FUTURE NOTES
-
"AudioUnitRender can provide pointers to its own buffers." Perhaps this is better for "generators" as it will spare a copy op?
-
in process(), is it safe to render to L,R only? Do we need to read the number of frames actually written? Do they get zeroed?
-
Should we get rid of the complex RCB since we're wrapping it for microphone data anyway? Instead just have a function (or block!) with numFrames, *L, *R, maybe inTimeStamp (as frames?). What about pre/post render business?
TODO
- Typedef Bell_Size and cleanup (can frame ever be < 0? Why the BellSize sometimes??)