s': '%s'\n", opt, param);               \
            else                                                              \
                av_log(avctx, AV_LOG_ERROR,                                   \
                        "bad value for '%s': '%s'\n", opt, param);            \
            return -1;                                                        \
        }                                                                     \
330
    } while (0)
331

332
static int convert_pix_fmt(enum AVPixelFormat pix_fmt)
333 334
{
    switch (pix_fmt) {
335 336 337 338 339
    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:
340
    case AV_PIX_FMT_YUVJ422P:
341 342
    case AV_PIX_FMT_YUV422P10: return X264_CSP_I422;
    case AV_PIX_FMT_YUV444P:
343
    case AV_PIX_FMT_YUVJ444P:
344 345
    case AV_PIX_FMT_YUV444P9:
    case AV_PIX_FMT_YUV444P10: return X264_CSP_I444;
346
#ifdef X264_CSP_BGR
347
    case AV_PIX_FMT_BGR24:
348 349
        return X264_CSP_BGR;

350
    case AV_PIX_FMT_RGB24:
351 352
        return X264_CSP_RGB;
#endif
353 354 355
    case AV_PIX_FMT_NV12:      return X264_CSP_NV12;
    case AV_PIX_FMT_NV16:
    case AV_PIX_FMT_NV20:      return X264_CSP_NV16;
356 357 358 359
    };
    return 0;
}

360 361 362 363 364 365
#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);\
    }

366
static av_cold int X264_init(AVCodecContext *avctx)
367 368
{
    X264Context *x4 = avctx->priv_data;
369
    int sw,sh;
370

371 372 373
    if (avctx->global_quality > 0)
        av_log(avctx, AV_LOG_WARNING, "-qscale is ignored, -crf is recommended.\n");

374 375
    x264_param_default(&x4->params);

376
    x4->params.b_deblocking_filter         = avctx->flags & CODEC_FLAG_LOOP_FILTER;
377

378 379
    if (x4->preset || x4->tune)
        if (x264_param_default_preset(&x4->params, x4->preset, x4->tune) < 0) {
380
            int i;
381
            av_log(avctx, AV_LOG_ERROR, "Error setting preset/tune %s/%s.\n", x4->preset, x4->tune);
382 383 384 385 386 387 388 389
            av_log(avctx, AV_LOG_INFO, "Possible presets:");
            for (i = 0; x264_preset_names[i]; i++)
                av_log(avctx, AV_LOG_INFO, " %s", x264_preset_names[i]);
            av_log(avctx, AV_LOG_INFO, "\n");
            av_log(avctx, AV_LOG_INFO, "Possible tunes:");
            for (i = 0; x264_tune_names[i]; i++)
                av_log(avctx, AV_LOG_INFO, " %s", x264_tune_names[i]);
            av_log(avctx, AV_LOG_INFO, "\n");
390 391 392
            return AVERROR(EINVAL);
        }

393 394
    if (avctx->level > 0)
        x4->params.i_level_idc = avctx->level;
395 396 397 398

    x4->params.pf_log               = X264_log;
    x4->params.p_log_private        = avctx;
    x4->params.i_log_level          = X264_LOG_DEBUG;
399
    x4->params.i_csp                = convert_pix_fmt(avctx->pix_fmt);
400

401
    OPT_STR("weightp", x4->wpredp);
402

403 404 405 406
    if (avctx->bit_rate) {
        x4->params.rc.i_bitrate   = avctx->bit_rate / 1000;
        x4->params.rc.i_rc_method = X264_RC_ABR;
    }
407 408 409 410 411 412
    x4->params.rc.i_vbv_buffer_size = avctx->rc_buffer_size / 1000;
    x4->params.rc.i_vbv_max_bitrate = avctx->rc_max_rate    / 1000;
    x4->params.rc.b_stat_write      = avctx->flags & CODEC_FLAG_PASS1;
    if (avctx->flags & CODEC_FLAG_PASS2) {
        x4->params.rc.b_stat_read = 1;
    } else {
413 414 415
        if (x4->crf >= 0) {
            x4->params.rc.i_rc_method   = X264_RC_CRF;
            x4->params.rc.f_rf_constant = x4->crf;
416 417 418
        } else if (x4->cqp >= 0) {
            x4->params.rc.i_rc_method   = X264_RC_CQP;
            x4->params.rc.i_qp_constant = x4->cqp;
419
        }
420 421 422

        if (x4->crf_max >= 0)
            x4->params.rc.f_rf_constant_max = x4->crf_max;
423 424
    }

425
    if (avctx->rc_buffer_size && avctx->rc_initial_buffer_occupancy > 0 &&
426
        (avctx->rc_initial_buffer_occupancy <= avctx->rc_buffer_size)) {
Robert Swain's avatar
Robert Swain committed
427
        x4->params.rc.f_vbv_buffer_init =
428
            (float)avctx->rc_initial_buffer_occupancy / avctx->rc_buffer_size;
429
    }
Robert Swain's avatar
Robert Swain committed
430

431 432
    OPT_STR("level", x4->level);

433 434
    if (avctx->i_quant_factor > 0)
        x4->params.rc.f_ip_factor         = 1 / fabs(avctx->i_quant_factor);
435 436 437 438
    if (avctx->b_quant_factor > 0)
        x4->params.rc.f_pb_factor         = avctx->b_quant_factor;
    if (avctx->chromaoffset)
        x4->params.analyse.i_chroma_qp_offset = avctx->chromaoffset;
Robert Swain's avatar
Robert Swain committed
439

440 441 442 443 444 445 446 447 448 449 450
    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;

451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468
    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;
469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490
    else if (x4->level) {
        int i;
        int mbn = FF_CEIL_RSHIFT(avctx->width, 4) * FF_CEIL_RSHIFT(avctx->height, 4);
        int level_id = -1;
        char *tail;
        int scale = X264_BUILD < 129 ? 384 : 1;

        if (!strcmp(x4->level, "1b")) {
            level_id = 9;
        } else if (strlen(x4->level) <= 3){
            level_id = av_strtod(x4->level, &tail) * 10 + 0.5;
            if (*tail)
                level_id = -1;
        }
        if (level_id <= 0)
            av_log(avctx, AV_LOG_WARNING, "Failed to parse level\n");

        for (i = 0; i<x264_levels[i].level_idc; i++)
            if (x264_levels[i].level_idc == level_id)
                x4->params.i_frame_reference = av_clip(x264_levels[i].dpb / mbn / scale, 1, x4->params.i_frame_reference);
    }

491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506
    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;
507

508 509
    if (x4->aq_mode >= 0)
        x4->params.rc.i_aq_mode = x4->aq_mode;
510 511
    if (x4->aq_strength >= 0)
        x4->params.rc.f_aq_strength = x4->aq_strength;
512 513
    PARSE_X264_OPT("psy-rd", psy_rd);
    PARSE_X264_OPT("deblock", deblock);
514
    PARSE_X264_OPT("partitions", partitions);
515
    PARSE_X264_OPT("stats", stats);
516 517
    if (x4->psy >= 0)
        x4->params.analyse.b_psy  = x4->psy;
518 519
    if (x4->rc_lookahead >= 0)
        x4->params.rc.i_lookahead = x4->rc_lookahead;
520 521
    if (x4->weightp >= 0)
        x4->params.analyse.i_weighted_pred = x4->weightp;
522 523
    if (x4->weightb >= 0)
        x4->params.analyse.b_weighted_bipred = x4->weightb;
524 525
    if (x4->cplxblur >= 0)
        x4->params.rc.f_complexity_blur = x4->cplxblur;
526

527 528
    if (x4->ssim >= 0)
        x4->params.analyse.b_ssim = x4->ssim;
529 530
    if (x4->intra_refresh >= 0)
        x4->params.b_intra_refresh = x4->intra_refresh;
531 532 533 534
    if (x4->bluray_compat >= 0) {
        x4->params.b_bluray_compat = x4->bluray_compat;
        x4->params.b_vfr_input = 0;
    }
535
    if (x4->avcintra_class >= 0)
536
#if X264_BUILD >= 142
537
        x4->params.i_avcintra_class = x4->avcintra_class;
538 539 540 541
#else
        av_log(avctx, AV_LOG_ERROR,
               "x264 too old for AVC Intra, at least version 142 needed\n");
#endif
542 543
    if (x4->b_bias != INT_MIN)
        x4->params.i_bframe_bias              = x4->b_bias;
544 545
    if (x4->b_pyramid >= 0)
        x4->params.i_bframe_pyramid = x4->b_pyramid;
546 547
    if (x4->mixed_refs >= 0)
        x4->params.analyse.b_mixed_references = x4->mixed_refs;
548 549
    if (x4->dct8x8 >= 0)
        x4->params.analyse.b_transform_8x8    = x4->dct8x8;
550 551
    if (x4->fast_pskip >= 0)
        x4->params.analyse.b_fast_pskip       = x4->fast_pskip;
552 553
    if (x4->aud >= 0)
        x4->params.b_aud                      = x4->aud;
554 555
    if (x4->mbtree >= 0)
        x4->params.rc.b_mb_tree               = x4->mbtree;
556 557
    if (x4->direct_pred >= 0)
        x4->params.analyse.i_direct_mv_pred   = x4->direct_pred;
558

559 560
    if (x4->slice_max_size >= 0)
        x4->params.i_slice_max_size =  x4->slice_max_size;
561 562 563 564 565 566 567 568 569 570
    else {
        /*
         * Allow x264 to be instructed through AVCodecContext about the maximum
         * size of the RTP payload. For example, this enables the production of
         * payload suitable for the H.264 RTP packetization-mode 0 i.e. single
         * NAL unit per RTP packet.
         */
        if (avctx->rtp_payload_size)
            x4->params.i_slice_max_size = avctx->rtp_payload_size;
    }
571

572 573 574
    if (x4->fastfirstpass)
        x264_param_apply_fastfirstpass(&x4->params);

575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598
    /* Allow specifying the x264 profile through AVCodecContext. */
    if (!x4->profile)
        switch (avctx->profile) {
        case FF_PROFILE_H264_BASELINE:
            x4->profile = av_strdup("baseline");
            break;
        case FF_PROFILE_H264_HIGH:
            x4->profile = av_strdup("high");
            break;
        case FF_PROFILE_H264_HIGH_10:
            x4->profile = av_strdup("high10");
            break;
        case FF_PROFILE_H264_HIGH_422:
            x4->profile = av_strdup("high422");
            break;
        case FF_PROFILE_H264_HIGH_444:
            x4->profile = av_strdup("high444");
            break;
        case FF_PROFILE_H264_MAIN:
            x4->profile = av_strdup("main");
            break;
        default:
            break;
        }
599

600 601 602
    if (x4->nal_hrd >= 0)
        x4->params.i_nal_hrd = x4->nal_hrd;

603
    if (x4->profile)
604
        if (x264_param_apply_profile(&x4->params, x4->profile) < 0) {
605
            int i;
606
            av_log(avctx, AV_LOG_ERROR, "Error setting profile %s.\n", x4->profile);
607 608
            av_log(avctx, AV_LOG_INFO, "Possible profiles:");
            for (i = 0; x264_profile_names[i]; i++)
Mahendra M's avatar
Mahendra M committed
609
                av_log(avctx, AV_LOG_INFO, " %s", x264_profile_names[i]);
610
            av_log(avctx, AV_LOG_INFO, "\n");
611 612
            return AVERROR(EINVAL);
        }
613

614 615
    x4->params.i_width          = avctx->width;
    x4->params.i_height         = avctx->height;
616 617 618
    av_reduce(&sw, &sh, avctx->sample_aspect_ratio.num, avctx->sample_aspect_ratio.den, 4096);
    x4->params.vui.i_sar_width  = sw;
    x4->params.vui.i_sar_height = sh;
619 620 621 622
    x4->params.i_timebase_den = avctx->time_base.den;
    x4->params.i_timebase_num = avctx->time_base.num;
    x4->params.i_fps_num = avctx->time_base.den;
    x4->params.i_fps_den = avctx->time_base.num * avctx->ticks_per_frame;
Robert Swain's avatar
Robert Swain committed
623

624
    x4->params.analyse.b_psnr = avctx->flags & CODEC_FLAG_PSNR;
Robert Swain's avatar
Robert Swain committed
625

626
    x4->params.i_threads      = avctx->thread_count;
627 628
    if (avctx->thread_type)
        x4->params.b_sliced_threads = avctx->thread_type == FF_THREAD_SLICE;
629

630
    x4->params.b_interlaced   = avctx->flags & CODEC_FLAG_INTERLACED_DCT;
631

632
    x4->params.b_open_gop     = !(avctx->flags & CODEC_FLAG_CLOSED_GOP);
633

634 635
    x4->params.i_slice_count  = avctx->slices;

636 637
    x4->params.vui.b_fullrange = avctx->pix_fmt == AV_PIX_FMT_YUVJ420P ||
                                 avctx->pix_fmt == AV_PIX_FMT_YUVJ422P ||
638 639 640 641 642 643 644 645 646
                                 avctx->pix_fmt == AV_PIX_FMT_YUVJ444P ||
                                 avctx->color_range == AVCOL_RANGE_JPEG;

    if (avctx->colorspace != AVCOL_SPC_UNSPECIFIED)
        x4->params.vui.i_colmatrix = avctx->colorspace;
    if (avctx->color_primaries != AVCOL_PRI_UNSPECIFIED)
        x4->params.vui.i_colorprim = avctx->color_primaries;
    if (avctx->color_trc != AVCOL_TRC_UNSPECIFIED)
        x4->params.vui.i_transfer  = avctx->color_trc;
647

648
    if (avctx->flags & CODEC_FLAG_GLOBAL_HEADER)
649 650
        x4->params.b_repeat_headers = 0;

651 652 653 654 655 656 657 658 659 660 661 662 663
    if(x4->x264opts){
        const char *p= x4->x264opts;
        while(p){
            char param[256]={0}, val[256]={0};
            if(sscanf(p, "%255[^:=]=%255[^:]", param, val) == 1){
                OPT_STR(param, "1");
            }else
                OPT_STR(param, val);
            p= strchr(p, ':');
            p+=!!p;
        }
    }

664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679
    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);
        }
    }

680
    // update AVCodecContext with x264 parameters
681 682
    avctx->has_b_frames = x4->params.i_bframe ?
        x4->params.i_bframe_pyramid ? 2 : 1 : 0;
683 684 685
    if (avctx->max_b_frames < 0)
        avctx->max_b_frames = 0;

686 687
    avctx->bit_rate = x4->params.rc.i_bitrate*1000;

688
    x4->enc = x264_encoder_open(&x4->params);
689
    if (!x4->enc)
690 691
        return -1;

692 693 694
    avctx->coded_frame = av_frame_alloc();
    if (!avctx->coded_frame)
        return AVERROR(ENOMEM);
695

696
    if (avctx->flags & CODEC_FLAG_GLOBAL_HEADER) {
697
        x264_nal_t *nal;
698
        uint8_t *p;
699
        int nnal, s, i;
700

701
        s = x264_encoder_headers(x4->enc, &nal, &nnal);
702
        avctx->extradata = p = av_malloc(s);
703 704
        if (!p)
            goto nomem;
705

706 707 708
        for (i = 0; i < nnal; i++) {
            /* Don't put the SEI in extradata. */
            if (nal[i].i_type == NAL_SEI) {
709
                av_log(avctx, AV_LOG_INFO, "%s\n", nal[i].p_payload+25);
710 711
                x4->sei_size = nal[i].i_payload;
                x4->sei      = av_malloc(x4->sei_size);
712 713
                if (!x4->sei)
                    goto nomem;
714 715 716 717 718 719 720
                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;
721 722
    }

723
    return 0;
724 725 726
nomem:
    X264_close(avctx);
    return AVERROR(ENOMEM);
727 728
}

729 730 731 732
static const enum AVPixelFormat pix_fmts_8bit[] = {
    AV_PIX_FMT_YUV420P,
    AV_PIX_FMT_YUVJ420P,
    AV_PIX_FMT_YUV422P,
733
    AV_PIX_FMT_YUVJ422P,
734
    AV_PIX_FMT_YUV444P,
735
    AV_PIX_FMT_YUVJ444P,
736 737
    AV_PIX_FMT_NV12,
    AV_PIX_FMT_NV16,
738
    AV_PIX_FMT_NONE
739
};
740 741 742 743
static const enum AVPixelFormat pix_fmts_9bit[] = {
    AV_PIX_FMT_YUV420P9,
    AV_PIX_FMT_YUV444P9,
    AV_PIX_FMT_NONE
744
};
745 746 747 748
static const enum AVPixelFormat pix_fmts_10bit[] = {
    AV_PIX_FMT_YUV420P10,
    AV_PIX_FMT_YUV422P10,
    AV_PIX_FMT_YUV444P10,
749
    AV_PIX_FMT_NV20,
750
    AV_PIX_FMT_NONE
751
};
752
static const enum AVPixelFormat pix_fmts_8bit_rgb[] = {
753
#ifdef X264_CSP_BGR
754 755
    AV_PIX_FMT_BGR24,
    AV_PIX_FMT_RGB24,
756
#endif
757
    AV_PIX_FMT_NONE
758
};
759 760 761 762 763 764 765 766 767 768 769

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;
}

770
#define OFFSET(x) offsetof(X264Context, x)
771 772
#define VE AV_OPT_FLAG_VIDEO_PARAM | AV_OPT_FLAG_ENCODING_PARAM
static const AVOption options[] = {
773 774 775
    { "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},
776
    { "fastfirstpass", "Use fast settings when encoding first pass",      OFFSET(fastfirstpass), AV_OPT_TYPE_INT,    { .i64 = 1 }, 0, 1, VE},
777 778 779 780
    {"level", "Specify level (as defined by Annex A)", OFFSET(level), AV_OPT_TYPE_STRING, {.str=NULL}, 0, 0, VE},
    {"passlogfile", "Filename for 2 pass stats", OFFSET(stats), AV_OPT_TYPE_STRING, {.str=NULL}, 0, 0, VE},
    {"wpredp", "Weighted prediction for P-frames", OFFSET(wpredp), AV_OPT_TYPE_STRING, {.str=NULL}, 0, 0, VE},
    {"x264opts", "x264 options", OFFSET(x264opts), AV_OPT_TYPE_STRING, {.str=NULL}, 0, 0, VE},
781 782
    { "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 },
783 784
    { "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"},
785 786 787
    { "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" },
788
    { "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},
789
    { "psy",           "Use psychovisual optimizations.",                 OFFSET(psy),           AV_OPT_TYPE_INT,    { .i64 = -1 }, -1, 1, VE },
790
    { "psy-rd",        "Strength of psychovisual optimization, in <psy-rd>:<psy-trellis> format.", OFFSET(psy_rd), AV_OPT_TYPE_STRING,  {0 }, 0, 0, VE},
791 792 793
    { "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" },
794 795 796
    { "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" },
797 798
    { "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 },
799
    { "bluray-compat", "Bluray compatibility workarounds.",               OFFSET(bluray_compat) ,AV_OPT_TYPE_INT,    { .i64 = -1 }, -1, 1, VE },
800 801
    { "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" },
802 803 804
    { "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" },
805 806 807 808 809
    { "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},
810
    { "deblock",       "Loop filter parameters, in <alpha:beta> form.",   OFFSET(deblock),       AV_OPT_TYPE_STRING, { 0 },  0, 0, VE},
811
    { "cplxblur",      "Reduce fluctuations in QP (before curve compression)", OFFSET(cplxblur), AV_OPT_TYPE_FLOAT,  {.dbl = -1 }, -1, FLT_MAX, VE},
812
    { "partitions",    "A comma-separated list of partitions to consider. "
813
                       "Possible values: p8x8, p4x4, b8x8, i8x8, i4x4, none, all", OFFSET(partitions), AV_OPT_TYPE_STRING, { 0 }, 0, 0, VE},
814
    { "direct-pred",   "Direct MV prediction mode",                       OFFSET(direct_pred),   AV_OPT_TYPE_INT,    { .i64 = -1 }, -1, INT_MAX, VE, "direct-pred" },
815 816 817 818
    { "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" },
819
    { "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 },
820
    { "stats",         "Filename for 2 pass stats",                       OFFSET(stats),         AV_OPT_TYPE_STRING, { 0 },  0,       0, VE },
821 822 823 824 825
    { "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" },
826
    { "avcintra-class","AVC-Intra class 50/100/200",                      OFFSET(avcintra_class),AV_OPT_TYPE_INT,     { .i64 = -1 }, -1, 200   , VE},
827
    { "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 },
828 829 830
    { NULL },
};

831
static const AVClass x264_class = {
832 833 834 835 836
    .class_name = "libx264",
    .item_name  = av_default_item_name,
    .option     = options,
    .version    = LIBAVUTIL_VERSION_INT,
};
837

838 839 840 841 842 843 844
static const AVClass rgbclass = {
    .class_name = "libx264rgb",
    .item_name  = av_default_item_name,
    .option     = options,
    .version    = LIBAVUTIL_VERSION_INT,
};

845 846
static const AVCodecDefault x264_defaults[] = {
    { "b",                "0" },
847
    { "bf",               "-1" },
848
    { "flags2",           "0" },
849
    { "g",                "-1" },
850
    { "i_qfactor",        "-1" },
851
    { "b_qfactor",        "-1" },
852 853 854 855 856
    { "qmin",             "-1" },
    { "qmax",             "-1" },
    { "qdiff",            "-1" },
    { "qblur",            "-1" },
    { "qcomp",            "-1" },
857
//     { "rc_lookahead",     "-1" },
858 859
    { "refs",             "-1" },
    { "sc_threshold",     "-1" },
860 861 862 863 864 865 866 867 868
    { "trellis",          "-1" },
    { "nr",               "-1" },
    { "me_range",         "-1" },
    { "me_method",        "-1" },
    { "subq",             "-1" },
    { "b_strategy",       "-1" },
    { "keyint_min",       "-1" },
    { "coder",            "-1" },
    { "cmp",              "-1" },
869
    { "threads",          AV_STRINGIFY(X264_THREADS_AUTO) },
870
    { "thread_type",      "0" },
871
    { "flags",            "+cgop" },
872
    { "rc_init_occupancy","-1" },
873 874 875
    { NULL },
};

876
AVCodec ff_libx264_encoder = {
877
    .name             = "libx264",
878
    .long_name        = NULL_IF_CONFIG_SMALL("libx264 H.264 / AVC / MPEG-4 AVC / MPEG-4 part 10"),
879
    .type             = AVMEDIA_TYPE_VIDEO,
880
    .id               = AV_CODEC_ID_H264,
881 882 883 884 885
    .priv_data_size   = sizeof(X264Context),
    .init             = X264_init,
    .encode2          = X264_frame,
    .close            = X264_close,
    .capabilities     = CODEC_CAP_DELAY | CODEC_CAP_AUTO_THREADS,
886
    .priv_class       = &x264_class,
887
    .defaults         = x264_defaults,
888
    .init_static_data = X264_init_static,
889
};
890 891 892

AVCodec ff_libx264rgb_encoder = {
    .name           = "libx264rgb",
893
    .long_name      = NULL_IF_CONFIG_SMALL("libx264 H.264 / AVC / MPEG-4 AVC / MPEG-4 part 10 RGB"),
894
    .type           = AVMEDIA_TYPE_VIDEO,
895
    .id             = AV_CODEC_ID_H264,
896 897
    .priv_data_size = sizeof(X264Context),
    .init           = X264_init,
898
    .encode2        = X264_frame,
899
    .close          = X264_close,
900
    .capabilities   = CODEC_CAP_DELAY | CODEC_CAP_AUTO_THREADS,
901 902 903 904
    .priv_class     = &rgbclass,
    .defaults       = x264_defaults,
    .pix_fmts       = pix_fmts_8bit_rgb,
};