VMD-L Mailing List

From: John Stone (johns_at_ks.uiuc.edu)
Date: Fri Oct 30 2020 - 18:49:32 CDT

Hi,
  Sorry I'm late replying on this, I've been exceptionally tied up.
VMD, Tachyon, and the VMD built-in OptiX version of my Tachyon code
all use a consistent model for their handling of VMD material properties,
transparency, and shadowing. OSPRay (written entirely by Intel)
is based on a very different material and shading model,
so VMD maps its internal material and shading parameters
to the closest matching OSPRay paraemters.

Where things are most different are in the cases where shadows
and transparent surfaces are involved.
In these cases, the OSPRay 1.x "scivis" renderer takes a very different
(simplified) approach compared to the one I took in all of the renderers
that I wrote. OSPRay 1.x and 2.x both have a second higher quality
"pathtracer" rendering mode which uses physically based rendering and
a far better approach for rendering transparent geometry and shadows.
However, the higher quality comes at a cost in speed.

In the case of OSPRay 1.x, the behavior of materials were different
among the two rendering modes, and I didn't have the time budget to
write-and-test two completely different rendering paths to
be able to expose this in VMD. OSPRay 2.x has a "pathtracer" renderer
also, but the OSPRay team (in part due to my requests) has made it possible
to have the same material properties usable in both paths.

In the future, I expect to be able to ship a VMD with OSPRay 2.x
and give the user the choice of rendering path to get either the
speed benefits of the fast path, or the quality benefits of the
full path tracing renderer back-end. I will likely stick with OSPRay 1.x
until this version of VMD ships, however, as I'm waiting for some missing
features (cylinders, cones, etc, which are in OSPRay 1.x but were removed
temporarily in early revs of OSPRay 2.x) to be reinstated in OSPRay 2.x.

Best regards,
  John Stone

On Fri, Oct 23, 2020 at 08:27:11AM +0000, McGuire, Kelly wrote:
> What are the differences between Tachyon-Optix and Tachyon-Ospray? Optix
> seems to render faster, but there appears to be some differences in
> rendered images as well. For example, if I render a molecule with AO and
> shadows enabled and set the material to transparent, the image quality and
> appearance is quite different for Ospray vs Optix. Ospray looks more
> realistic and higher quality. However, Ospray seems to ignore any
> lighting when the material is set to brushed metal or metallic pastel but
> Optix does render the lighting. Ospray appears as though the molecule is
> in the shadows with these materials even if I turn the AO Direct up to max
> and all lights on.
>
> Dr. Kelly L. McGuire
>
> PhD Biophysics
>
> Department of Physiology and Developmental Biology
>
> Brigham Young University
>
> LSB 3050
>
> Provo, UT 84602 

-- 
NIH Center for Macromolecular Modeling and Bioinformatics
Beckman Institute for Advanced Science and Technology
University of Illinois, 405 N. Mathews Ave, Urbana, IL 61801
http://www.ks.uiuc.edu/~johns/           Phone: 217-244-3349
http://www.ks.uiuc.edu/Research/vmd/