Hi, you could also try preloading libv4l, something like: LD_PRELOAD=/usr/lib/libv4l/v4l2convert.so skype
Regards, Paulo 2011/3/28 Benson Bear <benson.b...@gmail.com>: > On Sun, Mar 27, 2011 at 7:18 AM, I wrote: > >> This camera works nicely in everything like cheese, camstream, camorama, >> xawtv and guvcview. It doesn't work in the one thing I want it for: Skype. > >> This leads me to believe that perhaps the only problem is >> that Skype is not interpreting the output of the camera >> correctly, and all that perhaps need be done is to force >> the camera to put out the kind of output Skype thinks it >> is getting. > > In some practical sense at least I think I sort of solved > the problem, with a lot of googling. Having seen the > idea of gstreamer pipes as used years ago in that > "gst video stream" thing, I thought there should be a sink > to a virtual webcam, and it turns out there is, something > called v42loopback. > > I built this, loaded the kernel module it makes, getting > a /dev/video0, then plugged in the camera, getting a /dev/video1, > and then with some fiddling and guesswork ended up with: > > gst-launch v4l2src device=/dev/video1 ! > video/x-raw-yuv,width=640,height=480,framerate=20/1 ! v4l2sink > device=/dev/video0 > > and with this, skype produces good results reading from /dev/video0. > > I am still not sure what is happening here, perhaps an expert could explain. > I presume the v4l2src is putting out mjpg format and the second pipe > element forces it to yuv, which is what skype wants. > > Could Alexey or some other person say if there still could > be some issue with the webcam that could be fixed so this > would not be necessary? Also, could they advise on whether > this technique is an acceptable solution? For example, > I am a little concerned that the v4l2loopback is a kernel module > that is not distributed and I just causally built it and loaded > it without knowing much about how wise that is. (Maybe knowing > nothing about what is available for these gstreamer pipes > I have missed a more current technique for doing this > sort of thing). Also, this seems to be doing some > unnecessary work if the webcam can be forced to > serve up the correct pixel format from the beginning. > > Thanks. > _______________________________________________ > Linux-uvc-devel mailing list > Linux-uvc-devel@lists.berlios.de > https://lists.berlios.de/mailman/listinfo/linux-uvc-devel > _______________________________________________ Linux-uvc-devel mailing list Linux-uvc-devel@lists.berlios.de https://lists.berlios.de/mailman/listinfo/linux-uvc-devel
