Added async audio support to Python bindings and added noise cancelling channel to micview

[EgbertW]

've been working on the audio support on the kinect using libfreenect.

Since we are using Python for our project, we needed to access the audio data in Python. I changed the freenect cython file to claim the audio device and provide an audio callback, which will be supplied with the streams of the 4 microphones and the noise cancelled channel as a numpy array.

Also, to visualize it, I modified the micview.c example to include a 5th bar for the noise cancelled data. Unfortunately, apparently no data comes from this channel. This is also confirmed by the python example: the noise cancelled channel is silent. If anyone has any idea how to fix this, please help out.

[zarvox]

Thanks for the patches! They are quite handy, but to merge them I need to be able to also build the underlying library without audio support (ie, for OSX or Windows), lest we completely break the Python wrapper on those OSes. Additionally, I need each of your commits to be signed off as described here: http://openkinect.org/wiki/Contributing_Code#Sign_your_work .

If you can rework the patch to work correctly when no audio support exists in libfreenect and resubmit with signoffs, I'll figure out how to tweak it to merge atop unstable with marcan's recent python wrapper updates. :)

As for the cancelled channel: the noise cancelled data we receive from the Kinect is currently always silent. I'm not entirely sure why, but it might have to do with the fact that we don't do a cryptographic handshake with the Kinect that the Xbox360 does (that I haven't yet found the time to reimplement). Some time when I have a free weekend I'll take another look at this and see if I can find out anything else.

[pcp16]

Hi, this hasn't been merged yet? What would it need to be merged?

[ozzitheawesum]

Can't understand how to use the python wrapper to use the microphone array to record audio?
can you help me!!