[AG-TECH] Re: Full Screen Vic
Gavin W. Burris aka 86
ga5in at psu.edu
Tue Nov 16 14:48:40 CST 2004
Another solution would be to use a hardware scan converter with a zoom
function. That way you can zoom in on any window of any application
on your desktop and only output the actual video playback sans
widgets. I've used the Averkey products in the past:
I do prefer direct out from a video card's s-video for color
saturation and clarity, but that isn't always feasible.
Also the Nvidia cards have an overlay option in Windows, so that any
app using video overlay goes full-screen on your s-video output.
Many Ayromlou wrote (on Tue, 16 Nov 2004 at 12:01):
> Hi All,
> After meeting with the AG-Art group at SC2004, it became clear that
> something strange had happened to my posting from 2 years ago
> announcing a revamped version of vic we had developed in house.
> Since almost everyone in the group wanted to find out about it, I'm
> reposting the link to this version.
> Keep in mind that this version was developed when AG was still in 1.0
> realm. What that means is that it works properly, but is obviously not
> integrated into AG as a service or anything. I've included the compiled
> version + source in the package (yeah you also get all the .o
> files....hehehe). It will compile on Suse 9.1 and 9.2 for sure (We
> originally wrote it on Suse 7.3 I believe so it should compile on older
> versions as well). My version of the executable (under vic directory)
> should also work (just in case you run into trouble compiling it from
> This version is really meant for either a Display machine and/or
> another machine to record stuff offline (through the svideo output of a
> graphics card for example). Everything works pretty much as standard
> vic, except clicking on stamp sized video's does not open new windows.
> When you run the software from command line (vic -t 127
> ipmulticastaddress/port), there is a fullscreen button that brings up a
> dialog asking for the following:
> 1) # of screens: This is the physical number of screens you want to
> span (# of heads). For a full blown node this will be 4 (1 for monitor
> + 3 projectors)
> 2) Video's/Screen: simple enough....if you choose 3 screens above and
> choose 4 video's/screen you end up with a 6 by 2 presentation (6
> columns, 2 rows).
> 3) width/height: how tall/wide is the canvas to be (ie: for 3
> projectors at 1024x768 it will be 3072x768)
> 4) x offset/y offset: where to start the canvas (ie: if you have a quad
> headed machine with 3 projectors and your admin screen is screen 1
> running at 1280x1024 then that will be the offsets.
> The easiest setup (to test) is to get a linux box going with dual head
> working (say 2x1024x768 screens). Run the software from screen 1 (main
> screen) and tell it to use 1 display and 16 video's with a width/height
> of 1024x768 at an offset of 1024x768. This should create a 4x4 grid on
> your second display that covers the entire display area.
> These modifications were done about 2 years ago as the need arose for
> our performace group (synthops) to record it's performances on DV tape
> for offline viewing. We wanted to have a clean interface with no
> widgets on it so this is what we developed. I'm providing the software
> AS IS meaning that I probably won't have time to provide much support
> for it (I can answer the occasional questions). The student who did the
> modifications was under a tight deadline so as you might have noticed
> he was not able to create a patch file. In a way he's ended up with
> another fork for the software. The modifications were written by Ben
> Bogart during his time with SynthOps. We are releasing our
> modifications to the software under it's original license.
> Many Ayromlou
Gavin W. Burris aka 86
Senior Systems Programmer
Penn State Visualization Group
More information about the ag-tech