From joan@gorgonio.oac.ucla.edu Fri Aug 31 16:17:44 2001 Return-Path: Received: from gorgonio.oac.ucla.edu (gorgonio.oac.ucla.edu [128.97.60.12]) by ks.uiuc.edu (8.11.2/8.11.2) with ESMTP id f7VLHhr14039 for ; Fri, 31 Aug 2001 16:17:43 -0500 (CDT) Received: (from joan@localhost) by gorgonio.oac.ucla.edu (AIX4.3/8.9.3/8.9.3) id OAA20726; Fri, 31 Aug 2001 14:17:42 -0700 Message-Id: <200108312117.OAA20726@gorgonio.oac.ucla.edu> X-Mailer: exmh version 2.0.1 12/23/97 To: John Stone Subject: Re: WireGL + VMD etc... In-reply-to: Your message of Thu, 23 Aug 2001 17:24:29 -0500 Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Fri, 31 Aug 2001 14:17:42 -0700 From: Joan Slottow Status: RO Content-Length: 2503 Lines: 46 John I finally got some pictures for you. I will be sending them seperately. I work at UCLA Academic Technology Services in the Advanced Technologies group. We do high performance computing and visualization support for the campus. OUr web site is http://www.ats.ucla.edu/. In the pictures you will see Aeint deBoer and myself. He installs all the software used by the chemists on our IBM SP/cluster and our Beowulf Cluster and he has helped them to build a Beowulf Cluster in Chemistry. I do all the scientific visualization support and a lot of the work in the visualization portal http://www.ats.ucla.edu/portal/ on the scientific side. I am the one who is responsible for the building of the experimental tiled display and all the work we have done on it so far. The Portal display is produced by 3 projectors. We usually run them from our onyx2. We also have a 3 headed PC and a 3 headed mac that we project. Each of them has a different archive. The archives are set ways to focus the projectors. There are 4 archives for the onyx2, flat plane, flat plane stereo, spherically mapped, and sperically mapped stereo. Any application can be displayed in flat plane. In flat plane the video format files of the onyx2 are set up with 3 windows and each pair of windows has a 160 pixel overlap. The signal from the onyx2 then goes to the digiblend box that blends the edges and from there to the projectors. In spherically mapped, we run Performer applications set up with 3 slightly-overlapping viewing frustums. This then goes through the digiblend and out to the projectors but gives a much more immersive feel. On the tiled display, with wireGL, I set up 2 configurations, with and without a 160 pixel overlap. Pieter has been trying to create an archive for the tiled display (one for each since we dont have stereo coming from there now), but so far hasnt been that sucessful. He worked on it some this week. In the pictures taken in the portal and sent to you, I used flat plane on the tiled display, i.e., 160 pixel overlap between the screens, and we used the onyx's archive in the vis portal. Consequently you can see 2 vertical bands, about 1/3 of the way in from each side, where the blending is not exactly right and the picture looks fuzzy. When we took the pictures on the tiled display itself, we did not use that configuration because you see the overlap (i.e., those 160 pixels twice) and it's anoying. More in a next message Joan