Implement Miniaudio for sound and music processing/mixing/playback

[dashodanger]

This an initial implementation of Miniaudio for handling our various audio-related needs. It offers the following advantages over our current built-in mixing implementation:

1- Does not require managing device lock/unlocks from the main thread
2- Can accommodate new platforms/libraries without requiring a significant rework (for instance, can be used with SDL or sokol_app or what have you; custom backends can even be written if necessary although it supports the platforms we already target natively)
3 - True 3D spatialization of sounds
4 - Node system will not require allocating additional memory for mixed versions of underwater/vacuum/reverb effects or preprocessing the sound with effects added before playback
5 - More flexible regarding disparity of sound formats, channel counts, etc, and the underlying device's properties
6 - Simplified Sound Menu for the end user
7 - Custom decoders can be written to expand the audio types supported, although this PR implements support for all of the formats that we already supported, with the following exceptions:

• RAD v2 module support has been removed. The player code did not mesh well with the 'decoder' concept
• OPL and Emu de MIDI playback of MIDI tracks have been removed to simplify the MIDI 'decoder' implementation
  • IMF support, which relies on OPL support, is still in and is contained in its own source/decoder implementation

[pbdot]

Awesome, so many wins, not needing to lock the main thread, and the consistency on the audio format code is really nice. I looked it over, all looks good 👍