libavformat/chromaprint.c

   1 /*
   2  * Chromaprint fingerprinting muxer
   3  * Copyright (c) 2015 rcombs
   4  *
   5  * This file is part of FFmpeg.
   6  *
   7  * FFmpeg is free software; you can redistribute it and/or
   8  * modify it under the terms of the GNU Lesser General Public
   9  * 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  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  14  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  15  * Lesser General Public License for more details.
  16  *
  17  * 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 "avformat.h"
  23 #include "mux.h"
  24 #include "libavutil/opt.h"
  25 #include "libavutil/thread.h"
  26 #include <chromaprint.h>
  27
  28 #define CPR_VERSION_INT AV_VERSION_INT(CHROMAPRINT_VERSION_MAJOR, \
  29                                        CHROMAPRINT_VERSION_MINOR, \
  30                                        CHROMAPRINT_VERSION_PATCH)
  31
  32 static AVMutex chromaprint_mutex = AV_MUTEX_INITIALIZER;
  33
  34 typedef enum FingerprintFormat {
  35     FINGERPRINT_RAW,
  36     FINGERPRINT_COMPRESSED,
  37     FINGERPRINT_BASE64,
  38 } FingerprintFormat;
  39
  40 typedef struct ChromaprintMuxContext {
  41     const AVClass *class;
  42     int silence_threshold;
  43     int algorithm;
  44     FingerprintFormat fp_format;
  45 #if CPR_VERSION_INT >= AV_VERSION_INT(1, 4, 0)
  46     ChromaprintContext *ctx;
  47 #else
  48     ChromaprintContext ctx;
  49 #endif
  50 } ChromaprintMuxContext;
  51
  52 static void deinit(AVFormatContext *s)
  53 {
  54     ChromaprintMuxContext *const cpr = s->priv_data;
  55
  56     if (cpr->ctx) {
  57         ff_mutex_lock(&chromaprint_mutex);
  58         chromaprint_free(cpr->ctx);
  59         ff_mutex_unlock(&chromaprint_mutex);
  60     }
  61 }
  62
  63 static av_cold int init(AVFormatContext *s)
  64 {
  65     ChromaprintMuxContext *cpr = s->priv_data;
  66     AVStream *st;
  67
  68     ff_mutex_lock(&chromaprint_mutex);
  69     cpr->ctx = chromaprint_new(cpr->algorithm);
  70     ff_mutex_unlock(&chromaprint_mutex);
  71
  72     if (!cpr->ctx) {
  73         av_log(s, AV_LOG_ERROR, "Failed to create chromaprint context.\n");
  74         return AVERROR_EXTERNAL;
  75     }
  76
  77     if (cpr->silence_threshold != -1) {
  78 #if CPR_VERSION_INT >= AV_VERSION_INT(0, 7, 0)
  79         if (!chromaprint_set_option(cpr->ctx, "silence_threshold", cpr->silence_threshold)) {
  80             av_log(s, AV_LOG_ERROR, "Failed to set silence threshold. Setting silence_threshold requires -algorithm 3 option.\n");
  81             return AVERROR_EXTERNAL;
  82         }
  83 #else
  84         av_log(s, AV_LOG_ERROR, "Setting the silence threshold requires Chromaprint "
  85                                 "version 0.7.0 or later.\n");
  86         return AVERROR(ENOSYS);
  87 #endif
  88     }
  89
  90     st = s->streams[0];
  91
  92     if (st->codecpar->ch_layout.nb_channels > 2) {
  93         av_log(s, AV_LOG_ERROR, "Only up to 2 channels are supported\n");
  94         return AVERROR(EINVAL);
  95     }
  96
  97     if (st->codecpar->sample_rate < 1000) {
  98         av_log(s, AV_LOG_ERROR, "Sampling rate must be at least 1000\n");
  99         return AVERROR(EINVAL);
 100     }
 101
 102     if (!chromaprint_start(cpr->ctx, st->codecpar->sample_rate, st->codecpar->ch_layout.nb_channels)) {
 103         av_log(s, AV_LOG_ERROR, "Failed to start chromaprint\n");
 104         return AVERROR_EXTERNAL;
 105     }
 106
 107     return 0;
 108 }
 109
 110 static int write_packet(AVFormatContext *s, AVPacket *pkt)
 111 {
 112     ChromaprintMuxContext *cpr = s->priv_data;
 113     return chromaprint_feed(cpr->ctx, (const int16_t *)pkt->data, pkt->size / 2) ? 0 : AVERROR(EINVAL);
 114 }
 115
 116 static int write_trailer(AVFormatContext *s)
 117 {
 118     ChromaprintMuxContext *cpr = s->priv_data;
 119     AVIOContext *pb = s->pb;
 120     void *fp = NULL;
 121     char *enc_fp = NULL;
 122     int size, enc_size, ret = AVERROR_EXTERNAL;
 123
 124     if (!chromaprint_finish(cpr->ctx)) {
 125         av_log(s, AV_LOG_ERROR, "Failed to generate fingerprint\n");
 126         goto fail;
 127     }
 128
 129     if (!chromaprint_get_raw_fingerprint(cpr->ctx, (uint32_t **)&fp, &size)) {
 130         av_log(s, AV_LOG_ERROR, "Failed to retrieve fingerprint\n");
 131         goto fail;
 132     }
 133
 134     switch (cpr->fp_format) {
 135     case FINGERPRINT_RAW:
 136         avio_write(pb, fp, size * 4); //fp points to array of uint32_t
 137         break;
 138     case FINGERPRINT_COMPRESSED:
 139     case FINGERPRINT_BASE64:
 140         if (!chromaprint_encode_fingerprint(fp, size, cpr->algorithm, &enc_fp, &enc_size,
 141                                             cpr->fp_format == FINGERPRINT_BASE64)) {
 142             av_log(s, AV_LOG_ERROR, "Failed to encode fingerprint\n");
 143             goto fail;
 144         }
 145         avio_write(pb, enc_fp, enc_size);
 146         break;
 147     }
 148
 149     ret = 0;
 150 fail:
 151     if (fp)
 152         chromaprint_dealloc(fp);
 153     if (enc_fp)
 154         chromaprint_dealloc(enc_fp);
 155     return ret;
 156 }
 157
 158 #define OFFSET(x) offsetof(ChromaprintMuxContext, x)
 159 #define FLAGS AV_OPT_FLAG_ENCODING_PARAM
 160 static const AVOption options[] = {
 161     { "silence_threshold", "threshold for detecting silence", OFFSET(silence_threshold), AV_OPT_TYPE_INT, { .i64 = -1 }, -1, 32767, FLAGS },
 162     { "algorithm", "version of the fingerprint algorithm", OFFSET(algorithm), AV_OPT_TYPE_INT, { .i64 = CHROMAPRINT_ALGORITHM_DEFAULT }, CHROMAPRINT_ALGORITHM_TEST1, INT_MAX, FLAGS },
 163     { "fp_format", "fingerprint format to write", OFFSET(fp_format), AV_OPT_TYPE_INT, { .i64 = FINGERPRINT_BASE64 }, FINGERPRINT_RAW, FINGERPRINT_BASE64, FLAGS, .unit = "fp_format" },
 164     { "raw", "binary raw fingerprint", 0, AV_OPT_TYPE_CONST, {.i64 = FINGERPRINT_RAW }, INT_MIN, INT_MAX, FLAGS, .unit = "fp_format"},
 165     { "compressed", "binary compressed fingerprint", 0, AV_OPT_TYPE_CONST, {.i64 = FINGERPRINT_COMPRESSED }, INT_MIN, INT_MAX, FLAGS, .unit = "fp_format"},
 166     { "base64", "Base64 compressed fingerprint", 0, AV_OPT_TYPE_CONST, {.i64 = FINGERPRINT_BASE64 }, INT_MIN, INT_MAX, FLAGS, .unit = "fp_format"},
 167     { NULL },
 168 };
 169
 170 static const AVClass chromaprint_class = {
 171     .class_name = "chromaprint muxer",
 172     .item_name  = av_default_item_name,
 173     .option     = options,
 174     .version    = LIBAVUTIL_VERSION_INT,
 175 };
 176
 177 const FFOutputFormat ff_chromaprint_muxer = {
 178     .p.name            = "chromaprint",
 179     .p.long_name       = NULL_IF_CONFIG_SMALL("Chromaprint"),
 180     .priv_data_size    = sizeof(ChromaprintMuxContext),
 181     .p.audio_codec     = AV_NE(AV_CODEC_ID_PCM_S16BE, AV_CODEC_ID_PCM_S16LE),
 182     .p.video_codec     = AV_CODEC_ID_NONE,
 183     .p.subtitle_codec  = AV_CODEC_ID_NONE,
 184     .flags_internal    = FF_OFMT_FLAG_MAX_ONE_OF_EACH |
 185                          FF_OFMT_FLAG_ONLY_DEFAULT_CODECS,
 186     .init              = init,
 187     .write_packet      = write_packet,
 188     .write_trailer     = write_trailer,
 189     .deinit            = deinit,
 190     .p.flags           = AVFMT_NOTIMESTAMPS,
 191     .p.priv_class      = &chromaprint_class,
 192 };