First and foremost, I apologize if the inquiry I’m posting here might sound completely inaccurate, unfeasible or straight out of science fiction. I’m not a programmer by any means, simply an enthusiast who occasionally wonders about how things work.
I was dabbling around in the configurations of Retroarch and noticed that the Windows version currently provides the user with three different options for the Audio driver: directsound, xaudio and sdl2 (in addition to ‘nulling’ the audio entirely). Each one of these possibilities yields different results and, based on my experimentations, varying extents of tolerance with the Audio latency setting. Xaudio works fairly well for most cores and allows to bring the overall latency down by quite a considerable amount if compared to the default value of 64ms.
That being said, I’m curious whether an implementation of ASIO driver can be coded in Retroarch. It’s a popular solution that is commonly employed in both audio playback and audio production and it’s lauded for two significant benefits:
- bitperfect output, with the target application taking exclusive control of the audio and muting anything else (similarly to WASAPI);
- extremely low latency.
Once again, it’s an aspect I’m questioning without any in-depth technical competency on my part, but I was wondering whether this could actually be implemented one way or another. Thank you in advance for any insight!