From cdshaw@umbilicus.artsci.washington.edu Thu Jun 4 16:19:53 1998 Received: from burdell.cc.gatech.edu (root@burdell.cc.gatech.edu [130.207.3.207]) by lennon.cc.gatech.edu (8.8.4/8.6.9) with ESMTP id QAA25874 for ; Thu, 4 Jun 1998 16:19:48 -0400 (EDT) Received: from wheaten.hitl.washington.edu (IjK6CmoqyufNfKlGSVIooePTrbih1zRy@[128.95.73.60]) by burdell.cc.gatech.edu (8.8.4/8.6.9) with ESMTP id QAA01496 for ; Thu, 4 Jun 1998 16:19:47 -0400 (EDT) Received: from umbilicus.artsci.washington.edu (umbilicus.artsci.washington.edu [128.95.248.206]) by wheaten.hitl.washington.edu (8.8.8/8.6.12) with SMTP id NAA32670 for <3d-ui@hitl.washington.edu>; Thu, 4 Jun 1998 13:19:31 -0700 (PDT) Received: (from cdshaw@localhost) by umbilicus.artsci.washington.edu (950413.SGI.8.6.12/8.6.10) id NAA04880 for 3d-ui@hitl.washington.edu; Thu, 4 Jun 1998 13:19:31 -0700 From: Chris Shaw Message-Id: <199806042019.NAA04880@umbilicus.artsci.washington.edu> Subject: Re: LOD To: 3d-ui@hitl.washington.edu (3D UI List) Date: Thu, 4 Jun 1998 13:19:31 -0700 (PDT) In-Reply-To: <1.5.4.16.19970505123820.0fc77f20@pop.cc.ruu.nl> from "Ernst Kruijff" at Jun 4, 98 12:42:48 pm Reply-To: cdshaw@cs.URegina.ca X-Mailer: ELM [version 2.4 PL25] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Status: RO Ernst, > Can Chris maybe give a short definition on his perception on Fligtsim LOD? > It would clarify the situation for me, at least... The idea behind FlightSim LOD is that the computer graphics engine should display objects at a level of detail high enough for the user to perceive, but no higher. As a substitute for perceptual resolvability, one would typically use the number of pixels occupied by an object to determine the amount of effort that should go into rendering it. The more pixels that the object occupies on the screen, the greater the level of detail one ought to draw it with. For example, if a house takes up 5 pixels, just use one polygon (or a suitable multi-pixel dot). If a house takes up 1000 pixels, then many more polygons+textures are justified. At its simplest, FlightSim LOD uses a single metric to determine the level of detail that an object is to be drawn with, and that metric is projected image size (i.e. pixel count). Of course, you don't know the pixel count until you draw it, so a simple approximation is to compute the distance from the viewer to the object and to use the size of the object to calculate the approximate projected image size. In addition, one would maintain some sort of time control so that you might choose a less detailed LOD if you were running out of time on this update. So the punch line is: FlightSim LOD uses only the projected image size to choose LOD, and it's all automatic. It's supposed to be invisible to the user. The user doesn't control it. Architect LOD seems to allow LOD to be user selectable. Makes sense to me! -- Chris Shaw University of Regina cdshaw@cs.URegina.ca Assistant Professor http://www.cs.uregina.ca/~cdshaw