Hi Julien,
Yes, real time video input and output, composited with graphics is one
of SGI's primary markets. And dmSDK is being implemented to insure
that that functionality works smoothly and reliably. We are taking
what we learned from the O2, Octane, and Onyx video and graphics
systems, and applying them to the cross-platform architecture of
dmSDK. You should visit the http://www.khronos.org site for more
information on OpenML, which enhances the audio, video and graphics
interoperability.
As to specific devices, there are plans for a variety of platforms,
both at the low and high levels. Currently, we are supporting both
IRIX and Linux with NT on the horizon. For the open source release,
we currently support standard open source devices, like video4linux
and the Open Sound System. For the developer's release, there is
support for the high definition video available on Onyx3000. We will
be announcing support for other devices in the coming weeks that
will be demonstrated at NAB/2001. More information can be found at
http://www.sgi.com/developers.
We'd also like to hear about what types of applications and platforms
you are most interested in, and what kinds of functionality would be
useful to you.
Regards,
Mike
Julien writes:
>
> Hi all,
>
> I just discover dmsdk and I'm sure it will be usefull for me
> in the futur.
>
> Just a question :
> Is dmsdk able to use mixed real time video input
> and 3d real time imaging for augmented reality applications ?
> Using the video as a background (directly in the back buffer)
> or as a texture (directly in the texture memory) for 3d real-time
> rendering (OpenGL) ?
> On wich plateform and video/3d cards ?
>
> More generaly, is dmsdk a good tool for mixed media like video
> and 3d ?
>
> Thanks,
> Julien.
> --
> ______________________________ _______________________________
> Julien ROGER | Institut Image / ENSAM
> Tel: +33 (0) 385.424.350 | http://www.ai.cluny.ensam.fr
>
|