It appears that the quality of denoised audio deteriorates when the size of the buffer passed to rnnoise_process_frame drops below 512 samples. Is it because the frame size drops below 480? Any tradeoffs for processing of smaller buffers e.g. 128 samples per call to rnnoise_process_frame?
It appears that the quality of denoised audio deteriorates when the size of the buffer passed to rnnoise_process_frame drops below 512 samples. Is it because the frame size drops below 480? Any tradeoffs for processing of smaller buffers e.g. 128 samples per call to rnnoise_process_frame?