Files
dmf-studio-rnd/shared/VideoReader.hpp
T
JohannesItten bb7d1bb5a3 fix: VideoReader bugs — stride, memory, audio detection, UB
Pass MXL grain stride (from sliceSizes[0]) into get_next_frame so
YUV422P10toV210 writes with the correct line width instead of a
self-computed value that may not match the MXL buffer.

Fix audio stream detection: condition was inverted (!= -1 → == -1),
so the first audio stream was never picked up.

Add return false at end of get_next_frame to fix UB when g_running
goes false and the loop exits without returning.

Replace av_frame_unref/av_packet_unref with av_frame_free/av_packet_free
in destructor — unref only releases data, not the struct itself.

Add av_freep(&p10_data[0]) in destructor to free av_image_alloc memory.

Remove unused p10_buffer and v210_buffer allocations.

Read filename from config("file") instead of hardcoded path.
Add early return if mxlCreateFlowWriter fails.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-07-03 19:11:18 +03:00

205 lines
7.3 KiB
C++

#pragma once
extern "C" {
#include <libavformat/avformat.h>
#include <libavcodec/avcodec.h>
#include <libavutil/avutil.h>
#include <libswscale/swscale.h>
#include <libavutil/imgutils.h>
#include <libavutil/pixdesc.h>
}
#include <stdexcept>
#include <string>
#include "Signal.hpp"
#include "V210.hpp"
namespace dmf {
class VideoReader {
public:
struct SourceInfo {
int width = 0;
int height = 0;
int fps_num = 0;
int fps_den = 0;
int stride = 0;
AVPixelFormat pix_fmt{};
};
struct AudioInfo {
int sample_rate = 0;
int channels = 0;
int samples = 0;
int channel_stride = 0; // floats between channel planes
};
struct SourceInfo source_info{};
bool has_audio = false;
bool have_video = false;
VideoReader(std::string filename) {
if (!open_file(filename)) {
return;
}
get_source_info();
allocate_conversion_buffers();
}
~VideoReader() {
avcodec_free_context(&codec_context);
avformat_close_input(&format_context);
sws_freeContext(sws_ctx);
av_freep(&p10_data[0]);
av_frame_free(&frame);
av_packet_free(&packet);
}
bool get_next_frame(uint8_t* video_buf, uint32_t mxl_stride, uint8_t* audiobuf) {
while (dmf::g_running.load(std::memory_order_relaxed)) {
// Try to get a buffered frame from previous packet first
if (avcodec_receive_frame(codec_context, frame) == 0) {
sws_scale(
sws_ctx,
frame->data,
frame->linesize,
0,
source_info.height,
p10_data,
p10_linesizes
);
dmf::v210::YUV422P10toV210(
reinterpret_cast<uint16_t*>(p10_data[0]),
reinterpret_cast<uint16_t*>(p10_data[1]),
reinterpret_cast<uint16_t*>(p10_data[2]),
video_buf,
source_info.width,
source_info.height,
p10_linesizes[0],
p10_linesizes[1],
p10_linesizes[2],
mxl_stride
);
av_frame_unref(frame);
return true;
}
// No buffered frame — read next packet
av_packet_unref(packet);
if (av_read_frame(format_context, packet) < 0) {
// EOF — seek back to start and keep going
avformat_seek_file(format_context, -1, 0, 0, 0, AVSEEK_FLAG_BACKWARD);
avcodec_flush_buffers(codec_context);
av_packet_unref(packet);
continue;
}
if (packet->stream_index != video_stream_index) continue;
avcodec_send_packet(codec_context, packet);
}
return false;
}
private:
AVFormatContext* format_context = nullptr;
AVCodecContext* codec_context = nullptr;
AVPacket* packet = av_packet_alloc();
AVFrame* frame = av_frame_alloc();
int video_stream_index = -1;
int audio_stream_index = -1;
// conversion data
struct SwsContext *sws_ctx{};
int p10_linesizes[4] = {0, 0, 0, 0};
uint8_t* p10_data[4] = {nullptr, nullptr, nullptr, nullptr};
bool open_file(std::string filename) {
if (avformat_open_input(&format_context, filename.c_str(), nullptr, nullptr) != 0) {
throw std::runtime_error("Could not open file: " + filename);
return false;
}
// Find stream info
if (avformat_find_stream_info(format_context, nullptr) < 0) {
avformat_close_input(&format_context);
throw std::runtime_error("Could not find stream info");
return false;
}
// Find streams
for (unsigned int i = 0; i < format_context->nb_streams; i++) {
AVMediaType data_type = format_context->streams[i]->codecpar->codec_type;
if (data_type == AVMEDIA_TYPE_VIDEO) {
video_stream_index = i;
have_video = true;
} else if (data_type == AVMEDIA_TYPE_AUDIO && audio_stream_index == -1) {
// TODO: show list of available audio tracks and allow user to pick
// or handle multiple audio streams
audio_stream_index = i;
has_audio = true;
}
}
if (video_stream_index == -1 && audio_stream_index == -1) {
avformat_close_input(&format_context);
throw std::runtime_error("No audio/video stream found");
return false;
}
return true;
}
void get_source_info() {
// Get codec parameters
AVCodecParameters* codec_params = format_context->streams[video_stream_index]->codecpar;
const AVCodec* codec = avcodec_find_decoder(codec_params->codec_id);
if (!codec) {
avformat_close_input(&format_context);
throw std::runtime_error("Unsupported codec");
}
// Open codec
codec_context = avcodec_alloc_context3(codec);
avcodec_parameters_to_context(codec_context, codec_params);
if (avcodec_open2(codec_context, codec, nullptr) < 0) {
avcodec_free_context(&codec_context);
avformat_close_input(&format_context);
throw std::runtime_error("Could not open codec");
}
AVRational fps = codec_context->framerate;
if (fps.num == 0 || fps.den == 0) {
fps = format_context->streams[video_stream_index]->avg_frame_rate;
}
source_info.width = codec_context->width;
source_info.height = codec_context->height;
source_info.fps_num = fps.num;
source_info.fps_den = fps.den;
source_info.pix_fmt = codec_context->pix_fmt;
}
void allocate_conversion_buffers()
{
// Create Sws context to convert to planar YUV 4:2:2, 20bpp, (1 Cr & Cb sample per 2x1 Y samples), LE
sws_ctx = sws_getContext(
source_info.width, source_info.height, source_info.pix_fmt,
source_info.width, source_info.height, AV_PIX_FMT_YUV422P10LE,
0, NULL, NULL, NULL
);
if (!sws_ctx) {
throw std::runtime_error("Failed to create SwsContext");
return;
}
// Allocate P10 image (freed in destructor via av_freep(&p10_data[0]))
av_image_alloc(
p10_data, p10_linesizes,
source_info.width, source_info.height,
AV_PIX_FMT_YUV422P10LE, 64
);
}
};
}