This adds the following APIs:

  * ma_sound_set_fade_start_in_pcm_frames()
  * ma_sound_set_fade_start_in_milliseconds()

Public issue https://github.com/mackron/miniaudio/issues/669

The sample format of a WAV file is not always being set which results
in get_data_format() returning ma_format_unknown.

This change makes use of the onInitFile, onInitFileW and onInitMemory
backend callbacks which enables decoding backends to have optimized
implementations for reading from a file or a block of memory without
having to go through an abstraction layer on the miniaudio side.

Public issue https://github.com/mackron/miniaudio/issues/696

Public issue https://github.com/mackron/miniaudio/issues/707

Public issue https://github.com/mackron/miniaudio/issues/700

When this is used, the data callback should at some point call
ma_engine_read_pcm_frames() in order to do some processing.

This is optional and is fired at the end of each call to
ma_engine_read_process_pcm_frames(). The callback will be passed the
processed audio data so they can do their own processing such as
outputting to a file or whatnot.

The callback is configured via the engine config.