One of the things I'm trying to achieve is parallel encoding via FFmpeg's c API. This looks to work out of the box quite nicely; however, I've changed the goal posts slightly:
In an existing application, I already have a thread pool at hand. Instead of using another thread pool via FFmpeg, I would like reuse the existing thread pool in my application. Having studied the latest FFmpeg trunk docs, it very much looks possible.
Using some FFmpeg sample code, I've created a sample application to demonstrate what I'm trying to achieve (see below). The sample app generates a video-only mpeg2 ts using the mp2v codec.
The problem I'm experiencing is that the custom 'thread_execute' or 'thread_execute2' are never invoked. This is despite the fact that the codec appears to indicate that threading is supported. Please be aware that I have not yet plumbed in the thread pool just yet. My first goal is for it to call the custom function pointer.
I've tried to get assistance on the FFmpeg mailing lists but to no avail.
#include <iostream>
#include <thread>
#include <stdlib.h>
#include <string.h>
#include <math.h>
#include <cstring>
#include <future>
extern "C"
{
#include <libavutil/avassert.h>
#include <libavutil/channel_layout.h>
#include <libavutil/opt.h>
#include <libavutil/timestamp.h>
#include <libavformat/avformat.h>
//#include <libswscale/swscale.h>
#include <libswresample/swresample.h>
}
#define STREAM_DURATION 1000.0
#define STREAM_FRAME_RATE 25 /* 25 images/s */
#define STREAM_PIX_FMT AV_PIX_FMT_YUV420P /* default pix_fmt */
#define SCALE_FLAGS SWS_BICUBIC
// a wrapper around a single output AVStream
typedef struct OutputStream {
AVStream *st;
AVCodecContext *enc;
/* pts of the next frame that will be generated */
int64_t next_pts;
int samples_count;
AVFrame *frame;
AVFrame *tmp_frame;
float t, tincr, tincr2;
struct SwsContext *sws_ctx;
struct SwrContext *swr_ctx;
} OutputStream;
/////////////////////////////////////////////////////////////////////////////
// The ffmpeg variation raises compiler warnings.
char *cb_av_ts2str(char *buf, int64_t ts)
{
std::memset(buf,0,AV_TS_MAX_STRING_SIZE);
return av_ts_make_string(buf,ts);
}
/////////////////////////////////////////////////////////////////////////////
// The ffmpeg variation raises compiler warnings.
char *cb_av_ts2timestr(char *buf, int64_t ts, AVRational *tb)
{
std::memset(buf,0,sizeof(AV_TS_MAX_STRING_SIZE));
return av_ts_make_time_string(buf,ts,tb);
}
/////////////////////////////////////////////////////////////////////////////
// The ffmpeg variation raises compiler warnings.
char *cb_av_err2str(char *errbuf, size_t errbuf_size, int errnum)
{
std::memset(errbuf,0,errbuf_size);
return av_make_error_string(errbuf,errbuf_size,errnum);
}
int thread_execute(AVCodecContext* s, int (*func)(AVCodecContext *c2, void *arg2), void* arg, int* ret, int count, int size)
{
// Do it all serially for now
std::cout << "thread_execute" << std::endl;
for (int k = 0; k < count; ++k)
{
ret[k] = func(s, arg);
}
return 0;
}
int thread_execute2(AVCodecContext* s, int (*func)(AVCodecContext* c2, void* arg2, int, int), void* arg, int* ret, int count)
{
// Do it all serially for now
std::cout << "thread_execute2" << std::endl;
for (int k = 0; k < count; ++k)
{
ret[k] = func(s, arg, k, count);
}
return 0;
}
static void log_packet(const AVFormatContext *fmt_ctx, const AVPacket *pkt)
{
char s[AV_TS_MAX_STRING_SIZE];
AVRational *time_base = &fmt_ctx->streams[pkt->stream_index]->time_base;
printf("pts:%s pts_time:%s dts:%s dts_time:%s duration:%s duration_time:%s stream_index:%d\n",
cb_av_ts2str(s,pkt->pts), cb_av_ts2timestr(s,pkt->pts, time_base),
cb_av_ts2str(s,pkt->dts), cb_av_ts2timestr(s,pkt->dts, time_base),
cb_av_ts2str(s,pkt->duration), cb_av_ts2timestr(s,pkt->duration, time_base),
pkt->stream_index);
}
static int write_frame(AVFormatContext *fmt_ctx, const AVRational *time_base, AVStream *st, AVPacket *pkt)
{
/* rescale output packet timestamp values from codec to stream timebase */
av_packet_rescale_ts(pkt, *time_base, st->time_base);
pkt->stream_index = st->index;
/* Write the compressed frame to the media file. */
log_packet(fmt_ctx, pkt);
return av_interleaved_write_frame(fmt_ctx, pkt);
}
/* Add an output stream. */
static void add_stream(OutputStream *ost, AVFormatContext *oc,
AVCodec **codec,
enum AVCodecID codec_id)
{
AVCodecContext *c;
int i;
/* find the encoder */
*codec = avcodec_find_encoder(codec_id);
if (!(*codec)) {
fprintf(stderr, "Could not find encoder for '%s'\n",
avcodec_get_name(codec_id));
exit(1);
}
ost->st = avformat_new_stream(oc, NULL);
if (!ost->st) {
fprintf(stderr, "Could not allocate stream\n");
exit(1);
}
ost->st->id = oc->nb_streams-1;
c = avcodec_alloc_context3(*codec);
if (!c) {
fprintf(stderr, "Could not alloc an encoding context\n");
exit(1);
}
ost->enc = c;
switch ((*codec)->type)
{
case AVMEDIA_TYPE_AUDIO:
c->sample_fmt = (*codec)->sample_fmts ?
(*codec)->sample_fmts[0] : AV_SAMPLE_FMT_FLTP;
c->bit_rate = 64000;
c->sample_rate = 44100;
if ((*codec)->supported_samplerates) {
c->sample_rate = (*codec)->supported_samplerates[0];
for (i = 0; (*codec)->supported_samplerates[i]; i++) {
if ((*codec)->supported_samplerates[i] == 44100)
c->sample_rate = 44100;
}
}
c->channels = av_get_channel_layout_nb_channels(c->channel_layout);
c->channel_layout = AV_CH_LAYOUT_STEREO;
if ((*codec)->channel_layouts) {
c->channel_layout = (*codec)->channel_layouts[0];
for (i = 0; (*codec)->channel_layouts[i]; i++) {
if ((*codec)->channel_layouts[i] == AV_CH_LAYOUT_STEREO)
c->channel_layout = AV_CH_LAYOUT_STEREO;
}
}
c->channels = av_get_channel_layout_nb_channels(c->channel_layout);
ost->st->time_base = (AVRational){ 1, c->sample_rate };
break;
case AVMEDIA_TYPE_VIDEO:
c->codec_id = codec_id;
c->bit_rate = 400000;
/* Resolution must be a multiple of two. */
c->width = 352;
c->height = 288;
/* timebase: This is the fundamental unit of time (in seconds) in terms
* of which frame timestamps are represented. For fixed-fps content,
* timebase should be 1/framerate and timestamp increments should be
* identical to 1. */
ost->st->time_base = (AVRational){ 1, STREAM_FRAME_RATE };
c->time_base = ost->st->time_base;
c->gop_size = 12; /* emit one intra frame every twelve frames at most */
c->pix_fmt = STREAM_PIX_FMT;
if (c->codec_id == AV_CODEC_ID_MPEG2VIDEO) {
/* just for testing, we also add B-frames */
c->max_b_frames = 2;
}
if (c->codec_id == AV_CODEC_ID_MPEG1VIDEO) {
/* Needed to avoid using macroblocks in which some coeffs overflow.
* This does not happen with normal video, it just happens here as
* the motion of the chroma plane does not match the luma plane. */
c->mb_decision = 2;
}
break;
default:
break;
}
if (c->codec->capabilities & AV_CODEC_CAP_FRAME_THREADS ||
c->codec->capabilities & AV_CODEC_CAP_SLICE_THREADS)
{
if (c->codec->capabilities & AV_CODEC_CAP_FRAME_THREADS)
{
c->thread_type = FF_THREAD_FRAME;
}
if (c->codec->capabilities & AV_CODEC_CAP_SLICE_THREADS)
{
c->thread_type = FF_THREAD_SLICE;
}
c->execute = &thread_execute;
c->execute2 = &thread_execute2;
c->thread_count = 4;
// NOTE: Testing opaque.
c->opaque = (void*)0xff;
}
/* Some formats want stream headers to be separate. */
if (oc->oformat->flags & AVFMT_GLOBALHEADER)
c->flags |= AV_CODEC_FLAG_GLOBAL_HEADER;
}
/**************************************************************/
/* video output */
static AVFrame *alloc_picture(enum AVPixelFormat pix_fmt, int width, int height)
{
AVFrame *picture;
int ret;
picture = av_frame_alloc();
if (!picture)
return NULL;
picture->format = pix_fmt;
picture->width = width;
picture->height = height;
/* allocate the buffers for the frame data */
ret = av_frame_get_buffer(picture, 32);
if (ret < 0) {
fprintf(stderr, "Could not allocate frame data.\n");
exit(1);
}
return picture;
}
static void open_video(AVFormatContext *oc, AVCodec *codec, OutputStream *ost, AVDictionary *opt_arg)
{
int ret;
AVCodecContext *c = ost->enc;
//AVDictionary *opt = NULL;
//av_dict_copy(&opt, opt_arg, 0);
/* open the codec */
ret = avcodec_open2(c, codec, NULL);
//av_dict_free(&opt);
if (ret < 0) {
char s[AV_ERROR_MAX_STRING_SIZE];
fprintf(stderr, "Could not open video codec: %s\n", cb_av_err2str(s,AV_ERROR_MAX_STRING_SIZE,ret));
exit(1);
}
/* allocate and init a re-usable frame */
ost->frame = alloc_picture(c->pix_fmt, c->width, c->height);
if (!ost->frame) {
fprintf(stderr, "Could not allocate video frame\n");
exit(1);
}
/* If the output format is not YUV420P, then a temporary YUV420P
* picture is needed too. It is then converted to the required
* output format. */
ost->tmp_frame = NULL;
if (c->pix_fmt != AV_PIX_FMT_YUV420P) {
ost->tmp_frame = alloc_picture(AV_PIX_FMT_YUV420P, c->width, c->height);
if (!ost->tmp_frame) {
fprintf(stderr, "Could not allocate temporary picture\n");
exit(1);
}
}
/* copy the stream parameters to the muxer */
ret = avcodec_parameters_from_context(ost->st->codecpar, c);
if (ret < 0) {
fprintf(stderr, "Could not copy the stream parameters\n");
exit(1);
}
}
/* Prepare a dummy image. */
static void fill_yuv_image(AVFrame *pict, int frame_index,
int width, int height)
{
int x, y, i;
i = frame_index;
/* Y */
for (y = 0; y < height; y++)
for (x = 0; x < width; x++)
pict->data[0][y * pict->linesize[0] + x] = x + y + i * 3;
/* Cb and Cr */
for (y = 0; y < height / 2; y++) {
for (x = 0; x < width / 2; x++) {
pict->data[1][y * pict->linesize[1] + x] = 128 + y + i * 2;
pict->data[2][y * pict->linesize[2] + x] = 64 + x + i * 5;
}
}
}
static AVFrame *get_video_frame(OutputStream *ost)
{
AVCodecContext *c = ost->enc;
/* check if we want to generate more frames */
if (av_compare_ts(ost->next_pts, c->time_base,
STREAM_DURATION, (AVRational){ 1, 1 }) >= 0)
return NULL;
/* when we pass a frame to the encoder, it may keep a reference to it
* internally; make sure we do not overwrite it here */
if (av_frame_make_writable(ost->frame) < 0)
exit(1);
if (c->pix_fmt != AV_PIX_FMT_YUV420P) {
/* as we only generate a YUV420P picture, we must convert it
* to the codec pixel format if needed */
/*if (!ost->sws_ctx) {
ost->sws_ctx = sws_getContext(c->width, c->height,
AV_PIX_FMT_YUV420P,
c->width, c->height,
c->pix_fmt,
SCALE_FLAGS, NULL, NULL, NULL);
if (!ost->sws_ctx) {
fprintf(stderr,
"Could not initialize the conversion context\n");
exit(1);
}
}
fill_yuv_image(ost->tmp_frame, ost->next_pts, c->width, c->height);
sws_scale(ost->sws_ctx,
(const uint8_t * const *)ost->tmp_frame->data, ost->tmp_frame->linesize,
0, c->height, ost->frame->data, ost->frame->linesize);*/
} else {
fill_yuv_image(ost->frame, ost->next_pts, c->width, c->height);
}
ost->frame->pts = ost->next_pts++;
return ost->frame;
}
/*
* encode one video frame and send it to the muxer
* return 1 when encoding is finished, 0 otherwise
*/
static int write_video_frame(AVFormatContext *oc, OutputStream *ost)
{
int ret;
AVCodecContext *c;
AVFrame *frame;
int got_packet = 0;
AVPacket pkt = { 0 };
c = ost->enc;
frame = get_video_frame(ost);
if (frame)
{
ret = avcodec_send_frame(ost->enc, frame);
if (ret < 0)
{
char s[AV_ERROR_MAX_STRING_SIZE];
fprintf(stderr, "Error encoding video frame: %s\n", cb_av_err2str(s, AV_ERROR_MAX_STRING_SIZE, ret));
exit(1);
}
}
av_init_packet(&pkt);
ret = avcodec_receive_packet(ost->enc,&pkt);
if (ret < 0)
{
if (ret == AVERROR(EAGAIN)) { ret = 0; }
else
{
char s[AV_ERROR_MAX_STRING_SIZE];
fprintf(stderr, "Error receiving packet: %s\n", cb_av_err2str(s,AV_ERROR_MAX_STRING_SIZE,ret));
exit(1);
}
}
else
{
got_packet = 1;
ret = write_frame(oc, &c->time_base, ost->st, &pkt);
}
if (ret < 0) {
char s[AV_ERROR_MAX_STRING_SIZE];
fprintf(stderr, "Error while writing video frame: %s\n", cb_av_err2str(s,AV_ERROR_MAX_STRING_SIZE,ret));
exit(1);
}
return (frame || got_packet) ? 0 : 1;
}
static void close_stream(AVFormatContext *oc, OutputStream *ost)
{
avcodec_free_context(&ost->enc);
av_frame_free(&ost->frame);
av_frame_free(&ost->tmp_frame);
//sws_freeContext(ost->sws_ctx);
//swr_free(&ost->swr_ctx);
}
/**************************************************************/
/* media file output */
int main(int argc, char **argv)
{
OutputStream video_st = { 0 }, audio_st = { 0 };
const char *filename;
AVOutputFormat *fmt;
AVFormatContext *oc;
AVCodec /**audio_codec,*/ *video_codec;
int ret;
int have_video = 0, have_audio = 0;
int encode_video = 0, encode_audio = 0;
AVDictionary *opt = NULL;
int i;
/* Initialize libavcodec, and register all codecs and formats. */
av_register_all();
avformat_network_init();
if (argc < 2) {
printf("usage: %s output_file\n"
"API example program to output a media file with libavformat.\n"
"This program generates a synthetic audio and video stream, encodes and\n"
"muxes them into a file named output_file.\n"
"The output format is automatically guessed according to the file extension.\n"
"Raw images can also be output by using '%%d' in the filename.\n"
"\n", argv[0]);
return 1;
}
filename = argv[1];
for (i = 2; i+1 < argc; i+=2) {
if (!strcmp(argv[i], "-flags") || !strcmp(argv[i], "-fflags"))
av_dict_set(&opt, argv[i]+1, argv[i+1], 0);
}
const char *pfilename = filename;
/* allocate the output media context */
avformat_alloc_output_context2(&oc, NULL, "mpegts", pfilename);
if (!oc) {
printf("Could not deduce output format from file extension: using MPEG.\n");
avformat_alloc_output_context2(&oc, NULL, "mpeg", pfilename);
}
if (!oc)
return 1;
fmt = oc->oformat;
/* Add the audio and video streams using the default format codecs
* and initialize the codecs. */
if (fmt->video_codec != AV_CODEC_ID_NONE) {
add_stream(&video_st, oc, &video_codec, fmt->video_codec);
have_video = 1;
encode_video = 1;
}
/*if (fmt->audio_codec != AV_CODEC_ID_NONE) {
add_stream(&audio_st, oc, &audio_codec, fmt->audio_codec);
have_audio = 1;
encode_audio = 1;
}*/
/* Now that all the parameters are set, we can open the audio and
* video codecs and allocate the necessary encode buffers. */
if (have_video)
open_video(oc, video_codec, &video_st, opt);
//if (have_audio)
// open_audio(oc, audio_codec, &audio_st, opt);
av_dump_format(oc, 0, pfilename, 1);
/* open the output file, if needed */
if (!(fmt->flags & AVFMT_NOFILE)) {
ret = avio_open(&oc->pb, pfilename, AVIO_FLAG_WRITE);
if (ret < 0) {
char s[AV_ERROR_MAX_STRING_SIZE];
fprintf(stderr, "Could not open '%s': %s\n", pfilename,
cb_av_err2str(s,AV_ERROR_MAX_STRING_SIZE,ret));
return 1;
}
}
/* Write the stream header, if any. */
ret = avformat_write_header(oc, &opt);
if (ret < 0) {
char s[AV_ERROR_MAX_STRING_SIZE];
fprintf(stderr, "Error occurred when opening output file: %s\n",
cb_av_err2str(s,AV_ERROR_MAX_STRING_SIZE,ret));
return 1;
}
while (encode_video || encode_audio) {
/* select the stream to encode */
if (encode_video &&
(!encode_audio || av_compare_ts(video_st.next_pts, video_st.enc->time_base,
audio_st.next_pts, audio_st.enc->time_base) <= 0)) {
encode_video = !write_video_frame(oc, &video_st);
} else {
//encode_audio = !write_audio_frame(oc, &audio_st);
}
//std::this_thread::sleep_for(std::chrono::milliseconds(35));
}
/* Write the trailer, if any. The trailer must be written before you
* close the CodecContexts open when you wrote the header; otherwise
* av_write_trailer() may try to use memory that was freed on
* av_codec_close(). */
av_write_trailer(oc);
/* Close each codec. */
if (have_video)
close_stream(oc, &video_st);
if (have_audio)
close_stream(oc, &audio_st);
if (!(fmt->flags & AVFMT_NOFILE))
/* Close the output file. */
avio_closep(&oc->pb);
/* free the stream */
avformat_free_context(oc);
return 0;
}
//
Environment:
Ubuntu Zesty (17.04)
FFmpeg version 3.2.4 (via package manager)
gcc 6.3 (C++)
You have to do following:
call avcodec_alloc_context3(...). This call will set default execute and execute2 functions in new context
set c->thread_count = number_of_threads_in_your_thread_pool()
call avcodec_open2(...).
set c->execute and c->execute2 to point to your functions
call ff_thread_free(c). This function isnt exposed in libavcodec headers but you can add following line:
extern "C" void ff_thread_free(AVCodecContext *s);
Drawback is that libavcodec will create internal thread pool after avcodec_open2(...) call, and that pool will be deleted in ff_thread_free() call.
Internal thread pool is very efficient, but its not good if you plan to do parallel encoding of multiple video feeds. In that case libavcodec will create separate thread pool for each encoding video feed.
Related
I can't seem to understand why this doesn't work I’m trying to get a sound sample to a given function.
I've based my code on a version of the function that uses objective-c which works.
But the code below written in C++ doesn't it meant to function by passing a float buffer to OXY_DecodeAudioBuffer function that latter tries and looks for data in the buffer.
Question: Am I passing the right buffer size and output from the buffer to the function? I always get no data found in buffer. Can anyone see the issue?
The hardware I'm using is Raspberry Pi 2 with a USB microphone.
I’ve also included the function with the description:
//OXY_DecodeAudioBuffer function, receives an audiobuffer of specified size and outputs if encoded data is found
//* Parameters:
// audioBuffer: float array of bufferSize size with audio data to be decoded
// size: size of audioBuffer
// oxyingObject: OXY object instance, created in OXY_Create()
//* Returns: -1 if no decoded data is found, -2 if start token is found, -3 if complete word has been decoded, positive number if character is decoded (number is the token idx)
OXY_DLLEXPORT int32_t OXY_DecodeAudioBuffer(float *audioBuffer, int size, void *oxyingObject);
The float_buffer output from the code below:
1. -0.00354004 -0.00369263 -0.00338745 -0.00354004 -0.00341797 -0.00402832
Program Code:
#include <stdio.h>
#include <stdlib.h>
#include <alsa/asoundlib.h>
#include <unistd.h>
#include <math.h>
#include <time.h>
#include <stdio.h>
#include <stdlib.h>
#include <iostream>
#include "Globals.h"
#include "OxyCoreLib_api.h"
void* mCore;
using namespace std;
void GetDecodedMode(){
std::cerr << "DECODE_MODE ---> " << OXY_GetDecodedMode(mCore) << std::endl << std::endl;
}
int main(void)
{
int i,j;
int err;
int mode = 3;
int16_t *buffer;
float* float_buffer;
// Allocate our own buffers (1 channel, 16 bits per sample, thus 16 bits per frame, thus 2 bytes per frame).
// Practice learns the buffers used contain 512 frames, if this changes it will be fixed in processAudio.
int buffer_frames = 512; //Not sure this correct but reason above
unsigned int rate = 44100;
float sampleRate = 44100.f; //to configure
snd_pcm_t *capture_handle;
snd_pcm_hw_params_t *hw_params;
snd_pcm_format_t format = SND_PCM_FORMAT_S16_LE;
if ((err = snd_pcm_open(&capture_handle, "hw:1,0", SND_PCM_STREAM_CAPTURE, 0)) < 0) {
fprintf(stderr, "cannot open audio device %s (%s)\n","device",snd_strerror(err));
exit(1);
} else {fprintf(stdout, "audio interface opened\n");}
if ((err = snd_pcm_hw_params_malloc(&hw_params)) < 0) {
fprintf(stderr, "cannot allocate hardware parameter structure (%s)\n",
snd_strerror(err));
exit(1);
} else { fprintf(stdout, "hw_params allocated\n"); }
if ((err = snd_pcm_hw_params_any(capture_handle, hw_params)) < 0) {
fprintf(stderr, "cannot initialize hardware parameter structure (%s)\n",
snd_strerror(err));
exit(1);
} else { fprintf(stdout, "hw_params initialized\n"); }
if ((err = snd_pcm_hw_params_set_access(capture_handle, hw_params, SND_PCM_ACCESS_RW_INTERLEAVED)) < 0) {
fprintf(stderr, "cannot set access type (%s)\n",
snd_strerror(err));
exit(1);
} else { fprintf(stdout, "hw_params access set\n"); }
if ((err = snd_pcm_hw_params_set_format(capture_handle, hw_params, format)) < 0) {
fprintf(stderr, "cannot set sample format (%s)\n",
snd_strerror(err));
exit(1);
} else { fprintf(stdout, "hw_params format set\n"); }
if ((err = snd_pcm_hw_params_set_rate_near(capture_handle, hw_params, &rate, 0)) < 0) {
fprintf(stderr, "cannot set sample rate (%s)\n",
snd_strerror(err));
exit(1);
} else { fprintf(stdout, "hw_params rate set\n"); }
if ((err = snd_pcm_hw_params_set_channels(capture_handle, hw_params, 1)) < 0) {
fprintf(stderr, "cannot set channel count (%s)\n",
snd_strerror(err));
exit(1);
} else { fprintf(stdout, "hw_params channels set\n"); }
if ((err = snd_pcm_hw_params(capture_handle, hw_params)) < 0) {
fprintf(stderr, "cannot set parameters (%s)\n",
snd_strerror(err));
exit(1);
} else { fprintf(stdout, "hw_params set\n"); }
snd_pcm_hw_params_free(hw_params);
fprintf(stdout, "hw_params freed\n");
if ((err = snd_pcm_prepare(capture_handle)) < 0) {
fprintf(stderr, "cannot prepare audio interface for use (%s)\n",
snd_strerror(err));
exit(1);
} else { fprintf(stdout, "audio interface prepared\n"); }
//allocate buffer of 16bit ints, as specified in PCM_FORMAT
//initialise
mCore = OXY_Create();
//Configure - Mode 3 inaudible, 44100, bufferSize
OXY_Configure(mode, sampleRate, buffer_frames, mCore);
//Debug to make sure
GetDecodedMode();
buffer = static_cast<int16_t*>(malloc(buffer_frames * snd_pcm_format_width(format) / 8 * 2));
//buffer = malloc(buffer_frames * snd_pcm_format_width(format) / 8 * 2);
float_buffer = static_cast<float*>(malloc(buffer_frames*sizeof(float)));
//float_buffer = malloc(buffer_frames*sizeof(float));
fprintf(stdout, "buffer allocated\n");
//where did 10000 come from doubt its correct
for (i = 0; i < 10000; ++i) {
//read from audio device into buffer
if ((err = snd_pcm_readi(capture_handle, buffer, buffer_frames)) != buffer_frames) {
fprintf(stderr, "read from audio interface failed (%s)\n",
err, snd_strerror(err));
exit(1);
}
//try to change buffer from short ints to floats for transformation
for (i = 0; i < buffer_frames; i++){
//norm
float_buffer[i] = (float)buffer[i]/32768.0;
//Example output of float_buffer
/*
-0.00354004
-0.00369263
-0.00338745
-0.00354004
-0.00341797
-0.00402832
-0.00341797
-0.00427246
-0.00375366
-0.00378418
-0.00408936
-0.00332642
-0.00369263
-0.00350952
-0.00369263
-0.00369263
-0.00344849
-0.00354004
*/
}
//send to float_to be tested
int ret = OXY_DecodeAudioBuffer(float_buffer, buffer_frames, mCore);
if (ret == -2)
{
std::cerr << "FOUND_TOKEN ---> -2 " << std::endl << std::endl;
}
else if(ret>=0)
{
std::cerr << "Decode started ---> -2 " << ret << std::endl << std::endl;
}
else if (ret == -3)
{
//int sizeStringDecoded = OXY_GetDecodedData(mStringDecoded, mCore);
std::cerr << "STRING DECODED ---> -2 " << std::endl << std::endl;
// ...
}
else
{
std::cerr << "No data found in this buffer" << std::endl << std::endl;
//no data found in this buffer
}
}
free(buffer);
snd_pcm_close(capture_handle);
std::cerr << "memory freed\n" << std::endl << std::endl;
//snd_pcm_close(capture_handle);
return(0);
//exit(0);
}
Working objective-c version using the same API:
//
// IosAudioController.m
//
#import "IosAudioController.h"
#import <AudioToolbox/AudioToolbox.h>
#import "OxyCoreLib_api.h"
#define kOutputBus 0
#define kInputBus 1
IosAudioController* iosAudio;
void checkStatus(int status){
if (status) {
printf("Status not 0! %d\n", status);
exit(1);
}
}
static OSStatus recordingCallback(void *inRefCon,
AudioUnitRenderActionFlags *ioActionFlags,
const AudioTimeStamp *inTimeStamp,
UInt32 inBusNumber,
UInt32 inNumberFrames,
AudioBufferList *ioData) {
if (iosAudio->mOxyObject->mDecoding == 0)
return noErr;
// Because of the way our audio format (setup below) is chosen:
// we only need 1 buffer, since it is mono
// Samples are 16 bits = 2 bytes.
// 1 frame includes only 1 sample
AudioBuffer buffer;
buffer.mNumberChannels = 1;
buffer.mDataByteSize = inNumberFrames * 2;
buffer.mData = malloc( inNumberFrames * 2 );
// Put buffer in a AudioBufferList
AudioBufferList bufferList;
bufferList.mNumberBuffers = 1;
bufferList.mBuffers[0] = buffer;
// Then:
// Obtain recorded samples
OSStatus status;
status = AudioUnitRender([iosAudio audioUnit],
ioActionFlags,
inTimeStamp,
inBusNumber,
inNumberFrames,
&bufferList);
checkStatus(status);
// Now, we have the samples we just read sitting in buffers in bufferList
// Process the new data
[iosAudio processAudio:&bufferList];
//Now Decode Audio *******************
//convert from AudioBuffer format to *float buffer
iosAudio->floatBuffer = (float *)malloc(inNumberFrames * sizeof(float));
//UInt16 *frameBuffer = bufferList.mBuffers[0].mData;
SInt16 *frameBuffer = bufferList.mBuffers[0].mData;
for(int j=0;j<inNumberFrames;j++)
{
iosAudio->floatBuffer[j] = frameBuffer[j]/32768.0;
}
int ret = OXY_DecodeAudioBuffer(iosAudio->floatBuffer, inNumberFrames, (void*)iosAudio->mOxyObject->mOxyCore);
if (ret == -2)
{
// NSLog(#"BEGIN TOKEN FOUND!");
[iosAudio->mObject performSelector:iosAudio->mSelector withObject:[NSNumber numberWithInt:0]];
}
else if (ret >= 0)
{
NSLog(#"Decode started %#",#(ret).stringValue);
}
else if (ret == -3)
{
int sizeStringDecoded = OXY_GetDecodedData(iosAudio->mStringDecoded, (void*)iosAudio->mOxyObject->mOxyCore);
NSString *tmpString = [NSString stringWithUTF8String:iosAudio->mStringDecoded];
iosAudio->mOxyObject->mDecodedString = [NSString stringWithUTF8String:iosAudio->mStringDecoded];
if (sizeStringDecoded > 0)
{
iosAudio->mOxyObject->mDecodedOK = 1;
NSLog(#"Decoded OK! %# ", tmpString);
[iosAudio->mObject performSelector:iosAudio->mSelector withObject:[NSNumber numberWithInt:1]];
}
else
{
iosAudio->mOxyObject->mDecodedOK = -1;
NSLog(#"END DECODING BAD! %# ", tmpString);
[iosAudio->mObject performSelector:iosAudio->mSelector withObject:[NSNumber numberWithInt:2]];
}
}
else
{
//no data found in this buffer
}
// release the malloc'ed data in the buffer we created earlier
free(bufferList.mBuffers[0].mData);
free(iosAudio->floatBuffer);
return noErr;
}
static OSStatus playbackCallback(void *inRefCon,
AudioUnitRenderActionFlags *ioActionFlags,
const AudioTimeStamp *inTimeStamp,
UInt32 inBusNumber,
UInt32 inNumberFrames,
AudioBufferList *ioData) {
// Notes: ioData contains buffers (may be more than one!)
// Fill them up as much as you can. Remember to set the size value in each buffer to match how
// much data is in the buffer.
for (int i=0; i < ioData->mNumberBuffers; i++)
{ // in practice we will only ever have 1 buffer, since audio format is mono
AudioBuffer buffer = ioData->mBuffers[i];
// NSLog(#" Buffer %d has %d channels and wants %d bytes of data.", i, buffer.mNumberChannels, buffer.mDataByteSize);
// copy temporary buffer data to output buffer
UInt32 size = min(buffer.mDataByteSize, [iosAudio tempBuffer].mDataByteSize); // dont copy more data than we have, or than fits
memcpy(buffer.mData, [iosAudio tempBuffer].mData, size);
buffer.mDataByteSize = size; // indicate how much data we wrote in the buffer
// uncomment to hear random noise
/*UInt16 *frameBuffer = buffer.mData;
for (int j = 0; j < inNumberFrames; j++)
frameBuffer[j] = rand();*/
// Play encoded buffer
if (iosAudio->mOxyObject->mEncoding > 0)
{
int sizeSamplesRead;
float audioBuffer[2048];
sizeSamplesRead = OXY_GetEncodedAudioBuffer(audioBuffer, (void*)iosAudio->mOxyObject->mOxyCore);
if (sizeSamplesRead == 0)
iosAudio->mOxyObject->mEncoding = 0;
SInt16 *frameBuffer = buffer.mData;
for(int j=0;j<sizeSamplesRead;j++)
{
frameBuffer[j] = audioBuffer[j]*32768.0;
}
}
else
{
SInt16 *frameBuffer = buffer.mData;
for (int j = 0; j < inNumberFrames; j++)
frameBuffer[j] = 0;
}
}
return noErr;
}
#implementation IosAudioController
#synthesize audioUnit, tempBuffer;
- (id) init {
self = [super init];
OSStatus status;
// Describe audio component
AudioComponentDescription desc;
desc.componentType = kAudioUnitType_Output;
desc.componentSubType = kAudioUnitSubType_RemoteIO;
desc.componentFlags = 0;
desc.componentFlagsMask = 0;
desc.componentManufacturer = kAudioUnitManufacturer_Apple;
// Get component
AudioComponent inputComponent = AudioComponentFindNext(NULL, &desc);
// Get audio units
status = AudioComponentInstanceNew(inputComponent, &audioUnit);
checkStatus(status);
// Enable IO for recording
UInt32 flag = 1;
status = AudioUnitSetProperty(audioUnit,
kAudioOutputUnitProperty_EnableIO,
kAudioUnitScope_Input,
kInputBus,
&flag,
sizeof(flag));
checkStatus(status);
// Enable IO for playback
status = AudioUnitSetProperty(audioUnit,
kAudioOutputUnitProperty_EnableIO,
kAudioUnitScope_Output,
kOutputBus,
&flag,
sizeof(flag));
checkStatus(status);
// Describe format
AudioStreamBasicDescription audioFormat;
audioFormat.mSampleRate = 44100.0;
audioFormat.mFormatID = kAudioFormatLinearPCM;
audioFormat.mFormatFlags = kAudioFormatFlagIsSignedInteger | kAudioFormatFlagIsPacked;
audioFormat.mFramesPerPacket = 1;
audioFormat.mChannelsPerFrame = 1;
audioFormat.mBitsPerChannel = 16;
audioFormat.mBytesPerPacket = 2;
audioFormat.mBytesPerFrame = 2;
// Apply format
status = AudioUnitSetProperty(audioUnit,
kAudioUnitProperty_StreamFormat,
kAudioUnitScope_Output,
kInputBus,
&audioFormat,
sizeof(audioFormat));
checkStatus(status);
status = AudioUnitSetProperty(audioUnit,
kAudioUnitProperty_StreamFormat,
kAudioUnitScope_Input,
kOutputBus,
&audioFormat,
sizeof(audioFormat));
checkStatus(status);
// Set input callback
AURenderCallbackStruct callbackStruct;
callbackStruct.inputProc = recordingCallback;
callbackStruct.inputProcRefCon = (__bridge void *)self;
status = AudioUnitSetProperty(audioUnit,
kAudioOutputUnitProperty_SetInputCallback,
kAudioUnitScope_Global,
kInputBus,
&callbackStruct,
sizeof(callbackStruct));
checkStatus(status);
// Set output callback
callbackStruct.inputProc = playbackCallback;
callbackStruct.inputProcRefCon = (__bridge void *)self;
status = AudioUnitSetProperty(audioUnit,
kAudioUnitProperty_SetRenderCallback,
kAudioUnitScope_Global,
kOutputBus,
&callbackStruct,
sizeof(callbackStruct));
checkStatus(status);
// Disable buffer allocation for the recorder (optional - do this if we want to pass in our own)
flag = 0;
status = AudioUnitSetProperty(audioUnit,
kAudioUnitProperty_ShouldAllocateBuffer,
kAudioUnitScope_Output,
kInputBus,
&flag,
sizeof(flag));
// Allocate our own buffers (1 channel, 16 bits per sample, thus 16 bits per frame, thus 2 bytes per frame).
// Practice learns the buffers used contain 512 frames, if this changes it will be fixed in processAudio.
tempBuffer.mNumberChannels = 1;
int size = 512;
#if (TARGET_OS_SIMULATOR)
size = 256; //TODO check this value!! depends on play/record callback buffer size
#else
size = 512; //TODO check this value!! depends on play/record callback buffer size
#endif
tempBuffer.mDataByteSize = size * 2;
tempBuffer.mData = malloc( size * 2);
// Initialise
status = AudioUnitInitialize(audioUnit);
checkStatus(status);
return self;
}
- (void) start {
OSStatus status = AudioOutputUnitStart(audioUnit);
checkStatus(status);
}
- (void) stop {
OSStatus status = AudioOutputUnitStop(audioUnit);
checkStatus(status);
}
- (void) processAudio: (AudioBufferList*) bufferList{
AudioBuffer sourceBuffer = bufferList->mBuffers[0];
// fix tempBuffer size if it's the wrong size
if (tempBuffer.mDataByteSize != sourceBuffer.mDataByteSize) {
free(tempBuffer.mData);
tempBuffer.mDataByteSize = sourceBuffer.mDataByteSize;
tempBuffer.mData = malloc(sourceBuffer.mDataByteSize);
}
// copy incoming audio data to temporary buffer
memcpy(tempBuffer.mData, bufferList->mBuffers[0].mData, bufferList->mBuffers[0].mDataByteSize);
}
- (void) dealloc {
AudioUnitUninitialize(audioUnit);
free(tempBuffer.mData);
}
- (void) setOxyObject: (OxyCore*) oxyObject
{
mOxyObject = oxyObject;
}
- (void) setListenCallback:(id)object withSelector:(SEL)selector
{
mObject = object;
mSelector = selector;
}
#end
One problem that I can see is that you are using 2 nested loops with the same variable for iteration. The first loop for (i = 0; i < 10000; ++i) and the second one for (i = 0; i < buffer_frames; i++), if buffer_frames >= 10000 - 1 the first loop will be executed once and exit, otherwise it will enter an infinite loop.
I have two more remarks regarding the following line:
buffer = static_cast<int16_t*>(malloc(buffer_frames * snd_pcm_format_width(format) / 8 * 2));
According to the API reference snd_pcm_format_width(format) returns the number of bits per sample. As you have 16 bits per sample and each frame contains only one sample, you should allocate buffer_frames * snd_pcm_format_width(format) / 8 bytes of memory (that 2 from your multiplication represents the number of channels which in your case is 1). Also, I suggest to change your buffer type to char* as it is the only type that is not prone to violating the strict aliasing rule. Thus, the line becomes:
static_cast<char*>(malloc(buffer_frames * (snd_pcm_format_width(format) / 8)));
and when you do the trick to change from short ints to float, the second for loop becomes:
int16_t* sint_buffer = buffer;
for (j = 0; j < buffer_frames; ++j){
float_buffer[j] = (float)sint_buffer[j]/32768.0;
// everything else goes here
}
I want to decode H.264 video from a collection of MPEG-2 Transport Stream packets but I am not clear what to pass to avcodec_decode_video2
The documentation says to pass "the input AVPacket containing the input buffer."
But what should be in the input buffer?
A PES packet will be spread across the payload portion of several TS packets, with NALU(s) inside the PES. So pass a TS fragment? The entire PES? PES payload only?
This Sample Code mentions:
BUT some other codecs (msmpeg4, mpeg4) are inherently frame based, so
you must call them with all the data for one frame exactly. You must
also initialize 'width' and 'height' before initializing them.
But I can find no info on what "all the data" means...
Passing a fragment of a TS packet payload is not working:
AVPacket avDecPkt;
av_init_packet(&avDecPkt);
avDecPkt.data = inbuf_ptr;
avDecPkt.size = esBufSize;
len = avcodec_decode_video2(mpDecoderContext, mpFrameDec, &got_picture, &avDecPkt);
if (len < 0)
{
printf(" TS PKT #%.0f. Error decoding frame #%04d [rc=%d '%s']\n",
tsPacket.pktNum, mDecodedFrameNum, len, av_make_error_string(errMsg, 128, len));
return;
}
output
[h264 # 0x81cd2a0] no frame!
TS PKT #2973. Error decoding frame #0001 [rc=-1094995529 'Invalid data found when processing input']
EDIT
Using the excellent hits from WLGfx, I made this simple program to try decoding TS packets. As input, I prepared a file containing only TS packets from the Video PID.
It feels close but I don't know how to set up the FormatContext. The code below segfaults at av_read_frame() (and internally at ret = s->iformat->read_packet(s, pkt)). s->iformat is zero.
Suggestions?
EDIT II - Sorry, for got post source code **
**EDIT III - Sample code updated to simulate reading TS PKT Queue
/*
* Test program for video decoder
*/
#include <stdio.h>
#include <stdio.h>
#include <stdlib.h>
#include <unistd.h>
extern "C" {
#ifdef __cplusplus
#define __STDC_CONSTANT_MACROS
#ifdef _STDINT_H
#undef _STDINT_H
#endif
#include <stdint.h>
#endif
}
extern "C" {
#include "libavcodec/avcodec.h"
#include "libavformat/avformat.h"
#include "libswscale/swscale.h"
#include "libavutil/imgutils.h"
#include "libavutil/opt.h"
}
class VideoDecoder
{
public:
VideoDecoder();
bool rcvTsPacket(AVPacket &inTsPacket);
private:
AVCodec *mpDecoder;
AVCodecContext *mpDecoderContext;
AVFrame *mpDecodedFrame;
AVFormatContext *mpFmtContext;
};
VideoDecoder::VideoDecoder()
{
av_register_all();
// FORMAT CONTEXT SETUP
mpFmtContext = avformat_alloc_context();
mpFmtContext->flags = AVFMT_NOFILE;
// ????? WHAT ELSE ???? //
// DECODER SETUP
mpDecoder = avcodec_find_decoder(AV_CODEC_ID_H264);
if (!mpDecoder)
{
printf("Could not load decoder\n");
exit(11);
}
mpDecoderContext = avcodec_alloc_context3(NULL);
if (avcodec_open2(mpDecoderContext, mpDecoder, NULL) < 0)
{
printf("Cannot open decoder context\n");
exit(1);
}
mpDecodedFrame = av_frame_alloc();
}
bool
VideoDecoder::rcvTsPacket(AVPacket &inTsPkt)
{
bool ret = true;
if ((av_read_frame(mpFmtContext, &inTsPkt)) < 0)
{
printf("Error in av_read_frame()\n");
ret = false;
}
else
{
// success. Decode the TS packet
int got;
int len = avcodec_decode_video2(mpDecoderContext, mpDecodedFrame, &got, &inTsPkt);
if (len < 0)
ret = false;
if (got)
printf("GOT A DECODED FRAME\n");
}
return ret;
}
int
main(int argc, char **argv)
{
if (argc != 2)
{
printf("Usage: %s tsInFile\n", argv[0]);
exit(1);
}
FILE *tsInFile = fopen(argv[1], "r");
if (!tsInFile)
{
perror("Could not open TS input file");
exit(2);
}
unsigned int tsPktNum = 0;
uint8_t tsBuffer[256];
AVPacket tsPkt;
av_init_packet(&tsPkt);
VideoDecoder vDecoder;
while (!feof(tsInFile))
{
tsPktNum++;
tsPkt.size = 188;
tsPkt.data = tsBuffer;
fread(tsPkt.data, 188, 1, tsInFile);
vDecoder.rcvTsPacket(tsPkt);
}
}
I've got some code snippets that might help you out as I've been working with MPEG-TS also.
Starting with my packet thread which checks each packet against the stream ID's which I've already found and got the codec contexts:
void *FFMPEG::thread_packet_function(void *arg) {
FFMPEG *ffmpeg = (FFMPEG*)arg;
for (int c = 0; c < MAX_PACKETS; c++)
ffmpeg->free_packets[c] = &ffmpeg->packet_list[c];
ffmpeg->packet_pos = MAX_PACKETS;
Audio.start_decoding();
Video.start_decoding();
Subtitle.start_decoding();
while (!ffmpeg->thread_quit) {
if (ffmpeg->packet_pos != 0 &&
Audio.okay_add_packet() &&
Video.okay_add_packet() &&
Subtitle.okay_add_packet()) {
pthread_mutex_lock(&ffmpeg->packet_mutex); // get free packet
AVPacket *pkt = ffmpeg->free_packets[--ffmpeg->packet_pos]; // pre decrement
pthread_mutex_unlock(&ffmpeg->packet_mutex);
if ((av_read_frame(ffmpeg->fContext, pkt)) >= 0) { // success
int id = pkt->stream_index;
if (id == ffmpeg->aud_stream.stream_id) Audio.add_packet(pkt);
else if (id == ffmpeg->vid_stream.stream_id) Video.add_packet(pkt);
else if (id == ffmpeg->sub_stream.stream_id) Subtitle.add_packet(pkt);
else { // unknown packet
av_packet_unref(pkt);
pthread_mutex_lock(&ffmpeg->packet_mutex); // put packet back
ffmpeg->free_packets[ffmpeg->packet_pos++] = pkt;
pthread_mutex_unlock(&ffmpeg->packet_mutex);
//LOGI("Dumping unknown packet, id %d", id);
}
} else {
av_packet_unref(pkt);
pthread_mutex_lock(&ffmpeg->packet_mutex); // put packet back
ffmpeg->free_packets[ffmpeg->packet_pos++] = pkt;
pthread_mutex_unlock(&ffmpeg->packet_mutex);
//LOGI("No packet read");
}
} else { // buffers full so yield
//LOGI("Packet reader on hold: Audio-%d, Video-%d, Subtitle-%d",
// Audio.packet_pos, Video.packet_pos, Subtitle.packet_pos);
usleep(1000);
//sched_yield();
}
}
return 0;
}
Each decoder for audio, video and subtitles have their own threads which receive the packets from the above thread in ring buffers. I've had to separate the decoders into their own threads because CPU usage was increasing when I started using the deinterlace filter.
My video decoder reads the packets from the buffers and when it has finished with the packet sends it back to be unref'd and can be used again. Balancing the packet buffers doesn't take that much time once everything is running.
Here's the snipped from my video decoder:
void *VideoManager::decoder(void *arg) {
LOGI("Video decoder started");
VideoManager *mgr = (VideoManager *)arg;
while (!ffmpeg.thread_quit) {
pthread_mutex_lock(&mgr->packet_mutex);
if (mgr->packet_pos != 0) {
// fetch first packet to decode
AVPacket *pkt = mgr->packets[0];
// shift list down one
for (int c = 1; c < mgr->packet_pos; c++) {
mgr->packets[c-1] = mgr->packets[c];
}
mgr->packet_pos--;
pthread_mutex_unlock(&mgr->packet_mutex); // finished with packets array
int got;
AVFrame *frame = ffmpeg.vid_stream.frame;
avcodec_decode_video2(ffmpeg.vid_stream.context, frame, &got, pkt);
ffmpeg.finished_with_packet(pkt);
if (got) {
#ifdef INTERLACE_ALL
if (!frame->interlaced_frame) mgr->add_av_frame(frame, 0);
else {
if (!mgr->filter_initialised) mgr->init_filter_graph(frame);
av_buffersrc_add_frame_flags(mgr->filter_src_ctx, frame, AV_BUFFERSRC_FLAG_KEEP_REF);
int c = 0;
while (true) {
AVFrame *filter_frame = ffmpeg.vid_stream.filter_frame;
int result = av_buffersink_get_frame(mgr->filter_sink_ctx, filter_frame);
if (result == AVERROR(EAGAIN) ||
result == AVERROR(AVERROR_EOF) ||
result < 0)
break;
mgr->add_av_frame(filter_frame, c++);
av_frame_unref(filter_frame);
}
//LOGI("Interlaced %d frames, decode %d, playback %d", c, mgr->decode_pos, mgr->playback_pos);
}
#elif defined(INTERLACE_HALF)
if (!frame->interlaced_frame) mgr->add_av_frame(frame, 0);
else {
if (!mgr->filter_initialised) mgr->init_filter_graph(frame);
av_buffersrc_add_frame_flags(mgr->filter_src_ctx, frame, AV_BUFFERSRC_FLAG_KEEP_REF);
int c = 0;
while (true) {
AVFrame *filter_frame = ffmpeg.vid_stream.filter_frame;
int result = av_buffersink_get_frame(mgr->filter_sink_ctx, filter_frame);
if (result == AVERROR(EAGAIN) ||
result == AVERROR(AVERROR_EOF) ||
result < 0)
break;
mgr->add_av_frame(filter_frame, c++);
av_frame_unref(filter_frame);
}
//LOGI("Interlaced %d frames, decode %d, playback %d", c, mgr->decode_pos, mgr->playback_pos);
}
#else
mgr->add_av_frame(frame, 0);
#endif
}
//LOGI("decoded video packet");
} else {
pthread_mutex_unlock(&mgr->packet_mutex);
}
}
LOGI("Video decoder ended");
}
As you can see, I'm using a mutex when passing packets back and forth.
Once a frame has been got I just copy the YUV buffers from the frame for later use into another buffer list. I don't convert the YUV, I use a shader which converts the YUV to RGB on the GPU.
The next snippet adds my decoded frame to my buffer list. This may help understand how to deal with the data.
void VideoManager::add_av_frame(AVFrame *frame, int field_num) {
int y_linesize = frame->linesize[0];
int u_linesize = frame->linesize[1];
int hgt = frame->height;
int y_buffsize = y_linesize * hgt;
int u_buffsize = u_linesize * hgt / 2;
int buffsize = y_buffsize + u_buffsize + u_buffsize;
VideoBuffer *buffer = &buffers[decode_pos];
if (ffmpeg.is_network && playback_pos == decode_pos) { // patched 25/10/16 wlgfx
buffer->used = false;
if (!buffer->data) buffer->data = (char*)mem.alloc(buffsize);
if (!buffer->data) {
LOGI("Dropped frame, allocation error");
return;
}
} else if (playback_pos == decode_pos) {
LOGI("Dropped frame, ran out of decoder frame buffers");
return;
} else if (!buffer->data) {
buffer->data = (char*)mem.alloc(buffsize);
if (!buffer->data) {
LOGI("Dropped frame, allocation error.");
return;
}
}
buffer->y_frame = buffer->data;
buffer->u_frame = buffer->y_frame + y_buffsize;
buffer->v_frame = buffer->y_frame + y_buffsize + u_buffsize;
buffer->wid = frame->width;
buffer->hgt = hgt;
buffer->y_linesize = y_linesize;
buffer->u_linesize = u_linesize;
int64_t pts = av_frame_get_best_effort_timestamp(frame);
buffer->pts = pts;
buffer->buffer_size = buffsize;
double field_add = av_q2d(ffmpeg.vid_stream.context->time_base) * field_num;
buffer->frame_time = av_q2d(ts_stream) * pts + field_add;
memcpy(buffer->y_frame, frame->data[0], (size_t) (buffer->y_linesize * buffer->hgt));
memcpy(buffer->u_frame, frame->data[1], (size_t) (buffer->u_linesize * buffer->hgt / 2));
memcpy(buffer->v_frame, frame->data[2], (size_t) (buffer->u_linesize * buffer->hgt / 2));
buffer->used = true;
decode_pos = (++decode_pos) % MAX_VID_BUFFERS;
//if (field_num == 0) LOGI("Video %.2f, %d - %d",
// buffer->frame_time - Audio.pts_start_time, decode_pos, playback_pos);
}
If there's anything else that I may be able to help with just give me a shout. :-)
EDIT:
The snippet how I open my video stream context which automatically determines the codec, whether it is h264, mpeg2, or another:
void FFMPEG::open_video_stream() {
vid_stream.stream_id = av_find_best_stream(fContext, AVMEDIA_TYPE_VIDEO,
-1, -1, &vid_stream.codec, 0);
if (vid_stream.stream_id == -1) return;
vid_stream.context = fContext->streams[vid_stream.stream_id]->codec;
if (!vid_stream.codec || avcodec_open2(vid_stream.context,
vid_stream.codec, NULL) < 0) {
vid_stream.stream_id = -1;
return;
}
vid_stream.frame = av_frame_alloc();
vid_stream.filter_frame = av_frame_alloc();
}
EDIT2:
This is how I've opened the input stream, whether it be file or URL. The AVFormatContext is the main context for the stream.
bool FFMPEG::start_stream(char *url_, float xtrim, float ytrim, int gain) {
aud_stream.stream_id = -1;
vid_stream.stream_id = -1;
sub_stream.stream_id = -1;
this->url = url_;
this->xtrim = xtrim;
this->ytrim = ytrim;
Audio.volume = gain;
Audio.init();
Video.init();
fContext = avformat_alloc_context();
if ((avformat_open_input(&fContext, url_, NULL, NULL)) != 0) {
stop_stream();
return false;
}
if ((avformat_find_stream_info(fContext, NULL)) < 0) {
stop_stream();
return false;
}
// network stream will overwrite packets if buffer is full
is_network = url.substr(0, 4) == "udp:" ||
url.substr(0, 4) == "rtp:" ||
url.substr(0, 5) == "rtsp:" ||
url.substr(0, 5) == "http:"; // added for wifi broadcasting ability
// determine if stream is audio only
is_mp3 = url.substr(url.size() - 4) == ".mp3";
LOGI("Stream: %s", url_);
if (!open_audio_stream()) {
stop_stream();
return false;
}
if (is_mp3) {
vid_stream.stream_id = -1;
sub_stream.stream_id = -1;
} else {
open_video_stream();
open_subtitle_stream();
if (vid_stream.stream_id == -1) { // switch to audio only
close_subtitle_stream();
is_mp3 = true;
}
}
LOGI("Audio: %d, Video: %d, Subtitle: %d",
aud_stream.stream_id,
vid_stream.stream_id,
sub_stream.stream_id);
if (aud_stream.stream_id != -1) {
LOGD("Audio stream time_base {%d, %d}",
aud_stream.context->time_base.num,
aud_stream.context->time_base.den);
}
if (vid_stream.stream_id != -1) {
LOGD("Video stream time_base {%d, %d}",
vid_stream.context->time_base.num,
vid_stream.context->time_base.den);
}
LOGI("Starting packet and decode threads");
thread_quit = false;
pthread_create(&thread_packet, NULL, &FFMPEG::thread_packet_function, this);
Display.set_overlay_timout(3.0);
return true;
}
EDIT: (constructing an AVPacket)
Construct an AVPacket to send to the decoder...
AVPacket packet;
av_init_packet(&packet);
packet.data = myTSpacketdata; // pointer to the TS packet
packet.size = 188;
You should be able to reuse the packet. And it might need unref'ing.
You must first use the avcodec library to get the compressed frames out of the file. Then you can decode them using avcodec_decode_video2. look at this tutorial http://dranger.com/ffmpeg/
I am trying to play a wave file in RHEL6 using alsa library calls in my C Code in Qt. I am reading the wave file ("t15.wav") in a buffer(wave_buffer). The wave header has been stripped off since the alsa library requires raw PCM samples to be played. Further I have set up the PCM hardware & Software params using 'snd_pcm_hw_params(PCM, params)' & 'snd_pcm_sw_params_current(PCM, swparams)' and many other calls. I am writing the PCM samples on the PCM handle using 'snd_pcm_writei' command. For this purpose i am reading a chunk(32 or 1024 or 2048 or 4096 or 8192 bytes) of data from the wave_buffer and sending it for playing using the snd_pcm_writei command. If I choose a small chunk the audio quality falters but playback is uninterrupted. If I use a bigger chunk(greater than 4096 i.e. 8192) I get perfect audio quality but it is interrupted( When next chunk of data is required for playing ). My constraint is that I can have access to data in chunks only and not as a file or entire buffer. Can anybody help me in removing the interruptions in playing the wave data so that I can get uninterrupted audio playback. Following is my code :
The two variables buffer_time & period_time return the period size which is the size of chunk.
If buffer_time = 5000 & period_time=1000 the period_size returned by alsa library is 32 bytes //audio quality falters but no interruptions
If buffer_time = 500000 & period_time=100000 the period_size returned by alsa library is 8192 bytes //good audio quality but interrupted
Tuning these parameters seems useless as I have wasted a lot of time doing this. Please help me get through this problem-----
Stucture of Wave File :
Sample Rate : 44100
Bits per Sample : 16
Channels : 2
mainwindow.h----
#ifndef MAINWINDOW_H
#define MAINWINDOW_H
#include <QMainWindow>
#include <alsa/asoundlib.h>
#define BLOCKSIZE 44100 * 2 * 2 // Sample Rate * Channels * Byte per Sample(Bits per sample / 8)
namespace Ui {
class MainWindow;
}
class MainWindow : public QMainWindow
{
Q_OBJECT
public:
explicit MainWindow(QWidget *parent = 0);
int init_alsa();
int play_snd();
~MainWindow();
snd_pcm_t *PCM;
snd_pcm_sframes_t delayp;
snd_pcm_sframes_t availp;
snd_pcm_sw_params_t *swparams;
snd_pcm_hw_params_t *params;
static snd_pcm_sframes_t period_size;
static snd_pcm_sframes_t buffer_size;
unsigned char wave_buffer[900000];
unsigned char play_buffer[BLOCKSIZE];
int filesize;
FILE *fp;
private:
Ui::MainWindow *ui;
};
#endif // MAINWINDOW_H
mainwindow.cpp---
#include "mainwindow.h"
#include "ui_mainwindow.h"
snd_pcm_sframes_t MainWindow::period_size;
snd_pcm_sframes_t MainWindow::buffer_size;
MainWindow::MainWindow(QWidget *parent) :
QMainWindow(parent),
ui(new Ui::MainWindow)
{
ui->setupUi(this);
if((fp = fopen("t15.wav","rb"))==NULL)
printf("Error Opening File");
fseek(fp,0L,SEEK_END);
filesize = ftell(fp)-44;
fseek(fp,0L,SEEK_SET);
fseek(fp,44,SEEK_SET);
fread(wave_buffer,filesize,1,fp);
fclose(fp);
delayp = 0;
init_alsa();
play_snd();
}
MainWindow::~MainWindow()
{
delete ui;
}
int MainWindow::init_alsa()
{
unsigned int rate = 44100;
int err,dir;
unsigned int rrate = 44100;
snd_pcm_uframes_t size;
static unsigned int buffer_time = 500000;
static unsigned int period_time = 100000;
static int period_event = 0;
if ((err=snd_pcm_open(&PCM,"plughw:0,0",SND_PCM_STREAM_PLAYBACK, 0)) < 0)
{
fprintf(stderr, "Can't use sound: %s\n", snd_strerror(err));
return err;
}
snd_pcm_hw_params_alloca(¶ms);
snd_pcm_sw_params_alloca(&swparams);
//snd_pcm_nonblock(PCM,0);
/* choose all parameters */
err = snd_pcm_hw_params_any(PCM, params);
if (err < 0) {
printf("Broken configuration for playback: no configurations available: %s\n", snd_strerror(err));
return err;
}
/* set hardware resampling */
err = snd_pcm_hw_params_set_rate_resample(PCM, params, 1);
if (err < 0) {
printf("Resampling setup failed for playback: %s\n", snd_strerror(err));
return err;
}
/* set the interleaved read/write format */
err = snd_pcm_hw_params_set_access(PCM, params, SND_PCM_ACCESS_RW_INTERLEAVED);
if (err < 0) {
printf("Access type not available for playback: %s\n", snd_strerror(err));
return err;
}
/* set the sample format */
err = snd_pcm_hw_params_set_format(PCM, params, SND_PCM_FORMAT_S16_LE);
if (err < 0) {
printf("Sample format not available for playback: %s\n", snd_strerror(err));
return err;
}
/* set the count of channels */
err = snd_pcm_hw_params_set_channels(PCM, params, 2);
if (err < 0) {
printf("Channels count (%i) not available for playbacks: %s\n", 2, snd_strerror(err));
return err;
}
/* set the stream rate */
rrate = rate;
err = snd_pcm_hw_params_set_rate_near(PCM, params, &rrate, 0);
if (err < 0) {
printf("Rate %iHz not available for playback: %s\n", 44100, snd_strerror(err));
return err;
}
if (rrate != 44100) {
printf("Rate doesn't match (requested %iHz, get %iHz)\n", rrate, err);
return -EINVAL;
}
/* set the buffer time */
err = snd_pcm_hw_params_set_buffer_time_near(PCM, params, &buffer_time, &dir);
if (err < 0) {
printf("Unable to set buffer time %i for playback: %s\n", buffer_time, snd_strerror(err));
return err;
}
err = snd_pcm_hw_params_get_buffer_size(params, &size);
if (err < 0) {
printf("Unable to get buffer size for playback: %s\n", snd_strerror(err));
return err;
}
buffer_size = size;
/* set the period time */
err = snd_pcm_hw_params_set_period_time_near(PCM, params, &period_time, &dir);
if (err < 0) {
printf("Unable to set period time %i for playback: %s\n", period_time, snd_strerror(err));
return err;
}
err = snd_pcm_hw_params_get_period_size(params, &size, &dir);
if (err < 0) {
printf("Unable to get period size for playback: %s\n", snd_strerror(err));
return err;
}
period_size = size;
/* write the parameters to device */
err = snd_pcm_hw_params(PCM, params);
if (err < 0) {
printf("Unable to set hw params for playback: %s\n", snd_strerror(err));
return err;
}
printf("Size = %ld",period_size);
snd_pcm_sw_params_current(PCM, swparams); /* get the current swparams */
/* start the transfer when the buffer is almost full: */
/* (buffer_size / avail_min) * avail_min */
snd_pcm_sw_params_set_start_threshold(PCM, swparams, (buffer_size / period_size) * period_size);
/* allow the transfer when at least period_size samples can be processed */
/* or disable this mechanism when period event is enabled (aka interrupt like style processing) */
snd_pcm_sw_params_set_avail_min(PCM, swparams, period_event ? buffer_size : period_size);
snd_pcm_sw_params(PCM, swparams);/* write the parameters to the playback device */
return 1;
}
int MainWindow::play_snd()
{
int curr_pos = 0;
int buff_size = 0;
long val = 0;
while(curr_pos < filesize)
{
if(filesize-curr_pos >= period_size)
{
memcpy(play_buffer,wave_buffer+curr_pos,period_size);
buff_size = period_size;
curr_pos += buff_size;
}
else
{
memcpy(play_buffer,wave_buffer+curr_pos,filesize-curr_pos);
buff_size = filesize - curr_pos;
curr_pos += buff_size;
}
int i=1;
unsigned char *ptr = play_buffer;
while(buff_size > 0)
{
val = snd_pcm_writei(PCM,&play_buffer,buff_size);
if (val == -EAGAIN)
continue;
ptr += val * 2;
buff_size -= val;
}
}
return 0;
}
I have a similar C Code of alsa library which generates sine wave samples at runtime and plays them using same snd_pcm_writei command and it plays perfectly without any interruptions....This is the alsa library code---
/*
* This small demo sends a simple sinusoidal wave to your speakers.
*/
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <sched.h>
#include <errno.h>
#include <getopt.h>
#include "alsa/asoundlib.h"
#include <sys/time.h>
#include <math.h>
static char *device = "plughw:0,0"; /* playback device */
static snd_pcm_format_t format = SND_PCM_FORMAT_S16_LE; /* sample format */
static unsigned int rate = 44100; /* stream rate */
static unsigned int channels = 2; /* count of channels */
static unsigned int buffer_time = 5000; /* ring buffer length in us */
static unsigned int period_time = 1000; /* period time in us */
static double freq = 440; /* sinusoidal wave frequency in Hz */
static int resample = 1; /* enable alsa-lib resampling */
static int period_event = 0; /* produce poll event after each period */
static snd_pcm_sframes_t buffer_size;
static snd_pcm_sframes_t period_size;
static snd_output_t *output = NULL;
snd_pcm_sframes_t delayp;
snd_pcm_sframes_t availp;
static void generate_sine(const snd_pcm_channel_area_t *areas,
snd_pcm_uframes_t offset,
int count, double *_phase)
{
static double max_phase = 2. * M_PI;
double phase = *_phase;
double step = max_phase*freq/(double)rate;
unsigned char *samples[channels];
int steps[channels];
unsigned int chn;
int format_bits = snd_pcm_format_width(format);
unsigned int maxval = (1 << (format_bits - 1)) - 1;
int bps = format_bits / 8; /* bytes per sample */
int phys_bps = snd_pcm_format_physical_width(format) / 8;
int big_endian = snd_pcm_format_big_endian(format) == 1;
int to_unsigned = snd_pcm_format_unsigned(format) == 1;
int is_float = (format == SND_PCM_FORMAT_FLOAT_LE ||
format == SND_PCM_FORMAT_FLOAT_BE);
/* verify and prepare the contents of areas */
for (chn = 0; chn < channels; chn++) {
samples[chn] = /*(signed short *)*/(((unsigned char *)areas[chn].addr) + (areas[chn].first / 8));
steps[chn] = areas[chn].step / 8;
samples[chn] += offset * steps[chn];
}
/* fill the channel areas */
while (count-- > 0) {
union {
float f;
int i;
} fval;
int res, i;
if (is_float)
{
fval.f = sin(phase) * maxval;
res = fval.i;
}
else
res = sin(phase) * maxval;
if (to_unsigned)
res ^= 1U << (format_bits - 1);
for (chn = 0; chn < channels; chn++) {
/* Generate data in native endian format */
if (big_endian) {
for (i = 0; i < bps; i++)
*(samples[chn] + phys_bps - 1 - i) = (res >> i * 8) & 0xff;
} else {
for (i = 0; i < bps; i++)
*(samples[chn] + i) = (res >> i * 8) & 0xff;
}
samples[chn] += steps[chn];
}
phase += step;
if (phase >= max_phase)
phase -= max_phase;
}
*_phase = phase;
}
static int set_hwparams(snd_pcm_t *handle, snd_pcm_hw_params_t *params, snd_pcm_access_t access)
{
unsigned int rrate;
snd_pcm_uframes_t size;
int dir;
snd_pcm_hw_params_any(handle, params); /* choose all parameters */
snd_pcm_hw_params_set_rate_resample(handle, params, resample);/* set hardware resampling */
snd_pcm_hw_params_set_access(handle, params, access); /* set the interleaved read/write format */
snd_pcm_hw_params_set_format(handle, params, format); /* set the sample format */
snd_pcm_hw_params_set_channels(handle, params, channels); /* set the count of channels */
rrate = rate; /* set the stream rate */
snd_pcm_hw_params_set_rate_near(handle, params, &rrate, 0);
snd_pcm_hw_params_set_buffer_time_near(handle, params, &buffer_time, &dir);/* set the buffer time */
snd_pcm_hw_params_get_buffer_size(params, &size);
buffer_size = size;
snd_pcm_hw_params_set_period_time_near(handle, params, &period_time, &dir);/* set the period time */
snd_pcm_hw_params_get_period_size(params, &size, &dir);
period_size = size;
snd_pcm_hw_params(handle, params); /* write the parameters to device */
return 0;
}
static int set_swparams(snd_pcm_t *handle, snd_pcm_sw_params_t *swparams)
{
snd_pcm_sw_params_current(handle, swparams); /* get the current swparams */
/* start the transfer when the buffer is almost full: */
/* (buffer_size / avail_min) * avail_min */
snd_pcm_sw_params_set_start_threshold(handle, swparams, (buffer_size / period_size) * period_size);
/* allow the transfer when at least period_size samples can be processed */
/* or disable this mechanism when period event is enabled (aka interrupt like style processing) */
snd_pcm_sw_params_set_avail_min(handle, swparams, period_event ? buffer_size : period_size);
snd_pcm_sw_params(handle, swparams);/* write the parameters to the playback device */
return 0;
}
/*
* Transfer method - write only
*/
static int write_loop(snd_pcm_t *handle, signed short *samples, snd_pcm_channel_area_t *areas)
{
double phase = 0;
signed short *ptr;
int err, cptr;
int i=0;
printf("Period Size = %ld",period_size);
while (1) {
fflush(stdout);
generate_sine(areas, 0, period_size, &phase);
ptr = samples;
cptr = period_size;
i=1;
while (cptr > 0) {
err = snd_pcm_writei(handle, ptr, cptr);
snd_pcm_avail_delay(handle,&availp,&delayp);
printf("available frames =%ld delay = %ld i = %d\n",availp,delayp,i);
if (err == -EAGAIN)
continue;
ptr += err * channels;
cptr -= err;
i++;
}
}
}
/*
* Transfer method - asynchronous notification
*/
int main()
{
snd_pcm_t *handle;
snd_pcm_hw_params_t *hwparams;
snd_pcm_sw_params_t *swparams;
signed short *samples;
unsigned int chn;
snd_pcm_channel_area_t *areas;
snd_pcm_hw_params_alloca(&hwparams);
snd_pcm_sw_params_alloca(&swparams);
snd_output_stdio_attach(&output, stdout, 0);
printf("Playback device is %s\n", device);
printf("Stream parameters are %iHz, %s, %i channels\n", rate, snd_pcm_format_name(format), channels);
printf("Sine wave rate is %.4fHz\n", freq);
snd_pcm_open(&handle, device, SND_PCM_STREAM_PLAYBACK, 0);
set_hwparams(handle, hwparams, SND_PCM_ACCESS_RW_INTERLEAVED);
set_swparams(handle, swparams);
samples = malloc((period_size * channels * snd_pcm_format_physical_width(format)) / 8);
areas = calloc(channels, sizeof(snd_pcm_channel_area_t));
for (chn = 0; chn < channels; chn++) {
areas[chn].addr = samples;
areas[chn].first = chn * snd_pcm_format_physical_width(format);
areas[chn].step = channels * snd_pcm_format_physical_width(format);
}
write_loop(handle, samples, areas);
free(areas);
free(samples);
snd_pcm_close(handle);
return 0;
}
I solved the problem by altering the length argument of snd_pcm_writei...perviously i was giving it equal to the data contained in play_buffer...now i changed it to "buff_size/4" and the audio is playing perfectly without breaks. Actually it is the size after which the system should start buffering for new pcm samples as per my understanding. Previously it was buffering after playing the entire length buff_size and that resulted in breaks in audio output...
When I decode frames from avi file and then decode them in x264 and save to mp4 file, the fps of the output file is always 12,800. Therefore the file is played very fast. But, when I save the encoded in h264 frames in avi format and not mp4, so the fps is as I wanted - 25.
What could be the problem?
Here the code I wrote in VS2010:
#include "stdafx.h"
#include "inttypes.h"
extern "C" {
#include "libavcodec/avcodec.h"
#include "libavformat/avformat.h"
#include "libavutil/avutil.h"
#include <libswscale/swscale.h>
#include <libavutil/opt.h>
#include <libswscale/swscale.h>
#include <libavutil/imgutils.h>
}
#include <iostream>
using namespace std;
int main(int argc, char* argv[])
{
const char* inFileName = "C:\\000227_C1_GAME.avi";
const char* outFileName = "c:\\test.avi";
const char* outFileType = "avi";
av_register_all();
AVFormatContext* inContainer = NULL;
if(avformat_open_input(&inContainer, inFileName, NULL, NULL) < 0)
exit(1);
if(avformat_find_stream_info(inContainer, NULL) < 0)
exit(1);
// Find video stream
int videoStreamIndex = -1;
for (unsigned int i = 0; i < inContainer->nb_streams; ++i)
{
if (inContainer->streams[i] && inContainer->streams[i]->codec &&
inContainer->streams[i]->codec->codec_type == AVMEDIA_TYPE_VIDEO)
{
videoStreamIndex = i;
break;
}
}
if (videoStreamIndex == -1) exit(1);
AVFormatContext* outContainer = NULL;
if(avformat_alloc_output_context2(&outContainer, NULL, outFileType, outFileName) < 0)
exit(1);
// ----------------------------
// Decoder
// ----------------------------
AVStream const *const inStream = inContainer->streams[videoStreamIndex];
AVCodec *const decoder = avcodec_find_decoder(inStream->codec->codec_id);
if(!decoder)
exit(1);
if(avcodec_open2(inStream->codec, decoder, NULL) < 0)
exit(1);
// ----------------------------
// Encoder
// ----------------------------
AVCodec *encoder = avcodec_find_encoder(AV_CODEC_ID_H264);
if(!encoder)
exit(1);
AVStream *outStream = avformat_new_stream(outContainer, encoder);
if(!outStream)
exit(1);
avcodec_get_context_defaults3(outStream->codec, encoder);
// Construct encoder
if(outContainer->oformat->flags & AVFMT_GLOBALHEADER)
outStream->codec->flags |= CODEC_FLAG_GLOBAL_HEADER;
outStream->codec->coder_type = AVMEDIA_TYPE_VIDEO;
outStream->codec->pix_fmt = AV_PIX_FMT_YUV420P;
outStream->codec->width = inStream->codec->width;
outStream->codec->height = inStream->codec->height;
outStream->codec->codec_id = encoder->id;
outStream->codec->bit_rate = 500000;
//outStream->codec->rc_min_rate = 600000;
//outStream->codec->rc_max_rate = 800000;
outStream->codec->time_base.den = 25;
outStream->codec->time_base.num = 1;
outStream->codec->gop_size = 250; // Keyframe interval(=GOP length). Determines maximum distance distance between I-frames
outStream->codec->keyint_min = 25; // minimum GOP size
outStream->codec->max_b_frames = 3;//16; // maximum number of B-frames between non-B-frames
outStream->codec->b_frame_strategy = 1; // decides the best number of B-frames to use. Default mode in x264.
outStream->codec->scenechange_threshold = 40;
outStream->codec->refs = 6; // abillity to reference frames other than the one immediately prior to the current frame. specify how many references can be used.
outStream->codec->qmin = 0;//10;
outStream->codec->qmax = 69;//51;
outStream->codec->qcompress = 0.6;
outStream->codec->max_qdiff = 4;
outStream->codec->i_quant_factor = 1.4;//0.71;
outStream->codec->refs=1;//3;
outStream->codec->chromaoffset = -2;
outStream->codec->thread_count = 1;
outStream->codec->trellis = 1;
outStream->codec->me_range = 16;
outStream->codec->me_method = ME_HEX; //hex
outStream->codec->flags2 |= CODEC_FLAG2_FAST;
outStream->codec->coder_type = 1;
if(outStream->codec->codec_id == AV_CODEC_ID_H264)
{
av_opt_set(outStream->codec->priv_data, "preset", "slow", 0);
}
// Open encoder
if(avcodec_open2(outStream->codec, encoder, NULL) < 0)
exit(1);
// Open output container
if(avio_open(&outContainer->pb, outFileName, AVIO_FLAG_WRITE) < 0)
exit(1);
//close_o
AVFrame *decodedFrame = avcodec_alloc_frame();
if(!decodedFrame)
exit(1);
AVFrame *encodeFrame = avcodec_alloc_frame();
if(!encodeFrame)
exit(1);
encodeFrame->format = outStream->codec->pix_fmt;
encodeFrame->width = outStream->codec->width;
encodeFrame->height = outStream->codec->height;
if(av_image_alloc(encodeFrame->data, encodeFrame->linesize,
outStream->codec->width, outStream->codec->height,
outStream->codec->pix_fmt, 1) < 0)
exit(1);
av_dump_format(inContainer, 0, inFileName,0);
//Write header to ouput container
avformat_write_header(outContainer, NULL);
AVPacket decodePacket, encodedPacket;
int got_frame, len;
while(av_read_frame(inContainer, &decodePacket)>=0)
{
if (decodePacket.stream_index == videoStreamIndex)
{
len = avcodec_decode_video2(inStream->codec, decodedFrame, &got_frame, &decodePacket);
if(len < 0)
exit(1);
if(got_frame)
{
av_init_packet(&encodedPacket);
encodedPacket.data = NULL;
encodedPacket.size = 0;
if(avcodec_encode_video2(outStream->codec, &encodedPacket, decodedFrame, &got_frame) < 0)
exit(1);
if(got_frame)
{
if (outStream->codec->coded_frame->key_frame)
encodedPacket.flags |= AV_PKT_FLAG_KEY;
encodedPacket.stream_index = outStream->index;
if(av_interleaved_write_frame(outContainer, &encodedPacket) < 0)
exit(1);
av_free_packet(&encodedPacket);
}
}
}
av_free_packet(&decodePacket);
}
av_write_trailer(outContainer);
avio_close(outContainer->pb);
avcodec_free_frame(&encodeFrame);
avcodec_free_frame(&decodedFrame);
avformat_free_context(outContainer);
av_close_input_file(inContainer);
return 0;
}
The problem was with PTS and DTS of the packet. Before writing the packet to output( before av_interleaved_write_frame command) set PTS and DTS like this
if (encodedPacket.pts != AV_NOPTS_VALUE)
encodedPacket.pts = av_rescale_q(encodedPacket.pts, outStream->codec->time_base, outStream->time_base);
if (encodedPacket.dts != AV_NOPTS_VALUE)
encodedPacket.dts = av_rescale_q(encodedPacket.dts, outStream->codec->time_base, outStream->time_base);
I am processing video with opencv, but at the same time I need to play audio and simply control it, like loud or current frame number.
I think I should create a parallel process with ffmpeg, but I don't know how to do so. Can you explain what to do?
Or do you know another solution?
I think ffmpeg should be used to play audio and SDL for video in this case.
After opening the file with OpenCV and processing the frame, you can use OpenCV -> SDL to display it while retrieving the audio frames through ffmpeg and playing them with SDL.
Here is a nice collection of ffmpeg/SDL tutorials!
I also found a nice post that shows how to capture frames from a video file using ffmpeg, store them in OpenCV cv::Mat and display the result in a OpenCV window. But this way you can't play audio since OpenCV doesn't deal with that.
You might be interested in reading this post as well: How to avoid a growing delay with ffmpeg between sound and raw video data ?
EDIT:
I spent the last 4hrs coding a prototype to demonstrate how it's done. This demo reads video frames through OpenCV (so you can process them) and audio through ffmpeg, and SDL is used to play both! There are 2 limitations in this demo you must be aware: 1 - it assumes you are working with an OpenCV image packed as BGR (24bits), and 2 - audio and video are not being sync! Yes, I left have some work for you to do (yeeeey). But don't panic, page 6 has some ideas!
It's important to sync audio and video because you will be doing some processing on the frames, and that will certainly make the video and audio go out of sync real fast since they are being played independently of each other.
The ffmpeg tutorials I suggested above are very very important to understand the code, a lot of code from this demo came from there. They show how to deal with SDL, and how to read packets of audio/video streams.
#include <highgui.h>
#include <cv.h>
extern "C"
{
#include <SDL.h>
#include <SDL_thread.h>
#include <avcodec.h>
#include <avformat.h>
}
#include <iostream>
#include <stdio.h>
//#include <malloc.h>
using namespace cv;
#define SDL_AUDIO_BUFFER_SIZE 1024
typedef struct PacketQueue
{
AVPacketList *first_pkt, *last_pkt;
int nb_packets;
int size;
SDL_mutex *mutex;
SDL_cond *cond;
} PacketQueue;
PacketQueue audioq;
int audioStream = -1;
int videoStream = -1;
int quit = 0;
SDL_Surface* screen = NULL;
SDL_Surface* surface = NULL;
AVFormatContext* pFormatCtx = NULL;
AVCodecContext* aCodecCtx = NULL;
AVCodecContext* pCodecCtx = NULL;
void show_frame(IplImage* img)
{
if (!screen)
{
screen = SDL_SetVideoMode(img->width, img->height, 0, 0);
if (!screen)
{
fprintf(stderr, "SDL: could not set video mode - exiting\n");
exit(1);
}
}
// Assuming IplImage packed as BGR 24bits
SDL_Surface* surface = SDL_CreateRGBSurfaceFrom((void*)img->imageData,
img->width,
img->height,
img->depth * img->nChannels,
img->widthStep,
0xff0000, 0x00ff00, 0x0000ff, 0
);
SDL_BlitSurface(surface, 0, screen, 0);
SDL_Flip(screen);
}
void packet_queue_init(PacketQueue *q)
{
memset(q, 0, sizeof(PacketQueue));
q->mutex = SDL_CreateMutex();
q->cond = SDL_CreateCond();
}
int packet_queue_put(PacketQueue *q, AVPacket *pkt)
{
AVPacketList *pkt1;
if (av_dup_packet(pkt) < 0)
{
return -1;
}
//pkt1 = (AVPacketList*) av_malloc(sizeof(AVPacketList));
pkt1 = (AVPacketList*) malloc(sizeof(AVPacketList));
if (!pkt1) return -1;
pkt1->pkt = *pkt;
pkt1->next = NULL;
SDL_LockMutex(q->mutex);
if (!q->last_pkt)
q->first_pkt = pkt1;
else
q->last_pkt->next = pkt1;
q->last_pkt = pkt1;
q->nb_packets++;
q->size += pkt1->pkt.size;
SDL_CondSignal(q->cond);
SDL_UnlockMutex(q->mutex);
return 0;
}
static int packet_queue_get(PacketQueue *q, AVPacket *pkt, int block)
{
AVPacketList *pkt1;
int ret;
SDL_LockMutex(q->mutex);
for (;;)
{
if( quit)
{
ret = -1;
break;
}
pkt1 = q->first_pkt;
if (pkt1)
{
q->first_pkt = pkt1->next;
if (!q->first_pkt)
q->last_pkt = NULL;
q->nb_packets--;
q->size -= pkt1->pkt.size;
*pkt = pkt1->pkt;
//av_free(pkt1);
free(pkt1);
ret = 1;
break;
}
else if (!block)
{
ret = 0;
break;
}
else
{
SDL_CondWait(q->cond, q->mutex);
}
}
SDL_UnlockMutex(q->mutex);
return ret;
}
int audio_decode_frame(AVCodecContext *aCodecCtx, uint8_t *audio_buf, int buf_size)
{
static AVPacket pkt;
static uint8_t *audio_pkt_data = NULL;
static int audio_pkt_size = 0;
int len1, data_size;
for (;;)
{
while (audio_pkt_size > 0)
{
data_size = buf_size;
len1 = avcodec_decode_audio2(aCodecCtx, (int16_t*)audio_buf, &data_size,
audio_pkt_data, audio_pkt_size);
if (len1 < 0)
{
/* if error, skip frame */
audio_pkt_size = 0;
break;
}
audio_pkt_data += len1;
audio_pkt_size -= len1;
if (data_size <= 0)
{
/* No data yet, get more frames */
continue;
}
/* We have data, return it and come back for more later */
return data_size;
}
if (pkt.data)
av_free_packet(&pkt);
if (quit) return -1;
if (packet_queue_get(&audioq, &pkt, 1) < 0) return -1;
audio_pkt_data = pkt.data;
audio_pkt_size = pkt.size;
}
}
void audio_callback(void *userdata, Uint8 *stream, int len)
{
AVCodecContext *aCodecCtx = (AVCodecContext *)userdata;
int len1, audio_size;
static uint8_t audio_buf[(AVCODEC_MAX_AUDIO_FRAME_SIZE * 3) / 2];
static unsigned int audio_buf_size = 0;
static unsigned int audio_buf_index = 0;
while (len > 0)
{
if (audio_buf_index >= audio_buf_size)
{
/* We have already sent all our data; get more */
audio_size = audio_decode_frame(aCodecCtx, audio_buf, sizeof(audio_buf));
if(audio_size < 0)
{
/* If error, output silence */
audio_buf_size = 1024; // arbitrary?
memset(audio_buf, 0, audio_buf_size);
}
else
{
audio_buf_size = audio_size;
}
audio_buf_index = 0;
}
len1 = audio_buf_size - audio_buf_index;
if (len1 > len)
len1 = len;
memcpy(stream, (uint8_t *)audio_buf + audio_buf_index, len1);
len -= len1;
stream += len1;
audio_buf_index += len1;
}
}
void setup_ffmpeg(char* filename)
{
if (av_open_input_file(&pFormatCtx, filename, NULL, 0, NULL) != 0)
{
fprintf(stderr, "FFmpeg failed to open file %s!\n", filename);
exit(-1);
}
if (av_find_stream_info(pFormatCtx) < 0)
{
fprintf(stderr, "FFmpeg failed to retrieve stream info!\n");
exit(-1);
}
// Dump information about file onto standard error
dump_format(pFormatCtx, 0, filename, 0);
// Find the first video stream
int i = 0;
for (i; i < pFormatCtx->nb_streams; i++)
{
if (pFormatCtx->streams[i]->codec->codec_type == CODEC_TYPE_VIDEO && videoStream < 0)
{
videoStream = i;
}
if (pFormatCtx->streams[i]->codec->codec_type == CODEC_TYPE_AUDIO && audioStream < 0)
{
audioStream = i;
}
}
if (videoStream == -1)
{
fprintf(stderr, "No video stream found in %s!\n", filename);
exit(-1);
}
if (audioStream == -1)
{
fprintf(stderr, "No audio stream found in %s!\n", filename);
exit(-1);
}
// Get a pointer to the codec context for the audio stream
aCodecCtx = pFormatCtx->streams[audioStream]->codec;
// Set audio settings from codec info
SDL_AudioSpec wanted_spec;
wanted_spec.freq = aCodecCtx->sample_rate;
wanted_spec.format = AUDIO_S16SYS;
wanted_spec.channels = aCodecCtx->channels;
wanted_spec.silence = 0;
wanted_spec.samples = SDL_AUDIO_BUFFER_SIZE;
wanted_spec.callback = audio_callback;
wanted_spec.userdata = aCodecCtx;
SDL_AudioSpec spec;
if (SDL_OpenAudio(&wanted_spec, &spec) < 0)
{
fprintf(stderr, "SDL_OpenAudio: %s\n", SDL_GetError());
exit(-1);
}
AVCodec* aCodec = avcodec_find_decoder(aCodecCtx->codec_id);
if (!aCodec)
{
fprintf(stderr, "Unsupported codec!\n");
exit(-1);
}
avcodec_open(aCodecCtx, aCodec);
// audio_st = pFormatCtx->streams[index]
packet_queue_init(&audioq);
SDL_PauseAudio(0);
// Get a pointer to the codec context for the video stream
pCodecCtx = pFormatCtx->streams[videoStream]->codec;
// Find the decoder for the video stream
AVCodec* pCodec = avcodec_find_decoder(pCodecCtx->codec_id);
if (pCodec == NULL)
{
fprintf(stderr, "Unsupported codec!\n");
exit(-1); // Codec not found
}
// Open codec
if (avcodec_open(pCodecCtx, pCodec) < 0)
{
fprintf(stderr, "Unsupported codec!\n");
exit(-1); // Could not open codec
}
}
int main(int argc, char* argv[])
{
if (argc < 2)
{
std::cout << "Usage: " << argv[0] << " <video>" << std::endl;
return -1;
}
av_register_all();
// Init SDL
if (SDL_Init(SDL_INIT_VIDEO | SDL_INIT_AUDIO | SDL_INIT_TIMER))
{
fprintf(stderr, "Could not initialize SDL - %s\n", SDL_GetError());
return -1;
}
// Init ffmpeg and setup some SDL stuff related to Audio
setup_ffmpeg(argv[1]);
VideoCapture cap(argv[1]); // open the default camera
if (!cap.isOpened()) // check if we succeeded
{
std::cout << "Failed to load file!" << std::endl;
return -1;
}
AVPacket packet;
while (av_read_frame(pFormatCtx, &packet) >= 0)
{
if (packet.stream_index == videoStream)
{
// Actually this is were SYNC between audio/video would happen.
// Right now I assume that every VIDEO packet contains an entire video frame, and that's not true. A video frame can be made by multiple packets!
// But for the time being, assume 1 video frame == 1 video packet,
// so instead of reading the frame through ffmpeg, I read it through OpenCV.
Mat frame;
cap >> frame; // get a new frame from camera
// do some processing on the frame, either as a Mat or as IplImage.
// For educational purposes, applying a lame grayscale conversion
IplImage ipl_frame = frame;
for (int i = 0; i < ipl_frame.width * ipl_frame.height * ipl_frame.nChannels; i += ipl_frame.nChannels)
{
ipl_frame.imageData[i] = (ipl_frame.imageData[i] + ipl_frame.imageData[i+1] + ipl_frame.imageData[i+2])/3; //B
ipl_frame.imageData[i+1] = (ipl_frame.imageData[i] + ipl_frame.imageData[i+1] + ipl_frame.imageData[i+2])/3; //G
ipl_frame.imageData[i+2] = (ipl_frame.imageData[i] + ipl_frame.imageData[i+1] + ipl_frame.imageData[i+2])/3; //R
}
// Display it on SDL window
show_frame(&ipl_frame);
av_free_packet(&packet);
}
else if (packet.stream_index == audioStream)
{
packet_queue_put(&audioq, &packet);
}
else
{
av_free_packet(&packet);
}
SDL_Event event;
SDL_PollEvent(&event);
switch (event.type)
{
case SDL_QUIT:
SDL_FreeSurface(surface);
SDL_Quit();
break;
default:
break;
}
}
// the camera will be deinitialized automatically in VideoCapture destructor
// Close the codec
avcodec_close(pCodecCtx);
// Close the video file
av_close_input_file(pFormatCtx);
return 0;
}
On my Mac I compiled it with:
g++ ffmpeg_snd.cpp -o ffmpeg_snd -D_GNU_SOURCE=1 -D_THREAD_SAFE -I/usr/local/include/opencv -I/usr/local/include -I/usr/local/include/SDL -Wl,-framework,Cocoa -L/usr/local/lib -lopencv_core -lopencv_imgproc -lopencv_highgui -lopencv_ml -lopencv_video -lopencv_features2d -lopencv_calib3d -lopencv_objdetect -lopencv_contrib -lopencv_legacy -lopencv_flann -lSDLmain -lSDL -L/usr/local/lib -lavfilter -lavcodec -lavformat -I/usr/local/Cellar/ffmpeg/HEAD/include/libavcodec -I/usr/local/Cellar/ffmpeg/HEAD/include/libavformat