Page 1 of 1
Are coplanar brushes tolerable in this special case?
Posted: Sun Nov 02, 2008 12:59 pm
by monaster
Is it okay to have partly coplanar brushes (because it would need csg subtract to make them not being coplanar anymore) if that doesn't result in z-fighting anyway, because one of the coplanar sides will be textured with caulk and the other one normally with some kind of floor tile texture? Or can that result in exceptionally longer compiling times or something else not tolerable? Final bot play is intended to be implemented too and what I learned so far is that coplanar brushes can create problems for .aas files more easily than for bsp compiling; so this too should be taken into consideration.
(Editor is GTK Radiant for Q3A; bot files will be created with bspc2.1h)
Re: Are coplanar brushes tolerable in this special case?
Posted: Sun Nov 02, 2008 3:10 pm
by obsidian
That really all depends on the situation. As long as one of the coplanar brushes are textured with caulk, visually at least, you won't have any z-fighting.
As a general rule of thumb:
- Structural brushes should not be overlapping each other and should be more or less axial.
- Detail brushes can overlap each other and into structural brushes, but be careful of too much overlapping textured faces since this may cause overdraw.
Usually, I create all the visual stuff and convert all those brushes to detail. Then I add large structural caulk brushes between the walls to seal them off for hull and portals, etc.
If you post up a screenshot I can probably give a better suggestion in your particular case.
Re: Are coplanar brushes tolerable in this special case?
Posted: Sun Nov 02, 2008 4:28 pm
by Fjoggs
Let's bring kamarov back so we can have a caulk/no caulk discussion again. Those threads were a bundle of joy.
Re: Are coplanar brushes tolerable in this special case?
Posted: Sun Nov 02, 2008 5:59 pm
by monaster
[lvlshot]dead link[/lvlshot]
Erm... this is just a test picture to see if uploading works properly for me, the real screenshot will be posted in one or two days when I have more time, but thanks for your tipps so far, obsidian.
[EDIT: Killed the screenshot link so it's dead for now.]
Re: Are coplanar brushes tolerable in this special case?
Posted: Mon Nov 03, 2008 6:38 pm
by jal_
I, a lot of times, create a big structural brush in caulk and then insert smaller, detail, nonsolid brushes "drawing" the details of the wall. It's efficient cause the nonsolid ones don't split adjacent brushes (detail brushes unfortunately do at shared edges).
I'm using this less now, since I directly export the detail as a model now, if it wasn't made originally as one. But it is the same concept.
Re: Are coplanar brushes tolerable in this special case?
Posted: Sat Nov 15, 2008 2:47 pm
by monaster
[img]dead%20link[/img]
So this is what I meant with "partly coplanar": small parts of each brush are overlapping with others and though there's nothing to be seen inside Quake 3 Arena due to careful usage of caulk and normal textures, I'm not sure if it wouldn't be better to use "csg subtract" even if this is normally said to be one of the worst things you can do.
If anyone wonders how I constructed that thing: it's simply a hollowed brush which has been clipped on several edges in the first place.
EDIT: By the way, these are all structural brushes.
[EDIT: Killed the screenshot link so it's dead for now.]
Re: Are coplanar brushes tolerable in this special case?
Posted: Sat Nov 15, 2008 4:54 pm
by obsidian
So I take it that the outside of that shape is textured and the inside is going to be left as caulk? Why not just make it out of a 2 or 3 solid brushes clipped to those angles?
But in theory, you can convert the whole thing as detail and the slight overlap will be perfectly fine. Ideally you don't want to leave that as structural. If you need any vis blocking structures inside, add a square piece of caulk in the middle of it, fitting as closely to the detail object while remaining axial and on a relatively large grid.
Re: Are coplanar brushes tolerable in this special case?
Posted: Sun Nov 16, 2008 2:03 pm
by monaster
So I take it that the outside of that shape is textured and the inside is going to be left as caulk? Why not just make it out of a 2 or 3 solid brushes clipped to those angles?
Nope, in the end I want that thing to be a huge wall (inside an even bigger map) having some doors, through which you can walk into the middle of it; the texture outside was just placed to show the z-fighting inside the editor. Another reason to texture just a part of the outside wall was that on the inside there are no problems with z-fighting whatsoever and I just wanted to show the problematic areas this way. That means, finally, the whole brush will be textured once I'm done with it.
The second main aim is to have that structure surrounded by some environment that is completely accessible to the player; additionally this huge-wall-brush on the screenshot above will be clipped in the future so you can even walk in at least one long hallway inside it. So this could affect the whole structural-detail question, or am I mixing things up now?
In case this sounds a bit confusing... should I post another pic so it becomes more obvious what I mean?
Re: Are coplanar brushes tolerable in this special case?
Posted: Sun Nov 16, 2008 5:49 pm
by o'dium
Wont that produce light bleeding quite badly...? Not up to date on the older lightmapping stuff these days lol.
But yeah, if the outside is whats seen and the inside isnt needed, you can just make it from a few solid blocks.
Re: Are coplanar brushes tolerable in this special case?
Posted: Mon Nov 17, 2008 3:44 am
by obsidian
The outside is fine, but on the inside, you'll definitely have problems with overdraw (textured faces intersecting and extending behind another textured face). No amount of caulk will help you here.
Since the shape is pretty arbitrary anyway, I don't see why you don't just stick to a larger grid and bevel all the ends. Much simpler in construction and in the long run so you don't end up with other potential problems.
Re: Are coplanar brushes tolerable in this special case?
Posted: Mon Nov 17, 2008 8:54 am
by monaster
I really appreciate your advises, honestly, but unfortunately I can't change too much of this structure like beveling and the like due to a special look I need to apply to the final level. That's why my questions are kinda specific and leave not much room for other great methods bacuase I have to stick to a building method which fits best for this very special brush, sorry; I'm not too happy about that either.
At least I edited the scrnsht above so you can see pretty much everything which needs to be considered and that way you don't have to rely on my somewhat clumsy explanations.
[img]dead%20link[/img]
@o'dium: Gosh, I don't even know how light bleeding would look like, neither can I imagine how it could be affected by my brush structure. But as far as I have been experimenting with different light compiling commands the results were quite tolerable and didn't pose much of a problem, luckily.
[EDIT: Killed the screenshot link so it's dead for now.]
Re: Are coplanar brushes tolerable in this special case?
Posted: Mon Nov 17, 2008 9:52 am
by Kaz
Light bleeding occurs when you stab a fixture.
Re: Are coplanar brushes tolerable in this special case?
Posted: Wed Nov 19, 2008 1:34 am
by monaster
"stab a fixture"? Sounds strange... I'm not sure how this would look like - perhaps anyone got a screenshot? Yeah, I am a complete beginner in that respect.
Regarding my problem: since I'm still not sure what method would be best to avoid possible compiling problems (both bsp and aas compiling processes), I did something that surprisingly seemed best in that "special case" as I entitled it: I used csg subtract

Interestingly there were absolutely no problems like occuring phantom brushes and the like, since my huge brush had been created with "hollow" and thus the brushes were partly overlapping each other but at the same time they were perfectly aligned (see the grid window on the screenshot). That way, using csg subtract and thus getting 7 brushes with no coplanar sides, I can now do everything I pointed out with that weired drawings on the 2nd screenshot like adding an additional hallway inside the structure connecting all brushes. Plus, I don't have to think about texturing and z-fighting anymore since there are no more overlapping brush sides. Especially to obsidian always reminding me not to use csg subtract: it worked here, but I try not to use it anymore, promised! Thanks for all the other suggestions, too!
Re: Are coplanar brushes tolerable in this special case?
Posted: Wed Nov 19, 2008 1:47 am
by obsidian
I never said don't use subtract... ever... just that there are very few reasons to use it when other methods will usually have much better results, and not to use it if you don't really understand how to compensate for some of the problems it may create. Your situation may be one of those times when subtract might be somewhat useful. Just be careful of T-junctions, you may end up with a couple.
Re: Are coplanar brushes tolerable in this special case?
Posted: Wed Nov 19, 2008 1:57 am
by monaster
I'll take this into account, thanks!
Re: Are coplanar brushes tolerable in this special case?
Posted: Sat Nov 22, 2008 2:02 am
by pjw
This thread confuses me horribly.
I don't understand why you're leaving the overlapping brushes (and they are overlapping, "coplanar" means having faces exactly on the same plane, not what you have here); it just takes a few minutes to fix, and get everything back on the grid with no overlaps and have almost exactly the same shape (if you use a relatively small grid). Overlapping, in general, is messy and just plain bad, unless you're talking about detail brushes and structural brushes using the caulk hull method.
I just made a copy of your structure in radiant (click for large):
[lvlshot]http://pjw.planetquake.gamespy.com/screenshots/hollow_before.jpg[/lvlshot]
and then I used mostly edge-dragging, with a couple of uses of the clipper, to end up with this (click for large): [lvlshot]http://pjw.planetquake.gamespy.com/screenshots/hollow_after.jpg[/lvlshot]
Seriously, it took, like, five minutes, and there are now no overlaps, you can texture all the inner and outer faces as you like, and you don't need to worry about any sort of off-grid weirdness. Your brushes in the initial screenshot may be on the grid, but you'll probably have problems when you compile because the places where the brushes intersect (e.g. that top-most inside corner) are probably not going to be on the grid.
There's nothing wrong with using hollow or subtract, as long as you make sure to put the resulting brushes back on the grid afterward, and get rid of the overlap by using edge-dragging, the clipper, and/or snapping it all to the grid.
Re: Are coplanar brushes tolerable in this special case?
Posted: Sat Nov 22, 2008 6:25 pm
by wattro
pjw for the win... everything he said is gold
Re: Are coplanar brushes tolerable in this special case?
Posted: Tue Dec 02, 2008 10:35 am
by monaster
Amen.
The main reason, why I started that thread (if I remember correctly) was to find out if anyone knows the best method of building and/or editing that brush so it won't produce any compiling problems afterwards (IF there was a chance that bspc couldn't cope with some brush constructions). Fortunately the problem got solved, thanks to everyone involved!