From the latest 3D Mark 05

Moderator
Posts: 916
Joined: 2002.10
Post: #1
each blade is calculated in real time using vertex shaders... We should have people on the mac side doing experimental stuff like this (someone make this NOW)

[Image: post-14-1097212480.jpg]




same with this one:
[Image: post-14-1097212380.jpg]

details of the scene Wrote:The vegetation on the ground is dynamically distributed where it's needed, according to the camera movements. Its level of detail is also dynamically altered depending on the distance to the camera. The other key interest in this scene is the lighting and dynamic shadow system. This scene is really ideal for showing the benefits of perspective shadow maps-- the new shadow mapping technique used in all three 3DMark05 game scenes. We'll have more on perspective shadow maps later.

The ground material is like the metal material in Return to Proxycon, but with added diffuse, diffuse detail, normal, and normal detail maps. The rock surfaces also have a specular map. The tree branches are a modified metal material without a specular map and with a diffuse cube map and no bump mapping. The sky material does procedural light scattering.

The moonlight is directional, generating dynamic shadows using a 2048x2048 resolution depth map. The illuminating firefly is a masked point light, throwing shadows using a 512x512x6 cube depth map.
Quote this message in a reply
Luminary
Posts: 5,143
Joined: 2002.04
Post: #2
Doing this sort of thing on the Mac is difficult, due to poor support for the shading extensions (read: fix the ARB_vertex_program and ARB_fragment_program bugs NOW, or give us GLSL).

The grass one should be relatively easily doable -- is there an online paper somewhere about how it was done (lighting, movement, &c)?
Quote this message in a reply
Moderator
Posts: 529
Joined: 2003.03
Post: #3
OTOH, note the fps is 12. Maybe this isn't fast enough to do in an FPS game yet.
Quote this message in a reply
Moderator
Posts: 916
Joined: 2002.10
Post: #4
Daniel_Lurie Wrote:OTOH, note the fps is 12. Maybe this isn't fast enough to do in an FPS game yet.
3D Mark is a test on shaders and graphical displays of what games will do in 1 - 1.5 years down the line. (the demos and stuff don't even have true physics or collision detection)
If you look at 3D Mark 2001, you see a bunch of stuff that is almost downright standard today.
Quote this message in a reply
Moderator
Posts: 1,560
Joined: 2003.10
Post: #5
What, that grass isn't a photograph? Love

Alex Diener
Quote this message in a reply
Moderator
Posts: 529
Joined: 2003.03
Post: #6
ThemsAllTook Wrote:What, that grass isn't a photograph? Love

Alex Diener

Weren't you going to do a lawnmower game, Alex? How's that for inspiration?!
Quote this message in a reply
Sage
Posts: 1,232
Joined: 2002.10
Post: #7
I think that top pic wins the "in most dire need of antialiasing" award.
Quote this message in a reply
Moderator
Posts: 916
Joined: 2002.10
Post: #8
arekkusu Wrote:I think that top pic wins the "in most dire need of antialiasing" award.
may be.... but can you imagine how much that would kill the fps!!! LOL
Quote this message in a reply
Sage
Posts: 1,232
Joined: 2002.10
Post: #9
It's hard to imagine without trying it. That scene looks vertex-limited, not fillrate. So FSAA x4 might have a barely noticeable impact.

Ya know, if people are interested in getting some advanced GL demo stuff to try porting to OS X, the source for GPU Gems is available here.
Quote this message in a reply
Moderator
Posts: 916
Joined: 2002.10
Post: #10
arekkusu Wrote:It's hard to imagine without trying it. That scene looks vertex-limited, not fillrate. So FSAA x4 might have a barely noticeable impact.

Ya know, if people are interested in getting some advanced GL demo stuff to try porting to OS X, the source for GPU Gems is available here.
thanks for the link!

I've been meaning to grab all the game programming and the gpu gem... but very much strapped for cash... maybe if they were some how included annually in a programming contest dedicated for programming... hosted by this site... and was named uDevGames
Quote this message in a reply
Member
Posts: 79
Joined: 2002.08
Post: #11
Nice images indeed. The grass stuff is very easy to do though.
Like OSC I wish Apple would get vertex and fragment shader bugs fixed. It's very annoying and filing bug reports is pretty useless since nothing seem to happen with them anyway.

Very frustrating to say the least...

KenD

CodeBlender Software - http://www.codeblender.com
Quote this message in a reply
Sage
Posts: 1,232
Joined: 2002.10
Post: #12
The bugs get fixed, it just takes a while (longer than one OS revision) in some cases. Remember that when it comes to OpenGL, Apple isn't totally in control. It's up to the ATI and nvidia Mac driver guys to fix most of this stuff.

And FWIW, a bunch of my bugs (ATI stuff mostly) did get fixed in relatively short order. So go ahead and log the bugs!
Quote this message in a reply
Member
Posts: 79
Joined: 2002.08
Post: #13
I have filed a bug about fog being broken in verrtex programs on some cards. Granted I have no idea what's under the hood but one would think somehting like that should be a very easy fix. The fog coord value set in vertex programs is simply ignored on Radeon 9800 at least, so the polys are rendered without fog.

Trying the same thing on 10.2.8 with a Radeon 8500 locks up the machine completely. Unfortunately I can count on this one to never be fixed.

KenD

CodeBlender Software - http://www.codeblender.com
Quote this message in a reply
Member
Posts: 268
Joined: 2005.04
Post: #14
Random stuff:

http://www.futuremark.com/products/3dmark05/

http://www.futuremark.com/products/3dmar...creenshots

Quote:Feature Test - Vertex Shader (Complex)
This illuminates, but above all transforms a large number of grass straws. Each straw is skinned and bent separately, more towards the tip of the straw, like real grass straws waving in the wind. The straws are waved according to a fractal noise calculated on the CPU, but it is highly optimized to decrease the influence of the CPU performance on the measurement. The grass is kept at a distance from the camera, offering a less interesting visual effect, but this is necessary to decrease the influence of fill rate to the measurement.
Quote this message in a reply
Post Reply