POV-Ray : Newsgroups : povray.binaries.images : Flight over mountains and lakes [~53KB Jpg] : Re: Flight over mountains and lakes [~53KB Jpg] Server Time
2 Oct 2024 00:15:05 EDT (-0400)
  Re: Flight over mountains and lakes [~53KB Jpg]  
From: Greg M  Johnson
Date: 27 Jul 2000 14:47:09
Message: <3980827D.85A674C0@my-dejanews.com>
The point is that every time I started to mess with HF's, I wallowed in indecision
about the resolution. If I made it too coarse, I could never do an animation later
of a flyby close up or balls bouncing off of it or a man hiking up it. If I made it
finer, it still wasn't going to be fine enough for whatever use I haven't thought up
yet.

Isosurfaces have the same appeal over HF's that vector art has over 2d  bitmap
painting, and really, 3D over vector and bitmap painting.  When I do my rolling
algorithm, I set up my isosurface in an INC. When I need to accurately compute
normals for the traces which determine particle trajectories, I set my
accuracy=0.00001, when I trace, it's 0.01 or 0.001.

Vector programs didn't render bitmap painting obsolete, per se..............

Christoph Hormann wrote:

> Bob Hughes wrote:
> >
> [...]
> >
> > You're right on the mark with that comment Greg.  The circumventing of a
> > need for smooth height fields to do this sort of thing may make the HF
> > obsolete, dare I say.  The one drawback is the need to get a fine enough
> > 'accuracy' apparently, among other things I might not know about, which can
> > slow things up.  The 'eval' and 'method 2' are superb additions.
> > Oh, yeah.  Yes, just the isosurface with ridged multifractal function and a
> > water plane, some ground fog, sky_sphere too.  A little detailed info:  24
> > by 24 unit square isosurface used with the camera traveling about 3 units at
> > a height of 0.41*y average above the water level.
> >
> > Bob
>
> Even though your scene shows the strength of isosurface terrain very good i have
> to disagree about heightfields becoming obsolete.
>
> IMO, their strength is where you apply many functions to the surface that would
> take very long to render as an isosurface and most erosion functions would be
> nearly impossible with isosurfaces, because they often use iterative algorithms.
>
> Christoph
>
> --
> Christoph Hormann <chr### [at] gmxde>
> Homepage: http://www.schunter.etc.tu-bs.de/~chris/


Post a reply to this message

Copyright 2003-2023 Persistence of Vision Raytracer Pty. Ltd.