SDL 3 Migration - Audio API

[Ravbug]

SDL 3 Migration - Audio API

[icculus]

Is it possible to configure SDL3 to invoke the audio callback such that it always asks for the same number of samples every time it is called?

No, but the callback also does not care how much data you actually put to the audiostream on each call. If you give it less than requested, it will fill the rest of the playback time with silence (which could make sense at the end of playback), and if you give it more, it'll just have extra already available when it goes to consume more data.

Which is to say: you can just feed the thing a fixed amount every time it runs, regardless of what it asks for.

The only gotcha is that you need to make sure you give it at least what it asks for so you don't get skips in the audio. After opening the audio device, you can call SDL_GetAudioDeviceFormat to find out how much data it intends to consume at a time, and then set up the middleware to output that much audio per iteration.

[Ravbug]

This works, thanks!