| 1 |
|
|---|
| 2 | #include "stdafx.h"
|
|---|
| 3 | #include <iostream>
|
|---|
| 4 |
|
|---|
| 5 | extern "C"
|
|---|
| 6 | {
|
|---|
| 7 | #include "libavcodec/avcodec.h"
|
|---|
| 8 | #include "libavformat/avformat.h"
|
|---|
| 9 | //#include "swscale.h"
|
|---|
| 10 | #include "libswresample/swresample.h"
|
|---|
| 11 | };
|
|---|
| 12 |
|
|---|
| 13 | FILE *fin,
|
|---|
| 14 | *fout;
|
|---|
| 15 |
|
|---|
| 16 | int ffmpeg_audio_decode( const char * inFile, const char * outFile)
|
|---|
| 17 | {
|
|---|
| 18 | // Initialize FFmpeg
|
|---|
| 19 | av_register_all();
|
|---|
| 20 |
|
|---|
| 21 | AVFrame* frame = avcodec_alloc_frame();
|
|---|
| 22 | if (!frame)
|
|---|
| 23 | {
|
|---|
| 24 | std::cout << "Error allocating the frame" << std::endl;
|
|---|
| 25 | return 1;
|
|---|
| 26 | }
|
|---|
| 27 |
|
|---|
| 28 | // you can change the file name "01 Push Me to the Floor.wav" to whatever the file is you're reading, like "myFile.ogg" or
|
|---|
| 29 | // "someFile.webm" and this should still work
|
|---|
| 30 | AVFormatContext* formatContext = NULL;
|
|---|
| 31 | //if (avformat_open_input(&formatContext, "01 Push Me to the Floor.wav", NULL, NULL) != 0)
|
|---|
| 32 | if (avformat_open_input(&formatContext, inFile, NULL, NULL) != 0)
|
|---|
| 33 | {
|
|---|
| 34 | av_free(frame);
|
|---|
| 35 | std::cout << "Error opening the file" << std::endl;
|
|---|
| 36 | return 1;
|
|---|
| 37 | }
|
|---|
| 38 |
|
|---|
| 39 | if (avformat_find_stream_info(formatContext, NULL) < 0)
|
|---|
| 40 | {
|
|---|
| 41 | av_free(frame);
|
|---|
| 42 | av_close_input_file(formatContext);
|
|---|
| 43 | std::cout << "Error finding the stream info" << std::endl;
|
|---|
| 44 | return 1;
|
|---|
| 45 | }
|
|---|
| 46 |
|
|---|
| 47 | AVStream* audioStream = NULL;
|
|---|
| 48 | // Find the audio stream (some container files can have multiple streams in them)
|
|---|
| 49 | for (unsigned int i = 0; i < formatContext->nb_streams; ++i)
|
|---|
| 50 | {
|
|---|
| 51 | if (formatContext->streams[i]->codec->codec_type == AVMEDIA_TYPE_AUDIO)
|
|---|
| 52 | {
|
|---|
| 53 | audioStream = formatContext->streams[i];
|
|---|
| 54 | break;
|
|---|
| 55 | }
|
|---|
| 56 | }
|
|---|
| 57 |
|
|---|
| 58 | if (audioStream == NULL)
|
|---|
| 59 | {
|
|---|
| 60 | av_free(frame);
|
|---|
| 61 | av_close_input_file(formatContext);
|
|---|
| 62 | std::cout << "Could not find any audio stream in the file" << std::endl;
|
|---|
| 63 | return 1;
|
|---|
| 64 | }
|
|---|
| 65 |
|
|---|
| 66 | AVCodecContext* codecContext = audioStream->codec;
|
|---|
| 67 |
|
|---|
| 68 | codecContext->codec = avcodec_find_decoder(codecContext->codec_id);
|
|---|
| 69 | if (codecContext->codec == NULL)
|
|---|
| 70 | {
|
|---|
| 71 | av_free(frame);
|
|---|
| 72 | av_close_input_file(formatContext);
|
|---|
| 73 | std::cout << "Couldn't find a proper decoder" << std::endl;
|
|---|
| 74 | return 1;
|
|---|
| 75 | }
|
|---|
| 76 | else if (avcodec_open2(codecContext, codecContext->codec, NULL) != 0)
|
|---|
| 77 | {
|
|---|
| 78 | av_free(frame);
|
|---|
| 79 | av_close_input_file(formatContext);
|
|---|
| 80 | std::cout << "Couldn't open the context with the decoder" << std::endl;
|
|---|
| 81 | return 1;
|
|---|
| 82 | }
|
|---|
| 83 |
|
|---|
| 84 | std::cout << "This stream has " << codecContext->channels << " channels and a sample rate of " << codecContext->sample_rate << "Hz" << std::endl;
|
|---|
| 85 | std::cout << "The data is in the format " << av_get_sample_fmt_name(codecContext->sample_fmt) << std::endl;
|
|---|
| 86 |
|
|---|
| 87 | //codecContext->sample_fmt = AV_SAMPLE_FMT_S16;
|
|---|
| 88 |
|
|---|
| 89 | int64_t outChannelLayout = AV_CH_LAYOUT_MONO; //AV_CH_LAYOUT_STEREO;
|
|---|
| 90 | AVSampleFormat outSampleFormat = AV_SAMPLE_FMT_S16; // Packed audio, non-planar (this is the most common format, and probably what you want; also, WAV needs it)
|
|---|
| 91 | int outSampleRate = 8000;//44100;
|
|---|
| 92 | /*
|
|---|
| 93 | Wav wav;
|
|---|
| 94 | wav.sampleRate = outSampleRate;
|
|---|
| 95 | wav.sampleSize = av_get_bytes_per_sample(outSampleFormat);
|
|---|
| 96 | wav.channels = av_get_channel_layout_nb_channels(outChannelLayout);
|
|---|
| 97 | */
|
|---|
| 98 | // Note that AVCodecContext::channel_layout may or may not be set by libavcodec. Because of this,
|
|---|
| 99 | // we won't use it, and will instead try to guess the layout from the number of channels.
|
|---|
| 100 | SwrContext* swrContext = swr_alloc_set_opts(NULL,
|
|---|
| 101 | outChannelLayout,
|
|---|
| 102 | outSampleFormat,
|
|---|
| 103 | outSampleRate,
|
|---|
| 104 | av_get_default_channel_layout(codecContext->channels),
|
|---|
| 105 | codecContext->sample_fmt,
|
|---|
| 106 | codecContext->sample_rate,
|
|---|
| 107 | 0,
|
|---|
| 108 | NULL);
|
|---|
| 109 |
|
|---|
| 110 | if (swrContext == NULL)
|
|---|
| 111 | {
|
|---|
| 112 | av_free(frame);
|
|---|
| 113 | avcodec_close(codecContext);
|
|---|
| 114 | avformat_close_input(&formatContext);
|
|---|
| 115 | std::cout << "Couldn't create the SwrContext" << std::endl;
|
|---|
| 116 | return 1;
|
|---|
| 117 | }
|
|---|
| 118 |
|
|---|
| 119 | if (swr_init(swrContext) != 0)
|
|---|
| 120 | {
|
|---|
| 121 | av_free(frame);
|
|---|
| 122 | avcodec_close(codecContext);
|
|---|
| 123 | avformat_close_input(&formatContext);
|
|---|
| 124 | swr_free(&swrContext);
|
|---|
| 125 | std::cout << "Couldn't initialize the SwrContext" << std::endl;
|
|---|
| 126 | return 1;
|
|---|
| 127 | }
|
|---|
| 128 |
|
|---|
| 129 | fout = fopen(outFile, "wb+");
|
|---|
| 130 |
|
|---|
| 131 | AVPacket packet;
|
|---|
| 132 | av_init_packet(&packet);
|
|---|
| 133 |
|
|---|
| 134 | // Read the packets in a loop
|
|---|
| 135 | while (av_read_frame(formatContext, &packet) == 0)
|
|---|
| 136 | {
|
|---|
| 137 | if (packet.stream_index == audioStream->index)
|
|---|
| 138 | {
|
|---|
| 139 | AVPacket decodingPacket = packet;
|
|---|
| 140 |
|
|---|
| 141 | while (decodingPacket.size > 0)
|
|---|
| 142 | {
|
|---|
| 143 | // Try to decode the packet into a frame
|
|---|
| 144 | int frameFinished = 0;
|
|---|
| 145 | int result = avcodec_decode_audio4(codecContext, frame, &frameFinished, &decodingPacket);
|
|---|
| 146 |
|
|---|
| 147 | if (result < 0 || frameFinished == 0)
|
|---|
| 148 | {
|
|---|
| 149 | break;
|
|---|
| 150 | }
|
|---|
| 151 |
|
|---|
| 152 | //std::vector<unsigned char> buffer(wav.channels * wav.sampleRate * wav.sampleSize);
|
|---|
| 153 |
|
|---|
| 154 | unsigned char buffer[100000] = {NULL};
|
|---|
| 155 | unsigned char* pointers[SWR_CH_MAX] = {NULL};
|
|---|
| 156 | pointers[0] = &buffer[0];
|
|---|
| 157 |
|
|---|
| 158 | int numSamplesOut = swr_convert(swrContext,
|
|---|
| 159 | pointers,
|
|---|
| 160 | outSampleRate, //wav.sampleRate,
|
|---|
| 161 | (const unsigned char**)frame->extended_data,
|
|---|
| 162 | //(const uint8_t**)frame->extended_data[0],
|
|---|
| 163 | frame->nb_samples);
|
|---|
| 164 |
|
|---|
| 165 | //processFrame(frame, swrContext, wav);
|
|---|
| 166 |
|
|---|
| 167 | //fwrite( frame->data[0], sizeof(short), (size_t)(frame->nb_samples), fout);
|
|---|
| 168 | //fwrite( frame->extended_data, sizeof(short), (size_t)(frame->nb_samples), fout);
|
|---|
| 169 | //uint16_t uiCnt_1 = (uint16_t )frame->extended_data[0];
|
|---|
| 170 | //uint16_t uiCnt_2 = (uint16_t )frame->extended_data[1];
|
|---|
| 171 |
|
|---|
| 172 | /*
|
|---|
| 173 | ReSampleContext *rs_ctx = NULL;
|
|---|
| 174 | // resample to 44100, stereo, s16
|
|---|
| 175 | rs_ctx = av_audio_resample_init(
|
|---|
| 176 | 1, codecContext->channels,
|
|---|
| 177 | 8000, codecContext->sample_rate,
|
|---|
| 178 | AV_SAMPLE_FMT_S16, codecContext->sample_fmt,
|
|---|
| 179 | 16, 10, 0, 1);
|
|---|
| 180 |
|
|---|
| 181 | //outbuff = (uint8_t*)av_malloc(AVCODEC_MAX_AUDIO_FRAME_SIZE);
|
|---|
| 182 | short bufferSh[100000] = {NULL};
|
|---|
| 183 | // resampling
|
|---|
| 184 | //int after_sampled_len = audio_resample(rs_ctx, (short *)buffer, (short *)frame->extended_data[0], frame->nb_samples);
|
|---|
| 185 | int after_sampled_len = audio_resample(rs_ctx, bufferSh, (short *)frame->extended_data, frame->nb_samples);
|
|---|
| 186 | */
|
|---|
| 187 |
|
|---|
| 188 | fwrite( (short *)buffer, sizeof(short), (size_t)numSamplesOut, fout);
|
|---|
| 189 |
|
|---|
| 190 | decodingPacket.size -= result;
|
|---|
| 191 | decodingPacket.data += result;
|
|---|
| 192 | }
|
|---|
| 193 |
|
|---|
| 194 | /*
|
|---|
| 195 | // Try to decode the packet into a frame
|
|---|
| 196 | int frameFinished = 0;
|
|---|
| 197 | avcodec_decode_audio4(codecContext, frame, &frameFinished, &packet);
|
|---|
| 198 |
|
|---|
| 199 | // Some frames rely on multiple packets, so we have to make sure the frame is finished before
|
|---|
| 200 | // we can use it
|
|---|
| 201 | if (frameFinished)
|
|---|
| 202 | {
|
|---|
| 203 |
|
|---|
| 204 | //fwrite( (short *)&(frame->data[0]), sizeof(short), (size_t)(frame->nb_samples*2), fout);
|
|---|
| 205 | fwrite( (short *)&(frame->data[0]), sizeof(short), (size_t)(frame->nb_samples*2), fout);
|
|---|
| 206 |
|
|---|
| 207 | // frame now has usable audio data in it. How it's stored in the frame depends on the format of
|
|---|
| 208 | // the audio. If it's packed audio, all the data will be in frame->data[0]. If it's in planar format,
|
|---|
| 209 | // the data will be in frame->data and possibly frame->extended_data. Look at frame->data, frame->nb_samples,
|
|---|
| 210 | // frame->linesize, and other related fields on the FFmpeg docs. I don't know how you're actually using
|
|---|
| 211 | // the audio data, so I won't add any junk here that might confuse you. Typically, if I want to find
|
|---|
| 212 | // documentation on an FFmpeg structure or function, I just type "<name> doxygen" into google (like
|
|---|
| 213 | // "AVFrame doxygen" for AVFrame's docs)
|
|---|
| 214 | }
|
|---|
| 215 | */
|
|---|
| 216 | }
|
|---|
| 217 |
|
|---|
| 218 | // You *must* call av_free_packet() after each call to av_read_frame() or else you'll leak memory
|
|---|
| 219 | av_free_packet(&packet);
|
|---|
| 220 | }
|
|---|
| 221 |
|
|---|
| 222 | // Some codecs will cause frames to be buffered up in the decoding process. If the CODEC_CAP_DELAY flag
|
|---|
| 223 | // is set, there can be buffered up frames that need to be flushed, so we'll do that
|
|---|
| 224 | if (codecContext->codec->capabilities & CODEC_CAP_DELAY)
|
|---|
| 225 | {
|
|---|
| 226 | av_init_packet(&packet);
|
|---|
| 227 | // Decode all the remaining frames in the buffer, until the end is reached
|
|---|
| 228 | int frameFinished = 0;
|
|---|
| 229 | while (avcodec_decode_audio4(codecContext, frame, &frameFinished, &packet) >= 0 && frameFinished)
|
|---|
| 230 | {
|
|---|
| 231 | }
|
|---|
| 232 | }
|
|---|
| 233 |
|
|---|
| 234 | // Clean up!
|
|---|
| 235 | av_free(frame);
|
|---|
| 236 | avcodec_close(codecContext);
|
|---|
| 237 | av_close_input_file(formatContext);
|
|---|
| 238 | fclose(fout);
|
|---|
| 239 | }
|
|---|
| 240 |
|
|---|
| 241 |
|
|---|