ivfenc.c 36.7 KB
Newer Older
John Koleszar's avatar
John Koleszar committed
1 2 3
/*
 *  Copyright (c) 2010 The VP8 project authors. All Rights Reserved.
 *
4
 *  Use of this source code is governed by a BSD-style license
5 6
 *  that can be found in the LICENSE file in the root of the source
 *  tree. An additional intellectual property rights grant can be found
7
 *  in the file PATENTS.  All contributing project authors may
8
 *  be found in the AUTHORS file in the root of the source tree.
John Koleszar's avatar
John Koleszar committed
9 10 11 12 13 14
 */


/* This is a simple program that encodes YV12 files and generates ivf
 * files using the new interface.
 */
15
#if defined(_WIN32)
16 17 18 19
#define USE_POSIX_MMAP 0
#else
#define USE_POSIX_MMAP 1
#endif
John Koleszar's avatar
John Koleszar committed
20 21 22 23 24

#include <stdio.h>
#include <stdlib.h>
#include <stdarg.h>
#include <string.h>
25
#include "vpx/vpx_encoder.h"
John Koleszar's avatar
John Koleszar committed
26 27 28 29 30 31 32
#if USE_POSIX_MMAP
#include <sys/types.h>
#include <sys/stat.h>
#include <sys/mman.h>
#include <fcntl.h>
#include <unistd.h>
#endif
33
#include "vpx_config.h"
34
#include "vpx/vp8cx.h"
John Koleszar's avatar
John Koleszar committed
35 36
#include "vpx_ports/mem_ops.h"
#include "vpx_ports/vpx_timer.h"
37
#include "y4minput.h"
John Koleszar's avatar
John Koleszar committed
38 39 40 41 42 43 44 45 46 47

static const char *exec_name;

static const struct codec_item
{
    char const              *name;
    const vpx_codec_iface_t *iface;
    unsigned int             fourcc;
} codecs[] =
{
48 49 50
#if CONFIG_EXPERIMENTAL && CONFIG_VP8_ENCODER
    {"vp8x",  &vpx_codec_vp8x_cx_algo, 0x78385056},
#endif
John Koleszar's avatar
John Koleszar committed
51 52 53 54 55 56 57 58 59 60 61
#if CONFIG_VP8_ENCODER
    {"vp8",  &vpx_codec_vp8_cx_algo, 0x30385056},
#endif
};

static void usage_exit();

void die(const char *fmt, ...)
{
    va_list ap;
    va_start(ap, fmt);
62 63
    vfprintf(stderr, fmt, ap);
    fprintf(stderr, "\n");
John Koleszar's avatar
John Koleszar committed
64 65 66 67 68 69 70 71 72
    usage_exit();
}

static void ctx_exit_on_error(vpx_codec_ctx_t *ctx, const char *s)
{
    if (ctx->err)
    {
        const char *detail = vpx_codec_error_detail(ctx);

73
        fprintf(stderr, "%s: %s\n", s, vpx_codec_error(ctx));
John Koleszar's avatar
John Koleszar committed
74 75

        if (detail)
76
            fprintf(stderr, "    %s\n", detail);
John Koleszar's avatar
John Koleszar committed
77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225

        exit(EXIT_FAILURE);
    }
}

/* This structure is used to abstract the different ways of handling
 * first pass statistics.
 */
typedef struct
{
    vpx_fixed_buf_t buf;
    int             pass;
    FILE           *file;
    char           *buf_ptr;
    size_t          buf_alloc_sz;
} stats_io_t;

int stats_open_file(stats_io_t *stats, const char *fpf, int pass)
{
    int res;

    stats->pass = pass;

    if (pass == 0)
    {
        stats->file = fopen(fpf, "wb");
        stats->buf.sz = 0;
        stats->buf.buf = NULL,
                   res = (stats->file != NULL);
    }
    else
    {
#if 0
#elif USE_POSIX_MMAP
        struct stat stat_buf;
        int fd;

        fd = open(fpf, O_RDONLY);
        stats->file = fdopen(fd, "rb");
        fstat(fd, &stat_buf);
        stats->buf.sz = stat_buf.st_size;
        stats->buf.buf = mmap(NULL, stats->buf.sz, PROT_READ, MAP_PRIVATE,
                              fd, 0);
        res = (stats->buf.buf != NULL);
#else
        size_t nbytes;

        stats->file = fopen(fpf, "rb");

        if (fseek(stats->file, 0, SEEK_END))
        {
            fprintf(stderr, "First-pass stats file must be seekable!\n");
            exit(EXIT_FAILURE);
        }

        stats->buf.sz = stats->buf_alloc_sz = ftell(stats->file);
        rewind(stats->file);

        stats->buf.buf = malloc(stats->buf_alloc_sz);

        if (!stats->buf.buf)
        {
            fprintf(stderr, "Failed to allocate first-pass stats buffer (%d bytes)\n",
                    stats->buf_alloc_sz);
            exit(EXIT_FAILURE);
        }

        nbytes = fread(stats->buf.buf, 1, stats->buf.sz, stats->file);
        res = (nbytes == stats->buf.sz);
#endif
    }

    return res;
}

int stats_open_mem(stats_io_t *stats, int pass)
{
    int res;
    stats->pass = pass;

    if (!pass)
    {
        stats->buf.sz = 0;
        stats->buf_alloc_sz = 64 * 1024;
        stats->buf.buf = malloc(stats->buf_alloc_sz);
    }

    stats->buf_ptr = stats->buf.buf;
    res = (stats->buf.buf != NULL);
    return res;
}


void stats_close(stats_io_t *stats)
{
    if (stats->file)
    {
        if (stats->pass == 1)
        {
#if 0
#elif USE_POSIX_MMAP
            munmap(stats->buf.buf, stats->buf.sz);
#else
            free(stats->buf.buf);
#endif
        }

        fclose(stats->file);
        stats->file = NULL;
    }
    else
    {
        if (stats->pass == 1)
            free(stats->buf.buf);
    }
}

void stats_write(stats_io_t *stats, const void *pkt, size_t len)
{
    if (stats->file)
    {
        fwrite(pkt, 1, len, stats->file);
    }
    else
    {
        if (stats->buf.sz + len > stats->buf_alloc_sz)
        {
            size_t  new_sz = stats->buf_alloc_sz + 64 * 1024;
            char   *new_ptr = realloc(stats->buf.buf, new_sz);

            if (new_ptr)
            {
                stats->buf_ptr = new_ptr + (stats->buf_ptr - (char *)stats->buf.buf);
                stats->buf.buf = new_ptr;
                stats->buf_alloc_sz = new_sz;
            } /* else ... */
        }

        memcpy(stats->buf_ptr, pkt, len);
        stats->buf.sz += len;
        stats->buf_ptr += len;
    }
}

vpx_fixed_buf_t stats_get(stats_io_t *stats)
{
    return stats->buf;
}

226 227 228 229 230 231 232
enum video_file_type
{
    FILE_TYPE_RAW,
    FILE_TYPE_IVF,
    FILE_TYPE_Y4M
};

233 234 235 236 237 238
struct detect_buffer {
    char buf[4];
    int  valid;
};


John Koleszar's avatar
John Koleszar committed
239
#define IVF_FRAME_HDR_SZ (4+8) /* 4 byte size + 8 byte timestamp */
240
static int read_frame(FILE *f, vpx_image_t *img, unsigned int file_type,
241
                      y4m_input *y4m, struct detect_buffer *detect)
John Koleszar's avatar
John Koleszar committed
242 243 244
{
    int plane = 0;

245
    if (file_type == FILE_TYPE_Y4M)
John Koleszar's avatar
John Koleszar committed
246
    {
247 248
        if (y4m_input_fetch_frame(y4m, f, img) < 0)
           return 0;
John Koleszar's avatar
John Koleszar committed
249
    }
250
    else
John Koleszar's avatar
John Koleszar committed
251
    {
252
        if (file_type == FILE_TYPE_IVF)
John Koleszar's avatar
John Koleszar committed
253
        {
254 255 256 257 258 259 260
            char junk[IVF_FRAME_HDR_SZ];

            /* Skip the frame header. We know how big the frame should be. See
             * write_ivf_frame_header() for documentation on the frame header
             * layout.
             */
            fread(junk, 1, IVF_FRAME_HDR_SZ, f);
John Koleszar's avatar
John Koleszar committed
261 262
        }

263
        for (plane = 0; plane < 3; plane++)
John Koleszar's avatar
John Koleszar committed
264
        {
265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287
            unsigned char *ptr;
            int w = (plane ? (1 + img->d_w) / 2 : img->d_w);
            int h = (plane ? (1 + img->d_h) / 2 : img->d_h);
            int r;

            /* Determine the correct plane based on the image format. The for-loop
             * always counts in Y,U,V order, but this may not match the order of
             * the data on disk.
             */
            switch (plane)
            {
            case 1:
                ptr = img->planes[img->fmt==VPX_IMG_FMT_YV12? VPX_PLANE_V : VPX_PLANE_U];
                break;
            case 2:
                ptr = img->planes[img->fmt==VPX_IMG_FMT_YV12?VPX_PLANE_U : VPX_PLANE_V];
                break;
            default:
                ptr = img->planes[plane];
            }

            for (r = 0; r < h; r++)
            {
288 289 290 291 292 293 294 295 296
                if (detect->valid)
                {
                    memcpy(ptr, detect->buf, 4);
                    fread(ptr+4, 1, w-4, f);
                    detect->valid = 0;
                }
                else
                    fread(ptr, 1, w, f);

297 298
                ptr += img->stride[plane];
            }
John Koleszar's avatar
John Koleszar committed
299 300 301 302 303 304 305
        }
    }

    return !feof(f);
}


306
unsigned int file_is_y4m(FILE      *infile,
307 308
                         y4m_input *y4m,
                         char       detect[4])
309
{
310
    if(memcmp(detect, "YUV4", 4) == 0)
311 312 313 314 315 316
    {
        return 1;
    }
    return 0;
}

John Koleszar's avatar
John Koleszar committed
317 318 319 320
#define IVF_FILE_HDR_SZ (32)
unsigned int file_is_ivf(FILE *infile,
                         unsigned int *fourcc,
                         unsigned int *width,
321 322
                         unsigned int *height,
                         char          detect[4])
John Koleszar's avatar
John Koleszar committed
323 324 325 326
{
    char raw_hdr[IVF_FILE_HDR_SZ];
    int is_ivf = 0;

327 328 329
    if(memcmp(detect, "DKIF", 4) != 0)
        return 0;

John Koleszar's avatar
John Koleszar committed
330 331 332
    /* See write_ivf_file_header() for more documentation on the file header
     * layout.
     */
333 334
    if (fread(raw_hdr + 4, 1, IVF_FILE_HDR_SZ - 4, infile)
        == IVF_FILE_HDR_SZ - 4)
John Koleszar's avatar
John Koleszar committed
335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511
    {
        {
            is_ivf = 1;

            if (mem_get_le16(raw_hdr + 4) != 0)
                fprintf(stderr, "Error: Unrecognized IVF version! This file may not"
                        " decode properly.");

            *fourcc = mem_get_le32(raw_hdr + 8);
        }
    }

    if (is_ivf)
    {
        *width = mem_get_le16(raw_hdr + 12);
        *height = mem_get_le16(raw_hdr + 14);
    }

    return is_ivf;
}


static void write_ivf_file_header(FILE *outfile,
                                  const vpx_codec_enc_cfg_t *cfg,
                                  unsigned int fourcc,
                                  int frame_cnt)
{
    char header[32];

    if (cfg->g_pass != VPX_RC_ONE_PASS && cfg->g_pass != VPX_RC_LAST_PASS)
        return;

    header[0] = 'D';
    header[1] = 'K';
    header[2] = 'I';
    header[3] = 'F';
    mem_put_le16(header + 4,  0);                 /* version */
    mem_put_le16(header + 6,  32);                /* headersize */
    mem_put_le32(header + 8,  fourcc);            /* headersize */
    mem_put_le16(header + 12, cfg->g_w);          /* width */
    mem_put_le16(header + 14, cfg->g_h);          /* height */
    mem_put_le32(header + 16, cfg->g_timebase.den); /* rate */
    mem_put_le32(header + 20, cfg->g_timebase.num); /* scale */
    mem_put_le32(header + 24, frame_cnt);         /* length */
    mem_put_le32(header + 28, 0);                 /* unused */

    fwrite(header, 1, 32, outfile);
}


static void write_ivf_frame_header(FILE *outfile,
                                   const vpx_codec_cx_pkt_t *pkt)
{
    char             header[12];
    vpx_codec_pts_t  pts;

    if (pkt->kind != VPX_CODEC_CX_FRAME_PKT)
        return;

    pts = pkt->data.frame.pts;
    mem_put_le32(header, pkt->data.frame.sz);
    mem_put_le32(header + 4, pts & 0xFFFFFFFF);
    mem_put_le32(header + 8, pts >> 32);

    fwrite(header, 1, 12, outfile);
}

#include "args.h"

static const arg_def_t use_yv12 = ARG_DEF(NULL, "yv12", 0,
                                  "Input file is YV12 ");
static const arg_def_t use_i420 = ARG_DEF(NULL, "i420", 0,
                                  "Input file is I420 (default)");
static const arg_def_t codecarg = ARG_DEF(NULL, "codec", 1,
                                  "Codec to use");
static const arg_def_t passes           = ARG_DEF("p", "passes", 1,
        "Number of passes (1/2)");
static const arg_def_t pass_arg         = ARG_DEF(NULL, "pass", 1,
        "Pass to execute (1/2)");
static const arg_def_t fpf_name         = ARG_DEF(NULL, "fpf", 1,
        "First pass statistics file name");
static const arg_def_t limit = ARG_DEF(NULL, "limit", 1,
                                       "Stop encoding after n input frames");
static const arg_def_t deadline         = ARG_DEF("d", "deadline", 1,
        "Deadline per frame (usec)");
static const arg_def_t best_dl          = ARG_DEF(NULL, "best", 0,
        "Use Best Quality Deadline");
static const arg_def_t good_dl          = ARG_DEF(NULL, "good", 0,
        "Use Good Quality Deadline");
static const arg_def_t rt_dl            = ARG_DEF(NULL, "rt", 0,
        "Use Realtime Quality Deadline");
static const arg_def_t verbosearg       = ARG_DEF("v", "verbose", 0,
        "Show encoder parameters");
static const arg_def_t psnrarg          = ARG_DEF(NULL, "psnr", 0,
        "Show PSNR in status line");
static const arg_def_t *main_args[] =
{
    &codecarg, &passes, &pass_arg, &fpf_name, &limit, &deadline, &best_dl, &good_dl, &rt_dl,
    &verbosearg, &psnrarg,
    NULL
};

static const arg_def_t usage            = ARG_DEF("u", "usage", 1,
        "Usage profile number to use");
static const arg_def_t threads          = ARG_DEF("t", "threads", 1,
        "Max number of threads to use");
static const arg_def_t profile          = ARG_DEF(NULL, "profile", 1,
        "Bitstream profile number to use");
static const arg_def_t width            = ARG_DEF("w", "width", 1,
        "Frame width");
static const arg_def_t height           = ARG_DEF("h", "height", 1,
        "Frame height");
static const arg_def_t timebase         = ARG_DEF(NULL, "timebase", 1,
        "Stream timebase (frame duration)");
static const arg_def_t error_resilient  = ARG_DEF(NULL, "error-resilient", 1,
        "Enable error resiliency features");
static const arg_def_t lag_in_frames    = ARG_DEF(NULL, "lag-in-frames", 1,
        "Max number of frames to lag");

static const arg_def_t *global_args[] =
{
    &use_yv12, &use_i420, &usage, &threads, &profile,
    &width, &height, &timebase, &error_resilient,
    &lag_in_frames, NULL
};

static const arg_def_t dropframe_thresh   = ARG_DEF(NULL, "drop-frame", 1,
        "Temporal resampling threshold (buf %)");
static const arg_def_t resize_allowed     = ARG_DEF(NULL, "resize-allowed", 1,
        "Spatial resampling enabled (bool)");
static const arg_def_t resize_up_thresh   = ARG_DEF(NULL, "resize-up", 1,
        "Upscale threshold (buf %)");
static const arg_def_t resize_down_thresh = ARG_DEF(NULL, "resize-down", 1,
        "Downscale threshold (buf %)");
static const arg_def_t end_usage          = ARG_DEF(NULL, "end-usage", 1,
        "VBR=0 | CBR=1");
static const arg_def_t target_bitrate     = ARG_DEF(NULL, "target-bitrate", 1,
        "Bitrate (kbps)");
static const arg_def_t min_quantizer      = ARG_DEF(NULL, "min-q", 1,
        "Minimum (best) quantizer");
static const arg_def_t max_quantizer      = ARG_DEF(NULL, "max-q", 1,
        "Maximum (worst) quantizer");
static const arg_def_t undershoot_pct     = ARG_DEF(NULL, "undershoot-pct", 1,
        "Datarate undershoot (min) target (%)");
static const arg_def_t overshoot_pct      = ARG_DEF(NULL, "overshoot-pct", 1,
        "Datarate overshoot (max) target (%)");
static const arg_def_t buf_sz             = ARG_DEF(NULL, "buf-sz", 1,
        "Client buffer size (ms)");
static const arg_def_t buf_initial_sz     = ARG_DEF(NULL, "buf-initial-sz", 1,
        "Client initial buffer size (ms)");
static const arg_def_t buf_optimal_sz     = ARG_DEF(NULL, "buf-optimal-sz", 1,
        "Client optimal buffer size (ms)");
static const arg_def_t *rc_args[] =
{
    &dropframe_thresh, &resize_allowed, &resize_up_thresh, &resize_down_thresh,
    &end_usage, &target_bitrate, &min_quantizer, &max_quantizer,
    &undershoot_pct, &overshoot_pct, &buf_sz, &buf_initial_sz, &buf_optimal_sz,
    NULL
};


static const arg_def_t bias_pct = ARG_DEF(NULL, "bias-pct", 1,
                                  "CBR/VBR bias (0=CBR, 100=VBR)");
static const arg_def_t minsection_pct = ARG_DEF(NULL, "minsection-pct", 1,
                                        "GOP min bitrate (% of target)");
static const arg_def_t maxsection_pct = ARG_DEF(NULL, "maxsection-pct", 1,
                                        "GOP max bitrate (% of target)");
static const arg_def_t *rc_twopass_args[] =
{
    &bias_pct, &minsection_pct, &maxsection_pct, NULL
};


static const arg_def_t kf_min_dist = ARG_DEF(NULL, "kf-min-dist", 1,
                                     "Minimum keyframe interval (frames)");
static const arg_def_t kf_max_dist = ARG_DEF(NULL, "kf-max-dist", 1,
                                     "Maximum keyframe interval (frames)");
512 513
static const arg_def_t kf_disabled = ARG_DEF(NULL, "disable-kf", 0,
                                     "Disable keyframe placement");
John Koleszar's avatar
John Koleszar committed
514 515
static const arg_def_t *kf_args[] =
{
516
    &kf_min_dist, &kf_max_dist, &kf_disabled, NULL
John Koleszar's avatar
John Koleszar committed
517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566
};


#if CONFIG_VP8_ENCODER
static const arg_def_t noise_sens = ARG_DEF(NULL, "noise-sensitivity", 1,
                                    "Noise sensitivity (frames to blur)");
static const arg_def_t sharpness = ARG_DEF(NULL, "sharpness", 1,
                                   "Filter sharpness (0-7)");
static const arg_def_t static_thresh = ARG_DEF(NULL, "static-thresh", 1,
                                       "Motion detection threshold");
#endif

#if CONFIG_VP8_ENCODER
static const arg_def_t cpu_used = ARG_DEF(NULL, "cpu-used", 1,
                                  "CPU Used (-16..16)");
#endif


#if CONFIG_VP8_ENCODER
static const arg_def_t token_parts = ARG_DEF(NULL, "token-parts", 1,
                                     "Number of token partitions to use, log2");
static const arg_def_t auto_altref = ARG_DEF(NULL, "auto-alt-ref", 1,
                                     "Enable automatic alt reference frames");
static const arg_def_t arnr_maxframes = ARG_DEF(NULL, "arnr-maxframes", 1,
                                        "alt_ref Max Frames");
static const arg_def_t arnr_strength = ARG_DEF(NULL, "arnr-strength", 1,
                                       "alt_ref Strength");
static const arg_def_t arnr_type = ARG_DEF(NULL, "arnr-type", 1,
                                   "alt_ref Type");

static const arg_def_t *vp8_args[] =
{
    &cpu_used, &auto_altref, &noise_sens, &sharpness, &static_thresh,
    &token_parts, &arnr_maxframes, &arnr_strength, &arnr_type, NULL
};
static const int vp8_arg_ctrl_map[] =
{
    VP8E_SET_CPUUSED, VP8E_SET_ENABLEAUTOALTREF,
    VP8E_SET_NOISE_SENSITIVITY, VP8E_SET_SHARPNESS, VP8E_SET_STATIC_THRESHOLD,
    VP8E_SET_TOKEN_PARTITIONS,
    VP8E_SET_ARNR_MAXFRAMES, VP8E_SET_ARNR_STRENGTH , VP8E_SET_ARNR_TYPE, 0
};
#endif

static const arg_def_t *no_args[] = { NULL };

static void usage_exit()
{
    int i;

567
    fprintf(stderr, "Usage: %s <options> src_filename dst_filename\n", exec_name);
John Koleszar's avatar
John Koleszar committed
568

569
    fprintf(stderr, "\n_options:\n");
John Koleszar's avatar
John Koleszar committed
570
    arg_show_usage(stdout, main_args);
571
    fprintf(stderr, "\n_encoder Global Options:\n");
John Koleszar's avatar
John Koleszar committed
572
    arg_show_usage(stdout, global_args);
573
    fprintf(stderr, "\n_rate Control Options:\n");
John Koleszar's avatar
John Koleszar committed
574
    arg_show_usage(stdout, rc_args);
575
    fprintf(stderr, "\n_twopass Rate Control Options:\n");
John Koleszar's avatar
John Koleszar committed
576
    arg_show_usage(stdout, rc_twopass_args);
577
    fprintf(stderr, "\n_keyframe Placement Options:\n");
John Koleszar's avatar
John Koleszar committed
578 579
    arg_show_usage(stdout, kf_args);
#if CONFIG_VP8_ENCODER
580
    fprintf(stderr, "\n_vp8 Specific Options:\n");
John Koleszar's avatar
John Koleszar committed
581 582
    arg_show_usage(stdout, vp8_args);
#endif
583
    fprintf(stderr, "\n"
John Koleszar's avatar
John Koleszar committed
584 585 586 587
           "Included encoders:\n"
           "\n");

    for (i = 0; i < sizeof(codecs) / sizeof(codecs[0]); i++)
588
        fprintf(stderr, "    %-6s - %s\n",
John Koleszar's avatar
John Koleszar committed
589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620
               codecs[i].name,
               vpx_codec_iface_name(codecs[i].iface));

    exit(EXIT_FAILURE);
}

#define ARG_CTRL_CNT_MAX 10


int main(int argc, const char **argv_)
{
    vpx_codec_ctx_t        encoder;
    const char                  *in_fn = NULL, *out_fn = NULL, *stats_fn = NULL;
    int                    i;
    FILE                  *infile, *outfile;
    vpx_codec_enc_cfg_t    cfg;
    vpx_codec_err_t        res;
    int                    pass, one_pass_only = 0;
    stats_io_t             stats;
    vpx_image_t            raw;
    const struct codec_item  *codec = codecs;
    int                    frame_avail, got_data;

    struct arg               arg;
    char                   **argv, **argi, **argj;
    int                      arg_usage = 0, arg_passes = 1, arg_deadline = 0;
    int                      arg_ctrls[ARG_CTRL_CNT_MAX][2], arg_ctrl_cnt = 0;
    int                      arg_limit = 0;
    static const arg_def_t **ctrl_args = no_args;
    static const int        *ctrl_args_map = NULL;
    int                      verbose = 0, show_psnr = 0;
    int                      arg_use_i420 = 1;
621
    int                      arg_have_timebase = 0;
John Koleszar's avatar
John Koleszar committed
622
    unsigned long            cx_time = 0;
623 624
    unsigned int             file_type, fourcc;
    y4m_input                y4m;
John Koleszar's avatar
John Koleszar committed
625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707

    exec_name = argv_[0];

    if (argc < 3)
        usage_exit();


    /* First parse the codec and usage values, because we want to apply other
     * parameters on top of the default configuration provided by the codec.
     */
    argv = argv_dup(argc - 1, argv_ + 1);

    for (argi = argj = argv; (*argj = *argi); argi += arg.argv_step)
    {
        arg.argv_step = 1;

        if (arg_match(&arg, &codecarg, argi))
        {
            int j, k = -1;

            for (j = 0; j < sizeof(codecs) / sizeof(codecs[0]); j++)
                if (!strcmp(codecs[j].name, arg.val))
                    k = j;

            if (k >= 0)
                codec = codecs + k;
            else
                die("Error: Unrecognized argument (%s) to --codec\n",
                    arg.val);

        }
        else if (arg_match(&arg, &passes, argi))
        {
            arg_passes = arg_parse_uint(&arg);

            if (arg_passes < 1 || arg_passes > 2)
                die("Error: Invalid number of passes (%d)\n", arg_passes);
        }
        else if (arg_match(&arg, &pass_arg, argi))
        {
            one_pass_only = arg_parse_uint(&arg);

            if (one_pass_only < 1 || one_pass_only > 2)
                die("Error: Invalid pass selected (%d)\n", one_pass_only);
        }
        else if (arg_match(&arg, &fpf_name, argi))
            stats_fn = arg.val;
        else if (arg_match(&arg, &usage, argi))
            arg_usage = arg_parse_uint(&arg);
        else if (arg_match(&arg, &deadline, argi))
            arg_deadline = arg_parse_uint(&arg);
        else if (arg_match(&arg, &best_dl, argi))
            arg_deadline = VPX_DL_BEST_QUALITY;
        else if (arg_match(&arg, &good_dl, argi))
            arg_deadline = VPX_DL_GOOD_QUALITY;
        else if (arg_match(&arg, &rt_dl, argi))
            arg_deadline = VPX_DL_REALTIME;
        else if (arg_match(&arg, &use_yv12, argi))
        {
            arg_use_i420 = 0;
        }
        else if (arg_match(&arg, &use_i420, argi))
        {
            arg_use_i420 = 1;
        }
        else if (arg_match(&arg, &verbosearg, argi))
            verbose = 1;
        else if (arg_match(&arg, &limit, argi))
            arg_limit = arg_parse_uint(&arg);
        else if (arg_match(&arg, &psnrarg, argi))
            show_psnr = 1;
        else
            argj++;
    }

    /* Ensure that --passes and --pass are consistent. If --pass is set and --passes=2,
     * ensure --fpf was set.
     */
    if (one_pass_only)
    {
        /* DWIM: Assume the user meant passes=2 if pass=2 is specified */
        if (one_pass_only > arg_passes)
        {
708
            fprintf(stderr, "Warning: Assuming --pass=%d implies --passes=%d\n",
John Koleszar's avatar
John Koleszar committed
709 710 711 712 713 714 715 716 717 718 719 720 721
                   one_pass_only, one_pass_only);
            arg_passes = one_pass_only;
        }

        if (arg_passes == 2 && !stats_fn)
            die("Must specify --fpf when --pass=%d and --passes=2\n", one_pass_only);
    }

    /* Populate encoder configuration */
    res = vpx_codec_enc_config_default(codec->iface, &cfg, arg_usage);

    if (res)
    {
722 723
        fprintf(stderr, "Failed to get config: %s\n",
                vpx_codec_err_to_string(res));
John Koleszar's avatar
John Koleszar committed
724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741
        return EXIT_FAILURE;
    }

    /* Now parse the remainder of the parameters. */
    for (argi = argj = argv; (*argj = *argi); argi += arg.argv_step)
    {
        arg.argv_step = 1;

        if (0);
        else if (arg_match(&arg, &threads, argi))
            cfg.g_threads = arg_parse_uint(&arg);
        else if (arg_match(&arg, &profile, argi))
            cfg.g_profile = arg_parse_uint(&arg);
        else if (arg_match(&arg, &width, argi))
            cfg.g_w = arg_parse_uint(&arg);
        else if (arg_match(&arg, &height, argi))
            cfg.g_h = arg_parse_uint(&arg);
        else if (arg_match(&arg, &timebase, argi))
742
        {
John Koleszar's avatar
John Koleszar committed
743
            cfg.g_timebase = arg_parse_rational(&arg);
744 745
            arg_have_timebase = 1;
        }
John Koleszar's avatar
John Koleszar committed
746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782
        else if (arg_match(&arg, &error_resilient, argi))
            cfg.g_error_resilient = arg_parse_uint(&arg);
        else if (arg_match(&arg, &lag_in_frames, argi))
            cfg.g_lag_in_frames = arg_parse_uint(&arg);
        else if (arg_match(&arg, &dropframe_thresh, argi))
            cfg.rc_dropframe_thresh = arg_parse_uint(&arg);
        else if (arg_match(&arg, &resize_allowed, argi))
            cfg.rc_resize_allowed = arg_parse_uint(&arg);
        else if (arg_match(&arg, &resize_up_thresh, argi))
            cfg.rc_resize_up_thresh = arg_parse_uint(&arg);
        else if (arg_match(&arg, &resize_down_thresh, argi))
            cfg.rc_resize_down_thresh = arg_parse_uint(&arg);
        else if (arg_match(&arg, &resize_down_thresh, argi))
            cfg.rc_resize_down_thresh = arg_parse_uint(&arg);
        else if (arg_match(&arg, &end_usage, argi))
            cfg.rc_end_usage = arg_parse_uint(&arg);
        else if (arg_match(&arg, &target_bitrate, argi))
            cfg.rc_target_bitrate = arg_parse_uint(&arg);
        else if (arg_match(&arg, &min_quantizer, argi))
            cfg.rc_min_quantizer = arg_parse_uint(&arg);
        else if (arg_match(&arg, &max_quantizer, argi))
            cfg.rc_max_quantizer = arg_parse_uint(&arg);
        else if (arg_match(&arg, &undershoot_pct, argi))
            cfg.rc_undershoot_pct = arg_parse_uint(&arg);
        else if (arg_match(&arg, &overshoot_pct, argi))
            cfg.rc_overshoot_pct = arg_parse_uint(&arg);
        else if (arg_match(&arg, &buf_sz, argi))
            cfg.rc_buf_sz = arg_parse_uint(&arg);
        else if (arg_match(&arg, &buf_initial_sz, argi))
            cfg.rc_buf_initial_sz = arg_parse_uint(&arg);
        else if (arg_match(&arg, &buf_optimal_sz, argi))
            cfg.rc_buf_optimal_sz = arg_parse_uint(&arg);
        else if (arg_match(&arg, &bias_pct, argi))
        {
            cfg.rc_2pass_vbr_bias_pct = arg_parse_uint(&arg);

            if (arg_passes < 2)
783 784 785
                fprintf(stderr,
                        "Warning: option %s ignored in one-pass mode.\n",
                        arg.name);
John Koleszar's avatar
John Koleszar committed
786 787 788 789 790 791
        }
        else if (arg_match(&arg, &minsection_pct, argi))
        {
            cfg.rc_2pass_vbr_minsection_pct = arg_parse_uint(&arg);

            if (arg_passes < 2)
792 793 794
                fprintf(stderr,
                        "Warning: option %s ignored in one-pass mode.\n",
                        arg.name);
John Koleszar's avatar
John Koleszar committed
795 796 797 798 799 800
        }
        else if (arg_match(&arg, &maxsection_pct, argi))
        {
            cfg.rc_2pass_vbr_maxsection_pct = arg_parse_uint(&arg);

            if (arg_passes < 2)
801 802 803
                fprintf(stderr,
                        "Warning: option %s ignored in one-pass mode.\n",
                        arg.name);
John Koleszar's avatar
John Koleszar committed
804 805 806 807 808
        }
        else if (arg_match(&arg, &kf_min_dist, argi))
            cfg.kf_min_dist = arg_parse_uint(&arg);
        else if (arg_match(&arg, &kf_max_dist, argi))
            cfg.kf_max_dist = arg_parse_uint(&arg);
809 810
        else if (arg_match(&arg, &kf_disabled, argi))
            cfg.kf_mode = VPX_KF_DISABLED;
John Koleszar's avatar
John Koleszar committed
811 812 813 814 815 816 817
        else
            argj++;
    }

    /* Handle codec specific options */
#if CONFIG_VP8_ENCODER

818 819 820 821 822
    if (codec->iface == &vpx_codec_vp8_cx_algo
#if CONFIG_EXPERIMENTAL
        || codec->iface == &vpx_codec_vp8x_cx_algo
#endif
        )
John Koleszar's avatar
John Koleszar committed
823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856
    {
        ctrl_args = vp8_args;
        ctrl_args_map = vp8_arg_ctrl_map;
    }

#endif

    for (argi = argj = argv; (*argj = *argi); argi += arg.argv_step)
    {
        int match = 0;

        arg.argv_step = 1;

        for (i = 0; ctrl_args[i]; i++)
        {
            if (arg_match(&arg, ctrl_args[i], argi))
            {
                match = 1;

                if (arg_ctrl_cnt < ARG_CTRL_CNT_MAX)
                {
                    arg_ctrls[arg_ctrl_cnt][0] = ctrl_args_map[i];
                    arg_ctrls[arg_ctrl_cnt][1] = arg_parse_int(&arg);
                    arg_ctrl_cnt++;
                }
            }
        }

        if (!match)
            argj++;
    }

    /* Check for unrecognized options */
    for (argi = argv; *argi; argi++)
857
        if (argi[0][0] == '-' && argi[0][1])
John Koleszar's avatar
John Koleszar committed
858 859 860 861 862 863 864 865 866 867 868 869 870 871 872
            die("Error: Unrecognized option %s\n", *argi);

    /* Handle non-option arguments */
    in_fn = argv[0];
    out_fn = argv[1];

    if (!in_fn || !out_fn)
        usage_exit();

    memset(&stats, 0, sizeof(stats));

    for (pass = one_pass_only ? one_pass_only - 1 : 0; pass < arg_passes; pass++)
    {
        int frames_in = 0, frames_out = 0;
        unsigned long nbytes = 0;
873
        struct detect_buffer detect;
John Koleszar's avatar
John Koleszar committed
874

875 876
        /* Parse certain options from the input file, if possible */
        infile = strcmp(in_fn, "-") ? fopen(in_fn, "rb") : stdin;
John Koleszar's avatar
John Koleszar committed
877 878 879

        if (!infile)
        {
880
            fprintf(stderr, "Failed to open input file\n");
John Koleszar's avatar
John Koleszar committed
881 882 883
            return EXIT_FAILURE;
        }

884 885 886 887 888
        fread(detect.buf, 1, 4, infile);
        detect.valid = 0;

        if (file_is_y4m(infile, &y4m, detect.buf))
        {
889 890 891 892 893 894 895 896 897 898 899 900
            if (y4m_input_open(&y4m, infile, detect.buf, 4) >= 0)
            {
                file_type = FILE_TYPE_Y4M;
                cfg.g_w = y4m.pic_w;
                cfg.g_h = y4m.pic_h;
                /* Use the frame rate from the file only if none was specified
                 * on the command-line.
                 */
                if (!arg_have_timebase)
                {
                    cfg.g_timebase.num = y4m.fps_d;
                    cfg.g_timebase.den = y4m.fps_n;
901 902
                    /* And don't reset it in the second pass.*/
                    arg_have_timebase = 1;
903 904 905 906
                }
                arg_use_i420 = 0;
            }
            else
907
            {
908 909
                fprintf(stderr, "Unsupported Y4M stream.\n");
                return EXIT_FAILURE;
910 911 912
            }
        }
        else if (file_is_ivf(infile, &fourcc, &cfg.g_w, &cfg.g_h, detect.buf))
913
        {
914 915 916 917 918 919 920 921 922 923 924 925 926
            file_type = FILE_TYPE_IVF;
            switch (fourcc)
            {
            case 0x32315659:
                arg_use_i420 = 0;
                break;
            case 0x30323449:
                arg_use_i420 = 1;
                break;
            default:
                fprintf(stderr, "Unsupported fourcc (%08x) in IVF\n", fourcc);
                return EXIT_FAILURE;
            }
927
        }
928 929 930 931 932 933 934 935
        else
        {
            file_type = FILE_TYPE_RAW;
            detect.valid = 1;
        }
#define SHOW(field) fprintf(stderr, "    %-28s = %d\n", #field, cfg.field)

        if (verbose && pass == 0)
936
        {
937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973
            fprintf(stderr, "Codec: %s\n", vpx_codec_iface_name(codec->iface));
            fprintf(stderr, "Source file: %s Format: %s\n", in_fn,
                    arg_use_i420 ? "I420" : "YV12");
            fprintf(stderr, "Destination file: %s\n", out_fn);
            fprintf(stderr, "Encoder parameters:\n");

            SHOW(g_usage);
            SHOW(g_threads);
            SHOW(g_profile);
            SHOW(g_w);
            SHOW(g_h);
            SHOW(g_timebase.num);
            SHOW(g_timebase.den);
            SHOW(g_error_resilient);
            SHOW(g_pass);
            SHOW(g_lag_in_frames);
            SHOW(rc_dropframe_thresh);
            SHOW(rc_resize_allowed);
            SHOW(rc_resize_up_thresh);
            SHOW(rc_resize_down_thresh);
            SHOW(rc_end_usage);
            SHOW(rc_target_bitrate);
            SHOW(rc_min_quantizer);
            SHOW(rc_max_quantizer);
            SHOW(rc_undershoot_pct);
            SHOW(rc_overshoot_pct);
            SHOW(rc_buf_sz);
            SHOW(rc_buf_initial_sz);
            SHOW(rc_buf_optimal_sz);
            SHOW(rc_2pass_vbr_bias_pct);
            SHOW(rc_2pass_vbr_minsection_pct);
            SHOW(rc_2pass_vbr_maxsection_pct);
            SHOW(kf_mode);
            SHOW(kf_min_dist);
            SHOW(kf_max_dist);
        }

974
        if(pass == (one_pass_only ? one_pass_only - 1 : 0)) {
975 976 977 978 979 980 981 982 983 984 985 986 987 988 989
            if (file_type == FILE_TYPE_Y4M)
                /*The Y4M reader does its own allocation.
                  Just initialize this here to avoid problems if we never read any
                   frames.*/
                memset(&raw, 0, sizeof(raw));
            else
                vpx_img_alloc(&raw, arg_use_i420 ? VPX_IMG_FMT_I420 : VPX_IMG_FMT_YV12,
                              cfg.g_w, cfg.g_h, 1);

            // This was added so that ivfenc will create monotically increasing
            // timestamps.  Since we create new timestamps for alt-reference frames
            // we need to make room in the series of timestamps.  Since there can
            // only be 1 alt-ref frame ( current bitstream) multiplying by 2
            // gives us enough room.
            cfg.g_timebase.den *= 2;
990 991
        }

992
        outfile = strcmp(out_fn, "-") ? fopen(out_fn, "wb") : stdout;
John Koleszar's avatar
John Koleszar committed
993 994 995

        if (!outfile)
        {
996
            fprintf(stderr, "Failed to open output file\n");
John Koleszar's avatar
John Koleszar committed
997 998 999 1000 1001 1002 1003
            return EXIT_FAILURE;
        }

        if (stats_fn)
        {
            if (!stats_open_file(&stats, stats_fn, pass))
            {
1004
                fprintf(stderr, "Failed to open statistics store\n");
John Koleszar's avatar
John Koleszar committed
1005 1006 1007 1008 1009 1010 1011
                return EXIT_FAILURE;
            }
        }
        else
        {
            if (!stats_open_mem(&stats, pass))
            {
1012
                fprintf(stderr, "Failed to open statistics store\n");
John Koleszar's avatar
John Koleszar committed
1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043
                return EXIT_FAILURE;
            }
        }

        cfg.g_pass = arg_passes == 2
                     ? pass ? VPX_RC_LAST_PASS : VPX_RC_FIRST_PASS
                 : VPX_RC_ONE_PASS;
#if VPX_ENCODER_ABI_VERSION > (1 + VPX_CODEC_ABI_VERSION)

        if (pass)
        {
            cfg.rc_twopass_stats_in = stats_get(&stats);
        }

#endif

        write_ivf_file_header(outfile, &cfg, codec->fourcc, 0);


        /* Construct Encoder Context */
        vpx_codec_enc_init(&encoder, codec->iface, &cfg,
                           show_psnr ? VPX_CODEC_USE_PSNR : 0);
        ctx_exit_on_error(&encoder, "Failed to initialize encoder");

        /* Note that we bypass the vpx_codec_control wrapper macro because
         * we're being clever to store the control IDs in an array. Real
         * applications will want to make use of the enumerations directly
         */
        for (i = 0; i < arg_ctrl_cnt; i++)
        {
            if (vpx_codec_control_(&encoder, arg_ctrls[i][0], arg_ctrls[i][1]))
1044 1045
                fprintf(stderr, "Error: Tried to set control %d = %d\n",
                        arg_ctrls[i][0], arg_ctrls[i][1]);
John Koleszar's avatar
John Koleszar committed
1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060

            ctx_exit_on_error(&encoder, "Failed to control codec");
        }

        frame_avail = 1;
        got_data = 0;

        while (frame_avail || got_data)
        {
            vpx_codec_iter_t iter = NULL;
            const vpx_codec_cx_pkt_t *pkt;
            struct vpx_usec_timer timer;

            if (!arg_limit || frames_in < arg_limit)
            {
1061 1062
                frame_avail = read_frame(infile, &raw, file_type, &y4m,
                                         &detect);
John Koleszar's avatar
John Koleszar committed
1063 1064 1065 1066

                if (frame_avail)
                    frames_in++;

1067 1068 1069
                fprintf(stderr,
                        "\rPass %d/%d frame %4d/%-4d %7ldB \033[K", pass + 1,
                        arg_passes, frames_in, frames_out, nbytes);
John Koleszar's avatar
John Koleszar committed
1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092
            }
            else
                frame_avail = 0;

            vpx_usec_timer_start(&timer);

            // since we halved our timebase we need to double the timestamps
            // and duration we pass in.
            vpx_codec_encode(&encoder, frame_avail ? &raw : NULL, (frames_in - 1) * 2,
                             2, 0, arg_deadline);
            vpx_usec_timer_mark(&timer);
            cx_time += vpx_usec_timer_elapsed(&timer);
            ctx_exit_on_error(&encoder, "Failed to encode frame");
            got_data = 0;

            while ((pkt = vpx_codec_get_cx_data(&encoder, &iter)))
            {
                got_data = 1;

                switch (pkt->kind)
                {
                case VPX_CODEC_CX_FRAME_PKT:
                    frames_out++;
1093 1094
                    fprintf(stderr, " %6luF",
                            (unsigned long)pkt->data.frame.sz);
John Koleszar's avatar
John Koleszar committed
1095 1096
                    write_ivf_frame_header(outfile, pkt);
                    fwrite(pkt->data.frame.buf, 1, pkt->data.frame.sz, outfile);
1097
                    nbytes += pkt->data.raw.sz;
John Koleszar's avatar
John Koleszar committed
1098 1099 1100
                    break;
                case VPX_CODEC_STATS_PKT:
                    frames_out++;
1101
                    fprintf(stderr, " %6luS",
John Koleszar's avatar
John Koleszar committed
1102 1103 1104 1105
                           (unsigned long)pkt->data.twopass_stats.sz);
                    stats_write(&stats,
                                pkt->data.twopass_stats.buf,
                                pkt->data.twopass_stats.sz);
1106
                    nbytes += pkt->data.raw.sz;
John Koleszar's avatar
John Koleszar committed
1107 1108 1109 1110 1111 1112 1113 1114
                    break;
                case VPX_CODEC_PSNR_PKT:

                    if (show_psnr)
                    {
                        int i;

                        for (i = 0; i < 4; i++)
1115
                            fprintf(stderr, "%.3lf ", pkt->data.psnr.psnr[i]);
John Koleszar's avatar
John Koleszar committed
1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129
                    }

                    break;
                default:
                    break;
                }
            }

            fflush(stdout);
        }

        /* this bitrate calc is simplified and relies on the fact that this
         * application uses 1/timebase for framerate.
         */
1130 1131
        fprintf(stderr,
               "\rPass %d/%d frame %4d/%-4d %7ldB %7ldb/f %7"PRId64"b/s"
John Koleszar's avatar
John Koleszar committed
1132 1133
               " %7lu %s (%.2f fps)\033[K", pass + 1,
               arg_passes, frames_in, frames_out, nbytes, nbytes * 8 / frames_in,
1134
               nbytes * 8 *(int64_t)cfg.g_timebase.den/2/ cfg.g_timebase.num / frames_in,
John Koleszar's avatar
John Koleszar committed
1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147
               cx_time > 9999999 ? cx_time / 1000 : cx_time,
               cx_time > 9999999 ? "ms" : "us",
               (float)frames_in * 1000000.0 / (float)cx_time);

        vpx_codec_destroy(&encoder);

        fclose(infile);

        if (!fseek(outfile, 0, SEEK_SET))
            write_ivf_file_header(outfile, &cfg, codec->fourcc, frames_out);

        fclose(outfile);
        stats_close(&stats);
1148
        fprintf(stderr, "\n");
John Koleszar's avatar
John Koleszar committed
1149 1150 1151 1152 1153 1154 1155 1156 1157

        if (one_pass_only)
            break;
    }

    vpx_img_free(&raw);
    free(argv);
    return EXIT_SUCCESS;
}