From ernst@kwetal.ms.mff.cuni.cz Thu May 28 14:41:46 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 OAA00742 for ; Thu, 28 May 1998 14:41:42 -0400 (EDT) Received: from wheaten.hitl.washington.edu (p9hK6tb3BhOdwURHO+xxtBXnClk4pLKF@[128.95.73.60]) by burdell.cc.gatech.edu (8.8.4/8.6.9) with ESMTP id OAA23176 for ; Thu, 28 May 1998 14:41:35 -0400 (EDT) Received: from relay.accu.uu.nl (relay.cc.ruu.nl [131.211.16.32]) by wheaten.hitl.washington.edu (8.8.8/8.6.12) with ESMTP id LAA27023 for <3d-ui@hitl.washington.edu>; Thu, 28 May 1998 11:41:28 -0700 (PDT) Received: from hydra.cc.ruu.nl (hydra.cc.ruu.nl [131.211.16.28]) by relay.accu.uu.nl (8.8.8/8.8.7) with ESMTP id UAA33596 for <3d-ui@hitl.washington.edu>; Thu, 28 May 1998 20:41:27 +0200 Received: from pop.cc.ruu.nl (hydra.cc.ruu.nl [131.211.16.28]) by hydra.cc.ruu.nl (8.8.7/8.8.7) with SMTP id UAA111094 for <3d-ui@hitl.washington.edu>; Thu, 28 May 1998 20:41:13 +0200 Date: Thu, 28 May 1998 20:41:13 +0200 Message-Id: <1.5.4.16.19970428203656.0dff93f0@pop.cc.ruu.nl> X-Sender: l9339493@pop.cc.ruu.nl X-Mailer: Windows Eudora Light Version 1.5.4 (16) Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" To: 3d-ui@hitl.washington.edu From: Ernst Kruijff Subject: LOD in architecture (2) Status: RO At 10:26 28-05-98 -0700, you wrote: > >> Ernst said: >> About LOD: architects have a bit of different view on LOD than CG people.. >> LOD is often reffered to as returning to a higher instance in a type >> instance tree, during top down design. This can hardly be dealed with, with >> a rendering technique, since it far more is dependable on a object grammar. > >Are you saying that there's a semantic difference in Levels Of Detail >in the architecture world that makes Computer-Graphics/Flight Sim LOD >management hard to do? Yes, I do. My current perception on CG LOD (and I might be wrong, do not misunderstand me!) is that it progressively (or regressively) can show detail in a graphic. So, it is an overall gradual refinement of graphical detail - first you see landscape, then you see trees and finally (??) you can see leafs. In architecture, LOD is referred to (as described before) as a type with instances. So, returning to a lower level of detail is reffering to a type which at a CERTAIN PART OF THE OBJECT lacks an element. Looking at architectural LOD, you cut off a certain branch of the total tree of parts which combine to an object, you don't cut off all branches at a certain level, as with flightsim LOD. Or am I wrong? >How many levels of detail would you typically see in an architectural >setup? Now this is up to a grammar, when it would be used. I have no idea how many levels a normal grammar has..I could check it out. >> far more is dependable on a object grammar. > >Is this a small matter of writing the appropriate grammar interpreter, >or is there more to it than that? Yes, and no - IF you would be using a tool with a specific grammar, this can be made. However, most of the conceptual design tools I know don't use a grammar, so you can not simply use a grammar interpreter. In my thesis I developed a theoretical framework for a 'locational undo' - the system should keep track of refinements in a certain part of the design, which is possible in a top down process. When an user wants to approximate, he can undo some developments in a particular part of the design, without killing developments in other parts of the design too (by referring to a previously saved model). To be sure I am telling the right thing about LOD, I'll check it out by a real architect (since I am not a real architect....) - Ernst .................. Ernst Kruijff ................. Westersingel 9 .............. 4101 ZG Culemborg ................ The Netherlands ................ (0)345 - 519397 .. e.p.c.kruijff@stud.let.ruu.nl .... ernst@kwetal.ms.mff.cuni.cz .. kwetal.ms.mff.cuni.cz/~ernst/