Floating point detection was backwards causing float-based audio to be detected as integer and vice versa (for >32-bit formats)
Floating point detection was backwards causing float-based audio to be detected as integer and vice versa (for >32-bit formats)