[openal] Format clarification

[Chris Robinson]

On 12/17/2014 03:20 AM, Miroslav Andel wrote:
> Hi,
>
> Are the multichannel formats such as AL_FORMAT_51CHN32,
> AL_FORMAT_61CHN32 and AL_FORMAT_71CHN32 signed 32-bit integers or 32-bit
> floats? Couldn't find any documentation.

Hi.

They are 32-bit floats, so they match up with AL_FORMAT_MONO_FLOAT32 and 
AL_FORMAT_STEREO_FLOAT32. So, AL_EXT_FLOAT32 and AL_EXT_MCFORMATS 
together provide for mono, stereo, rear, quad, 5.1, 6.1, and 7.1 channel 
configurations for unsigned 8-bit, signed 16-bit, and 32-bit float 
sample types.

If you want to use 32-bit integer samples, you'll need to convert them 
to 32-bit float or 16-bit integer samples.