From 86db1c6dcd6f69028c9c25d03e6e19ef9fc2b331 Mon Sep 17 00:00:00 2001
From: paolo <paolopr976@gmail.com>
Date: Mon, 6 Jun 2022 15:34:05 +0200
Subject: [PATCH] Added a complete example - with clear, well readable and
sequential code - for doing basic operations on a raw audio file. It improves
the current state of the examples directory
---
configure | 2 +
doc/examples/Makefile | 47 ++--
doc/examples/Makefile.example | 1 +
doc/examples/convert_encode_mux_audio.c | 351 ++++++++++++++++++++++++
4 files changed, 378 insertions(+), 23 deletions(-)
create mode 100644 doc/examples/convert_encode_mux_audio.c
@@ -1725,6 +1725,7 @@ COMPONENT_LIST="
EXAMPLE_LIST="
avio_list_dir_example
avio_reading_example
+ convert_encode_mux_audio_example
decode_audio_example
decode_video_example
demuxing_decoding_example
@@ -3760,6 +3761,7 @@ yadif_videotoolbox_filter_deps="metal corevideo videotoolbox"
# examples
avio_list_dir_deps="avformat avutil"
avio_reading_deps="avformat avcodec avutil"
+convert_encode_mux_audio_example_deps="avcodec avformat avutil swresample"
decode_audio_example_deps="avcodec avutil"
decode_video_example_deps="avcodec avutil"
demuxing_decoding_example_deps="avcodec avformat avutil"
@@ -1,26 +1,27 @@
-EXAMPLES-$(CONFIG_AVIO_LIST_DIR_EXAMPLE) += avio_list_dir
-EXAMPLES-$(CONFIG_AVIO_READING_EXAMPLE) += avio_reading
-EXAMPLES-$(CONFIG_DECODE_AUDIO_EXAMPLE) += decode_audio
-EXAMPLES-$(CONFIG_DECODE_VIDEO_EXAMPLE) += decode_video
-EXAMPLES-$(CONFIG_DEMUXING_DECODING_EXAMPLE) += demuxing_decoding
-EXAMPLES-$(CONFIG_ENCODE_AUDIO_EXAMPLE) += encode_audio
-EXAMPLES-$(CONFIG_ENCODE_VIDEO_EXAMPLE) += encode_video
-EXAMPLES-$(CONFIG_EXTRACT_MVS_EXAMPLE) += extract_mvs
-EXAMPLES-$(CONFIG_FILTER_AUDIO_EXAMPLE) += filter_audio
-EXAMPLES-$(CONFIG_FILTERING_AUDIO_EXAMPLE) += filtering_audio
-EXAMPLES-$(CONFIG_FILTERING_VIDEO_EXAMPLE) += filtering_video
-EXAMPLES-$(CONFIG_HTTP_MULTICLIENT_EXAMPLE) += http_multiclient
-EXAMPLES-$(CONFIG_HW_DECODE_EXAMPLE) += hw_decode
-EXAMPLES-$(CONFIG_METADATA_EXAMPLE) += metadata
-EXAMPLES-$(CONFIG_MUXING_EXAMPLE) += muxing
-EXAMPLES-$(CONFIG_QSVDEC_EXAMPLE) += qsvdec
-EXAMPLES-$(CONFIG_REMUXING_EXAMPLE) += remuxing
-EXAMPLES-$(CONFIG_RESAMPLING_AUDIO_EXAMPLE) += resampling_audio
-EXAMPLES-$(CONFIG_SCALING_VIDEO_EXAMPLE) += scaling_video
-EXAMPLES-$(CONFIG_TRANSCODE_AAC_EXAMPLE) += transcode_aac
-EXAMPLES-$(CONFIG_TRANSCODING_EXAMPLE) += transcoding
-EXAMPLES-$(CONFIG_VAAPI_ENCODE_EXAMPLE) += vaapi_encode
-EXAMPLES-$(CONFIG_VAAPI_TRANSCODE_EXAMPLE) += vaapi_transcode
+EXAMPLES-$(CONFIG_AVIO_LIST_DIR_EXAMPLE) += avio_list_dir
+EXAMPLES-$(CONFIG_AVIO_READING_EXAMPLE) += avio_reading
+EXAMPLES-$(CONFIG_CONVERT_ENCODE_MUX_AUDIO_EXAMPLE) += convert_encode_mux_audio
+EXAMPLES-$(CONFIG_DECODE_AUDIO_EXAMPLE) += decode_audio
+EXAMPLES-$(CONFIG_DECODE_VIDEO_EXAMPLE) += decode_video
+EXAMPLES-$(CONFIG_DEMUXING_DECODING_EXAMPLE) += demuxing_decoding
+EXAMPLES-$(CONFIG_ENCODE_AUDIO_EXAMPLE) += encode_audio
+EXAMPLES-$(CONFIG_ENCODE_VIDEO_EXAMPLE) += encode_video
+EXAMPLES-$(CONFIG_EXTRACT_MVS_EXAMPLE) += extract_mvs
+EXAMPLES-$(CONFIG_FILTER_AUDIO_EXAMPLE) += filter_audio
+EXAMPLES-$(CONFIG_FILTERING_AUDIO_EXAMPLE) += filtering_audio
+EXAMPLES-$(CONFIG_FILTERING_VIDEO_EXAMPLE) += filtering_video
+EXAMPLES-$(CONFIG_HTTP_MULTICLIENT_EXAMPLE) += http_multiclient
+EXAMPLES-$(CONFIG_HW_DECODE_EXAMPLE) += hw_decode
+EXAMPLES-$(CONFIG_METADATA_EXAMPLE) += metadata
+EXAMPLES-$(CONFIG_MUXING_EXAMPLE) += muxing
+EXAMPLES-$(CONFIG_QSVDEC_EXAMPLE) += qsvdec
+EXAMPLES-$(CONFIG_REMUXING_EXAMPLE) += remuxing
+EXAMPLES-$(CONFIG_RESAMPLING_AUDIO_EXAMPLE) += resampling_audio
+EXAMPLES-$(CONFIG_SCALING_VIDEO_EXAMPLE) += scaling_video
+EXAMPLES-$(CONFIG_TRANSCODE_AAC_EXAMPLE) += transcode_aac
+EXAMPLES-$(CONFIG_TRANSCODING_EXAMPLE) += transcoding
+EXAMPLES-$(CONFIG_VAAPI_ENCODE_EXAMPLE) += vaapi_encode
+EXAMPLES-$(CONFIG_VAAPI_TRANSCODE_EXAMPLE) += vaapi_transcode
EXAMPLES := $(EXAMPLES-yes:%=doc/examples/%$(PROGSSUF)$(EXESUF))
EXAMPLES_G := $(EXAMPLES-yes:%=doc/examples/%$(PROGSSUF)_g$(EXESUF))
@@ -13,6 +13,7 @@ LDLIBS := $(shell pkg-config --libs $(FFMPEG_LIBS)) $(LDLIBS)
EXAMPLES= avio_list_dir \
avio_reading \
+ convert_encode_mux_audio \
decode_audio \
decode_video \
demuxing_decoding \
new file mode 100644
@@ -0,0 +1,351 @@
+/*
+ * Copyright (c) 2019 Paolo Prete (paolopr976 at gmail.com)
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+ * THE SOFTWARE.
+ */
+
+/**
+ * @file
+ * API example for converting, encoding and muxing raw audio files.
+ * This example reads a raw audio input file, converts it to a sample format accepted by
+ * a given encoder, performs encoding and puts the encoded frames into a muxing container.
+ * The encoded stream is written to a file named "outfile".
+ * It can be adapted, with few changes, to a custom raw audio source (i.e: a live one).
+ * It uses a custom I/O write callback (write_muxed_data()) in order to show how to access
+ * muxed packets written in memory, before they are written to the output file.
+ *
+ * The raw input audio file can be created with:
+ *
+ * ffmpeg -i some_audio_file -f f32le -acodec pcm_f32le -ac 2 -ar 48000 raw_audio_file.raw
+ *
+ * @example convert_encode_mux_audio.c
+ */
+
+#include <libavcodec/avcodec.h>
+#include <libavformat/avformat.h>
+#include <libavutil/timestamp.h>
+#include <libswresample/swresample.h>
+
+// Customizable parameters
+#define SAMPLE_RATE 48000
+#define ENCODER_ID AV_CODEC_ID_AAC
+#define ENCODER_BITRATE 64000
+#define INPUT_SAMPLE_FMT AV_SAMPLE_FMT_FLT
+#define OUTPUT_SAMPLE_FMT AV_SAMPLE_FMT_FLTP
+#define INPUT_CHANNELS 2
+#define OUTPUT_CHANNELS 2
+
+static int encoded_pkt_counter = 1;
+
+static int write_muxed_data(void *opaque, uint8_t *data, int size)
+{
+ FILE *encoded_audio_file = (FILE *)opaque;
+ fwrite(data, 1, size, encoded_audio_file); //(h)
+ return size;
+}
+
+static int mux_encoded_packet(AVPacket *encoded_audio_packet, AVFormatContext *out_container_ctx)
+{
+ int ret_val;
+ if ((ret_val = av_write_frame(out_container_ctx, encoded_audio_packet)) < 0) {
+ av_log(NULL, AV_LOG_ERROR, "Error calling av_write_frame() (error '%s')\n", av_err2str(ret_val));
+ } else {
+ av_log(NULL, AV_LOG_INFO, "Encoded packet %d, size=%d, pts_time=%s\n",
+ encoded_pkt_counter, encoded_audio_packet->size,
+ av_ts2timestr(encoded_audio_packet->pts, &out_container_ctx->streams[0]->time_base));
+ }
+ return ret_val;
+}
+
+static int check_if_samplerate_is_supported(const AVCodec *audio_codec, int samplerate)
+{
+ const int *samplerates_list = audio_codec->supported_samplerates;
+ while (*samplerates_list) {
+ if (*samplerates_list == samplerate)
+ return 0;
+ samplerates_list++;
+ }
+ return 1;
+}
+
+int main(int argc, char **argv)
+{
+ FILE *input_audio_file = NULL, *encoded_audio_file = NULL;
+ AVChannelLayout in_ch_layout;
+ AVChannelLayout out_ch_layout;
+ const AVCodec *audio_codec = NULL;
+ AVCodecContext *audio_encoder_ctx = NULL;
+ AVFrame *input_audio_frame = NULL, *converted_audio_frame = NULL;
+ SwrContext *audio_convert_context = NULL;
+ const AVOutputFormat *out_container = NULL;
+ AVFormatContext *out_container_ctx = NULL;
+ uint8_t *out_container_buffer = NULL;
+ size_t out_container_buffer_size = 4096;
+ AVIOContext *avio_ctx = NULL;
+ AVStream *out_stream = NULL;
+ AVPacket *encoded_audio_packet = NULL;
+ int ret_val = 0;
+ int audio_bytes_to_encode;
+
+ if (argc != 3) {
+ printf("Usage: %s <raw audio input file (INPUT_CHANNELS, INPUT_SAMPLE_FMT, SAMPLE_RATE)>, <container format>\n", argv[0]);
+ return 1;
+ }
+
+ input_audio_file = fopen(argv[1], "rb");
+ if (!input_audio_file) {
+ av_log(NULL, AV_LOG_ERROR, "Could not open input audio file\n");
+ return AVERROR_EXIT;
+ }
+
+ encoded_audio_file = fopen("outfile", "wb");
+ if (!encoded_audio_file) {
+ av_log(NULL, AV_LOG_ERROR, "Could not open output audio file\n");
+ fclose(input_audio_file);
+ return AVERROR_EXIT;
+ }
+
+ /**
+ * Allocate the encoder's context and open the encoder
+ */
+ audio_codec = avcodec_find_encoder(ENCODER_ID);
+ if (!audio_codec) {
+ av_log(NULL, AV_LOG_ERROR, "Could not find encoder's codec\n");
+ ret_val = AVERROR_EXIT;
+ goto end;
+ }
+ if ((ret_val = check_if_samplerate_is_supported(audio_codec, SAMPLE_RATE)) != 0) {
+ av_log(NULL, AV_LOG_ERROR, "Audio codec doesn't support input samplerate %d\n", SAMPLE_RATE);
+ goto end;
+ }
+ audio_encoder_ctx = avcodec_alloc_context3(audio_codec);
+ if (!audio_codec) {
+ av_log(NULL, AV_LOG_ERROR, "Could not allocate the encoding context\n");
+ ret_val = AVERROR_EXIT;
+ goto end;
+ }
+ av_channel_layout_default(&out_ch_layout, OUTPUT_CHANNELS);
+ audio_encoder_ctx->sample_fmt = OUTPUT_SAMPLE_FMT;
+ audio_encoder_ctx->bit_rate = ENCODER_BITRATE;
+ audio_encoder_ctx->sample_rate = SAMPLE_RATE;
+ audio_encoder_ctx->ch_layout = out_ch_layout;
+ audio_encoder_ctx->time_base = (AVRational){1, SAMPLE_RATE};
+ audio_encoder_ctx->codec_type = AVMEDIA_TYPE_AUDIO ;
+ if ((ret_val = avcodec_open2(audio_encoder_ctx, audio_codec, NULL)) < 0) {
+ av_log(NULL, AV_LOG_ERROR, "Could not open input codec (error '%s')\n", av_err2str(ret_val));
+ goto end;
+ }
+
+ /**
+ * Allocate an AVFrame which will be filled with the input file's data.
+ */
+ if (!(input_audio_frame = av_frame_alloc())) {
+ av_log(NULL, AV_LOG_ERROR, "Could not allocate input frame\n");
+ ret_val = AVERROR(ENOMEM);
+ goto end;
+ }
+ av_channel_layout_default(&in_ch_layout, INPUT_CHANNELS);
+ input_audio_frame->nb_samples = audio_encoder_ctx->frame_size;
+ input_audio_frame->format = INPUT_SAMPLE_FMT;
+ input_audio_frame->sample_rate = SAMPLE_RATE;
+ input_audio_frame->ch_layout = in_ch_layout;
+ // Allocate the frame's data buffer
+ if ((ret_val = av_frame_get_buffer(input_audio_frame, 0)) < 0) {
+ av_log(NULL, AV_LOG_ERROR,
+ "Could not allocate container for input frame samples (error '%s')\n", av_err2str(ret_val));
+ ret_val = AVERROR(ENOMEM);
+ goto end;
+ }
+
+ /**
+ * Input data must be converted in the right format required by the encoder.
+ * We allocate a SwrContext and an AVFrame (which will contain the converted samples) for this task.
+ * The AVFrame will feed the encoding function (avcodec_send_frame())
+ */
+ swr_alloc_set_opts2(&audio_convert_context,
+ &out_ch_layout,
+ OUTPUT_SAMPLE_FMT,
+ SAMPLE_RATE,
+ &in_ch_layout,
+ INPUT_SAMPLE_FMT,
+ SAMPLE_RATE,
+ 0,
+ NULL);
+ if (!audio_convert_context) {
+ av_log(NULL, AV_LOG_ERROR, "Could not allocate resample context\n");
+ ret_val = AVERROR(ENOMEM);
+ goto end;
+ }
+ if (!(converted_audio_frame = av_frame_alloc())) {
+ av_log(NULL, AV_LOG_ERROR, "Could not allocate resampled frame\n");
+ ret_val = AVERROR(ENOMEM);
+ goto end;
+ }
+ converted_audio_frame->nb_samples = audio_encoder_ctx->frame_size;
+ converted_audio_frame->format = audio_encoder_ctx->sample_fmt;
+ converted_audio_frame->ch_layout = audio_encoder_ctx->ch_layout;
+ converted_audio_frame->sample_rate = SAMPLE_RATE;
+ if ((ret_val = av_frame_get_buffer(converted_audio_frame, 0)) < 0) {
+ av_log(NULL, AV_LOG_ERROR,
+ "Could not allocate a buffer for resampled frame samples (error '%s')\n", av_err2str(ret_val));
+ goto end;
+ }
+
+ /**
+ * Create the output container for the encoded frames
+ */
+ out_container = av_guess_format(argv[2], NULL, NULL);
+ if (!out_container) {
+ av_log(NULL, AV_LOG_ERROR, "Could not find output format\n");
+ ret_val = AVERROR_EXIT;
+ goto end;
+ }
+ if ((ret_val = avformat_alloc_output_context2(&out_container_ctx, out_container, "", NULL)) < 0) {
+ av_log(NULL, AV_LOG_ERROR, "Could not create output context (error '%s')\n", av_err2str(ret_val));
+ goto end;
+ }
+ if (!(out_container_buffer = av_malloc(out_container_buffer_size))) {
+ av_log(NULL, AV_LOG_ERROR, "Could not allocate a buffer for the I/O output context\n");
+ ret_val = AVERROR(ENOMEM);
+ goto end;
+ }
+
+ /**
+ * Create an I/O context for the muxer's container with a write callback (write_muxed_data()),
+ * so that muxed data will be accessed through this function and can be managed by the user.
+ */
+ if (!(avio_ctx = avio_alloc_context(out_container_buffer, out_container_buffer_size,
+ 1, encoded_audio_file, NULL,
+ &write_muxed_data, NULL))) {
+ av_log(NULL, AV_LOG_ERROR, "Could not create I/O output context\n");
+ ret_val = AVERROR_EXIT;
+ goto end;
+ }
+
+ /**
+ * Link the container's context to the previous I/O context
+ */
+ out_container_ctx->pb = avio_ctx;
+ if (!(out_stream = avformat_new_stream(out_container_ctx, NULL))) {
+ av_log(NULL, AV_LOG_ERROR, "Could not create new stream\n");
+ ret_val = AVERROR(ENOMEM);
+ goto end;
+ }
+ out_stream->id = out_container_ctx->nb_streams-1;
+
+ // Copy the encoder's parameters
+ avcodec_parameters_from_context(out_stream->codecpar, audio_encoder_ctx);
+ // Allocate the stream private data and write the stream header
+ if (avformat_write_header(out_container_ctx, NULL) < 0) {
+ av_log(NULL, AV_LOG_ERROR, "avformat_write_header() error\n");
+ ret_val = AVERROR_EXIT;
+ goto end;
+ }
+
+ /**
+ * (a) Fill the input frame's data buffer with input file data
+ * (b) Convert the input frame to output sample format
+ * (c) Assign a pts to the converted frame based on the number ot total samples read
+ * (d) Send the converted frame to the encoder
+ * (e) Get the encoded packet
+ * (f) Rescale the timestamps of the encoded packet to the muxer's timebase
+ * (g) Send the encoded packet, with rescaled timestamps, to the muxer
+ * (h) Muxed data is caught in write_muxed_data() callback (see above) and it is
+ * written to the output audio file
+ */
+ encoded_audio_packet = av_packet_alloc();
+ while (1) {
+
+ audio_bytes_to_encode = fread(input_audio_frame->data[0], 1,
+ input_audio_frame->linesize[0], input_audio_file); //(a)
+ if (audio_bytes_to_encode != input_audio_frame->linesize[0]) {
+ break;
+ } else {
+ if (av_frame_make_writable(converted_audio_frame) < 0) {
+ av_log(NULL, AV_LOG_ERROR, "av_frame_make_writable() error\n");
+ ret_val = AVERROR_EXIT;
+ goto end;
+ }
+
+ if ((ret_val = swr_convert_frame(audio_convert_context,
+ converted_audio_frame,
+ (const AVFrame *)input_audio_frame)) != 0) { //(b)
+ av_log(NULL, AV_LOG_ERROR,
+ "Error resampling input audio frame (error '%s')\n", av_err2str(ret_val));
+ goto end;
+ }
+
+ converted_audio_frame->pts = converted_audio_frame->nb_samples*(encoded_pkt_counter-1); //(c)
+
+ if ((ret_val = avcodec_send_frame(audio_encoder_ctx, converted_audio_frame)) == 0) //(d)
+ ret_val = avcodec_receive_packet(audio_encoder_ctx, encoded_audio_packet); //(e)
+ else {
+ av_log(NULL, AV_LOG_ERROR,
+ "Error encoding frame (error '%s')\n", av_err2str(ret_val));
+ goto end;
+ }
+
+ if (ret_val == 0) {
+ encoded_audio_packet->dts = encoded_audio_packet->pts = converted_audio_frame->pts;
+ av_packet_rescale_ts(encoded_audio_packet, audio_encoder_ctx->time_base,
+ out_stream->time_base); //(f)
+ if ((ret_val = mux_encoded_packet(encoded_audio_packet, out_container_ctx)) < 0) //(g)
+ goto end;
+ encoded_pkt_counter++;
+ } else if (ret_val != AVERROR(EAGAIN)) {
+ av_log(NULL, AV_LOG_ERROR,
+ "Error receiving encoded packet (error '%s')\n", av_err2str(ret_val));
+ goto end;
+ }
+ }
+ }
+
+ // Flush cached packets
+ if ((ret_val = avcodec_send_frame(audio_encoder_ctx, NULL)) == 0)
+ do {
+ ret_val = avcodec_receive_packet(audio_encoder_ctx, encoded_audio_packet);
+ if (ret_val == 0) {
+ encoded_audio_packet->dts = converted_audio_frame->nb_samples*(encoded_pkt_counter-1);
+ encoded_audio_packet->pts = encoded_audio_packet->dts;
+ av_packet_rescale_ts(encoded_audio_packet, audio_encoder_ctx->time_base,
+ out_stream->time_base);
+ if ((ret_val = mux_encoded_packet(encoded_audio_packet, out_container_ctx)) < 0)
+ goto end;
+ encoded_pkt_counter++;
+ }
+ } while (ret_val == 0);
+
+ av_write_trailer(out_container_ctx);
+
+end:
+
+ fclose(input_audio_file);
+ fclose(encoded_audio_file);
+ avcodec_free_context(&audio_encoder_ctx);
+ av_frame_free(&input_audio_frame);
+ swr_free(&audio_convert_context);
+ av_frame_free(&converted_audio_frame);
+ avformat_free_context(out_container_ctx);
+ av_freep(&avio_ctx);
+ av_freep(&out_container_buffer);
+ av_packet_free(&encoded_audio_packet);
+
+ return ret_val;
+
+}
--
2.32.0