enable >16 bit output with WMME api
Currently PA/WMME always uses 16 bit host buffers internally.
InitializeWaveHandles() could be updated to try a preferred format and fall back to 16 bit if necessary (it already tries waveformatex before waveformat. it could try waveformatex(preferred) then waveformatex(16) then waveformat(16).
Not sure exactly what else would need to change to support other buffer formats.
We'd want to support linear integer formats. not sure we need to support float since we probably want to use our own floating point converters rather than windows ones.
InitializeWaveHandles() could be updated to try a preferred format and fall back to 16 bit if necessary (it already tries waveformatex before waveformat. it could try waveformatex(preferred) then waveformatex(16) then waveformat(16).
Not sure exactly what else would need to change to support other buffer formats.
We'd want to support linear integer formats. not sure we need to support float since we probably want to use our own floating point converters rather than windows ones.
Leave a comment