winecoreaudio: mmdevapi's mixer format is 32bit floating point.

[Joerg-Cyril.Hoehle at t-systems.com]

Hi,

the line numbers probably make this patch depend on the GetMixFormat one.

16bit PCM just makes no sense, for 2 reasons.

First, MSDN and some MS blogs mention how the mixer uses floating
point and how this works very well with IEEE formats without loss of
precision.  The tests show that 32bit FP is what GetMixFormat returns.

Second, when playing with Mac OS X' CLI audio utilities when I
got my Mac mini, I was surprised that it looked like floating
point was CoreAudio's preferred format.  Esp. on a Mac, it makes no
sense to output 16bit integer PCM that is likely to get turned
into FP and be mixed by OSX's CoreAudio using FP.

Regards,
	Jörg Höhle
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0001-winecoreaudio-mmdevapi-s-mixer-format-is-32bit-float.patch
Type: application/octet-stream
Size: 1097 bytes
Desc: 0001-winecoreaudio-mmdevapi-s-mixer-format-is-32bit-float.patch
URL: <http://www.winehq.org/pipermail/wine-patches/attachments/20110606/0acfd680/attachment.obj>