Terrain
#3293 posted by . on 2005/02/06 15:33:10
Here's some shots of some terrain I'm working on using the triangle method. Generally I'm doing this in seperate groups of tris and then join all them together at some point.
www.phait-accompli.com/q/crap/ter1.jpg
www.phait-accompli.com/q/crap/ter2.jpg
www.phait-accompli.com/q/crap/ter3.jpg
www.phait-accompli.com/q/crap/ter_ed.jpg
In the 1st shot on the grey bit theres shadows, this is because that group isn't finished - some tri points aren't met with other points to provide a cleaner terrain - don't want the player getting snagged on this, although you could place clip brushes.
FUCK
#3294 posted by . on 2005/02/06 15:37:18
Terrain
#3295 posted by starbuck on 2005/02/06 17:50:29
in the last shot I can see a vertex meeting the middle of an edge in the top-down view; surely that's not a good idea, as it might not line up ingame?
Kinn
#3296 posted by . on 2005/02/06 17:54:38
Could you circle the area you see this in? Worldcraft reports no problems, and neither does quake. I've made a flatter, bigger mound recently.
A Good Rule Of Thumb
#3297 posted by Blitz on 2005/02/06 21:21:34
When using the triangle method is to never have more than three triangles sharing the same point...otherwise it can look too artificial.
There's a good tutoral map at gamedesign.net for doing the triangle method.
#3298 posted by - on 2005/02/06 22:30:19
"triangle method" is called 'trisouping'.
"Hi My Name Is Sean, I Work At Raven"
#3299 posted by Blitz on 2005/02/06 22:50:51
I am the definitive Quake resource on this board because they pay me to glue models together at Raven. But when the Bobs stop by Madison, I'll have the same fate as Michael Bolton.
All Right, All Right
#3300 posted by Lunaran on 2005/02/06 23:12:09
"triangle method" is also called 'trisouping'
Is that better?
Phait
#3301 posted by Kinn on 2005/02/07 00:37:31
Are you confusing me with Starbuck again? (Hint: same colour name != same person)
Ahhh
#3302 posted by . on 2005/02/07 00:54:23
D'oh
Terrain With No More Than Three Tris Sharing Any Given Point?
#3303 posted by BlackDog on 2005/02/07 03:15:49
Good rule of thumb? Try "geometrically impossible".
Six, maybe (four is the minimum). I'd be more inclined to fuck rules of thumb off entirely in this case.
#3304 posted by starbuck on 2005/02/07 03:40:03
Are you confusing me with Starbuck again? (Hint: same colour name != same person)
No, but if you have the same 1st letter and the same number of letters in your name, then you're the same person ;)
Starbuck
#3305 posted by Kell on 2005/02/07 08:49:04
Well I was only implying the confusion was caused by nick color and not.....
oh shit
Heh
#3306 posted by pjw on 2005/02/07 11:37:29
Try "geometrically impossible".
You just aren't trying hard enough
AguirRe
#3307 posted by R.P.G. on 2005/02/07 15:26:43
I got a *** WARNING 22: Healing degenerate edge at (43 -168 224) in all 3 hulls in a map I'm working on. Your tooltips say to not really worry about it, but I'm wondering why I'm getting this error now when I haven't gotten it before. I have indeed changed a couple things, but none of them were in the region that gave the warning.
That's Not Unusual
#3308 posted by aguirRe on 2005/02/07 16:06:16
having such warnings pop up or disappear while changing seemingly unrelated stuff. It can be the editor shuffling things around in the map file or qbsp balancing on the edge due to the float calculations.
Most likely you don't have to worry about it, otherwise check out the position in hull 0 and see if you can find any off-grid vertex. Hulls 1/2 will probably follow a fix for hull 0.
You could also try Tyrann's new qbsp; it has a lot of changed logic that might help.
OMG!
#3309 posted by distrans on 2005/02/07 17:16:49
No, but if you have the same 1st letter and the same number of letters in your name, then you're the same person ;)
Starbuck is Shambler!
OK Then
#3310 posted by Blitz on 2005/02/07 17:17:02
I just wasted a good hour of my night trying to prove a point, but it was all worth it.
You'll notice that no more than 3 of the "terrain" triangles share any point. You'll also notice that you have the ability to move through some of the brushes, despite the fact that there are no func_illusionary brushes :(
It will eventually compile without those brushes that you can move through if you play with the size and style of the troubled brushes a bit. The .map file is included as well as a .bsp, so have at it.
http://blitz.circa1984.com/terrain.zip
P.S. I suspect that this map file would work flawlessly (i.e. without the move-through-able brushes, in Q2 or Q3)
#3311 posted by Kell on 2005/02/07 17:51:28
You'll notice that no more than 3 of the "terrain" triangles share any point
You'll notice the pertinent inclusion of quotation marks around the word 'terrain' :P
No more than 3 triangles share a vertex because this isn't terrain, it's a path.
Thanks, AguirRe
#3312 posted by R.P.G. on 2005/02/07 19:28:57
.
Triangle Method :)
#3313 posted by Ankh on 2005/02/08 01:28:08
Hey - this solves many of my problems!
Ha
#3314 posted by BlackDog on 2005/02/08 03:19:14
OK, I retract "geometrically impossible", since I admit that you've got some "terrain" here, and that of it's vertexes, none has more than 3 attached triangles. But that's only becuase...
a) I'm a very sweet, lonely guy;
b) "Terrain" is a term flexible enough to encompass not only general solutions to the problem of tiling an area with triangles, but also freaky special cases like this;
c) You're cheating.
PS Nice maplet...especially the trigger_changelevel. :p
Blitz
#3315 posted by Tyrann on 2005/02/08 04:01:36
This map compiles perfectly first time with the latest tyr-qbsp. No modifications required.
Distrans
#3316 posted by starbuck on 2005/02/08 07:17:57
Starbuck is Shambler!
Go map.
Blitz
#3317 posted by Mike Woodham on 2005/02/08 10:22:33
Was the point simply to show '3 sharing a point' or were you looking at something else to do with terrain?
I have more than a passing interest in building terrain so am keen to learn anything even if it's not everything.
|