2 Questions...

[King Slayer]

First off, I was wondering if UT renders by zones, only what you see or renders everything?
Second, for my "visual interface", I have a Voodoo Banshee and a Hitachi TTX 17" monitor.
I don't know much about refresh rates and all of that, but I know your cards' refresh rate has to be the same as your monitors'. I run the game with all video settings maxxed, but my card and monitor settings are factory set still. I would like to know if I can give my system a kick in the *** , so to speak. If you can help, please do.

[King Slayer]

Oh yeah a third question, how do youget bot teammates to follow you of high ledges?

[Tetris L]

1. UT renders only what you see. Doesn't any 3D engine do that?? If you activate the "Time Demo Statistics" in the "Tools" menu, UT shows you your framerate and you'll see it going up and down depending on where you look.

2. There is a vsynch setting in the advanced options, but AFAIK, the effect is minor, so don't bother with it.

3. If you order a bot "cover me", he'll follow you as far as he can. What limits him is that he has to use invisible natigation points, that the map maker has put in the map. So if you use very tricky jumps or TL manouvers to get to areas that have no such points, he may not be able to follow you.

http://www.planetunreal.com/images/bullet2.gif
©leaner.CH.[$FCE2]
$FCE2 - Respawn Inc. - (http://www.respawn.onlinehome.de)
UT ©leaned (http://www.unreality.org/cleaned) - Tips'n'Tricks for UT Config and Gameplay

[The Lamb]

3D engines actually render everything within your field of view, and then everything connected with that. Your graphics card will also render this - eg your voodoo will render both the triangles you can see on a wall, and those on the player standing behind it who you can't see - because it sorts the z-buffer after rendering the everything. This means that if you are looking at a scene with a depth complexity of three, your graphics card has to render three times more than you ever get to see on your screen.

Some graphics cards, like the VDO powerVR series of boards, the Kyro and the as yet unproduced Gigapixel/3Dfx parts all use something called 'defered rendering'. They sort everything by z-buffer first then render, so only the parts of the scene you can see are rendered - eg the unseen player behind the wall is not rendered. This is a much more effecient way of rendering, which not only deals with higly depth complex scenes better than do normal graphics cards but also provides a more stable framerate.

I think you may have also been talking about 'Tiling', which is again only used by PowerVR boards. This technique involves splitting the screen into segments of maybe 16x16 and updating only those 'tiles' which need it using the defered rendering process.

http://www.planetunreal.com/images/bullet2.gif
don't worry, it's UTCentre (http://www.utcenter.com) Humour

[This message has been edited by The Lamb (edited 06-15-2000).]

[henke]

intresting....
this info can make me a better mapper...
i always thought that the card renderd the whole map and kept it in the RAM or something and then the player is just a camera that moves and shows what u see

http://www.planetunreal.com/images/bullet2.gif
If a million people say a foolish thing, it is still a foolish thing.

Anatole France [Jacques Anatole Thibault] (1844-1924)

[The Lamb]

Nope. When you vis a map you set down what places can be seen from different positions, and the normal vid card renders every vis'ed triangle within your FOV - including any which might be hidden, and doesn't render anything which is classed as unseen in the vis info.
Vid cards store the textures for a map in their memory, but not the whole map itself. These textures are then rendered onto the polygons classed as visible and the finished frame is stored in a frame buffer before being pumped into the RAMDAC and displayed.

So the card is the 'camera', not the player.

http://www.planetunreal.com/images/bullet2.gif
don't worry, it's UTCentre (http://www.utcenter.com) Humour