Re: 3 Channel OTW display
Joaquin Casillas (jcasilla++at++glup.irobot.uv.es)
Thu, 08 Apr 1999 12:08:20 +0200
Angus Dorbie wrote:
>
> > Bemis, Suzie CEI-TACCSF wrote:
> >
> > We currently have performer running our front view OTW for a cockpit.
> > I want to add two channels for side views. Is this as simple as
> > creating two channels and offsetting the views from the front-view
> > eyepoint? And just updating the two new channel views when the front
> > view is updated?
>
> Yes, or you could simple chan share the position, they need to share the
> same scene and sharing stuff like LOD & earth sky (with fog) can also
> help. There are a few other tricks which are less well known like making
> sure that localviewer lighting is enabled in the light model to minimize
> lighting discontinuities across channels.
>
> >
> > Is anyone familiar with the ONYX 2 and the IR graphics? Do we need
> > the multi-channel option hardware to send the two new channels to the
> > new displays? Or is the hardware built-in? Or is my question too
> > vague and requires more information?
>
> You can get 2 channels from a basic system or if you go for the channel
> option you can get 8 channels. You always have the video bandwidth to
> drive 3++at++1280x1024_60Hz but obviously you'd need the 8 output DG for the
> 3 channels.
>
> Ensuring the right graphics channel is output to the right video channel
> is simply a matter of opening viewports in the right areas of the
> framebuffer to match the channel layout you create using the ircombine
> program.
>
> Cheers,Angus.
>
Hello. We are also managing a 3 channel display system (150º FOV, 3
projectors, curved screen). The projectors need an overlap area to blend
the edges of the adjacent channels, and we are not sure about the
correct way to achieve this.
One approach is to overlap the channels from performer code, letting the
left and right channel to have a heading offset smaller than the fov for
each channel. Another approach is to overlap the video channels with
ircombine, and have the performer channels with the same heading offset
and fov. We think the former is more correct than the latter, although
if it would be possible to achieve the correct effect with the second
approach we would be saving some extra drawing.
Any suggestion would be greatly appreciated.
-----------------------------------------------
Joaquin Casillas Melendez mailto:jcasilla++at++glup.irobot.uv.es
LISITT (ARTEC) Phone: +34 6 364 2253
Dept. Electr. & Informatics Addr: Poligono de la Coma s/n
University of Valencia Valencia (SPAIN)
This archive was generated by hypermail 2.0b2
on Thu Apr 08 1999 - 02:12:21 PDT