#include "video_encoder.h"
class ALSAOutput;
+class ChromaSubsampler;
class QSurface;
+class QSurfaceFormat;
namespace movit {
class Effect;
class FlatInput;
class ResourcePool;
} // namespace movit
-class QSurfaceFormat;
// For any card that's not the master (where we pick out the frames as they
// come, as fast as we can process), there's going to be a queue. The question
void schedule_audio_resampling_tasks(unsigned dropped_frames, int num_samples_per_frame, int length_per_frame);
void render_one_frame(int64_t duration);
void audio_thread_func();
- void subsample_chroma(GLuint src_tex, GLuint dst_dst);
void release_display_frame(DisplayFrame *frame);
double pts() { return double(pts_int) / TIMEBASE; }
std::atomic<unsigned> audio_source_channel{0};
std::atomic<unsigned> master_clock_channel{0};
std::unique_ptr<movit::EffectChain> display_chain;
- GLuint cbcr_program_num; // Owned by <resource_pool>.
- GLuint cbcr_texture_sampler_uniform;
- GLuint cbcr_vbo; // Holds position and texcoord data.
- GLuint cbcr_position_attribute_index, cbcr_texcoord_attribute_index;
+ std::unique_ptr<ChromaSubsampler> chroma_subsampler;
std::unique_ptr<VideoEncoder> video_encoder;
// Effects part of <display_chain>. Owned by <display_chain>.