Hollowed-Out Brush vs. 6 Sided Rooms

(CaptM) #1

I may be about to make some enemies here with this next statement.
I’ve done some work with the Unreal Editor (please save the boo’s) and when making maps, most mappers use the technique of creating one large brush and hollowing it out, then placing all features inside.
I’ve come across one tutorial for GtkRadiant that follows this method, but every other one goes with building 6 individual walls for the shell and building all features inside. Is this considered the preferred method for the Radiant editor or just mappers choice? I’d really like to know on this one.

Thanks All.

Ok, now you can hiss and boo about Unreal (I do love their editor though).

(chavo_one) #2

From what I understand of unreal editing, the schools of thought are completely different.

In unreal editing, you take a solid block, and carve out a world, much like a sculptor would. In q3 editing, you are taking a void and adding in structure.

In its essence, q3 editing is additive, while unreal editing is subtractive. So if you go about using a subtractive method in an additive system (q3 editing), then you are essentially going against the grain, and things will not be optimized.

([rD]MrPink) #3

The reason most mappers make 6 walls is to not have overlapping brushes, as that looks horrible, and using the hollow button makes a lot of overlapping brushes.

(DeAtHmAsTeR) #4

overlapping structural brushes = longer vis times

apparently something to do with, the engine sees the overlapping brushes as a CSG carve…

(JIM_BOB7813) #5

Doesn’t the hollow button prepare the brushes for mitering?

But then most of the time I use it for my skybox, which has to be structural, therefore it shouldn’t be mitered.

Yeah, with unreal you have to subtract your world from the void, then add brushes into your hollow space. But the thing I don’t understand is, how can you edit the size of your skybox/map after it has been subtracted.

Can you merge it up again? By adding void?

(redfella) #6

6 brushes > hollow (always)

:banana:

(G0-Gerbil) #7

shrugs Do it either way, they both work.
I tend to build individual brushes to make up rooms, but then that’s because it’s incredibly rare for me to have a cube room, so the hollow option is a bit irrelevant.

(damocles) #8

There are two reasons why you should not use the hollow brush for most rooms…

  1. The ends. If the room is relatively simple in shape then you need to miter the corners. Hollowing will make mitering a longer process because the walls don’t line up correctly. I can give you more info on mitering if you have no idea what it is.

  2. Vis portals. This relates to mitering but also to experience. If you don’t understand how the vis process creates portals for visibility calculations, find some tutorials (there’ll be plenty around) it is a very important aspect of Q3 mapping. The walls made by hollowing create vis portals in an unpredictable manner. When you make the 6 walls by hand, you have total control over the vis creation process, meaning that with experience you can learn how to make more efficient levels which means you can make higher poly counts in scenes which results in more pleasing eye candy for the end user.

I sympathise with you though bud, I came from unreal to Q3 as well. At first I thought it a pain in the ass to make 6 walls for a room, but in practice, you will make far mroe interesting and varied levels when you have to make the walls by hand. When I mapped for unresl, it became too easy to make blocky levels because you always end up extracting cubes. I actually find now that I know how to map well in Radiant I can map maybe twice as fast as I can with UnrealED and I was pretty handy with it. Radiant is probably the best level editor I have used (and I’ve used a lot). Give yourself the time to learn it and you won’t be dissapointed - it’s bloody marvellous.

(CaptM) #9

I want to thank all of you for your help and thoughts on this it really helps a lot.

I am new to editing in Radiant and have never submitted a map in Unreal, but have been playing around with the different versions of their editor for a few years.

One thing I’d like to say, whether it be UnrealED, Radiant or any other of the publicly released editors, YOU GUYS ARE THE GREATEST, It’s great that there is so much help out here for noobs to ask questions and get honest help and replies (and straightforward critiquing), if it weren’t for the great communities, we would never see some of the incredible maps and mods that have come out in the last few years.

My hat’s off to all of you.

(G0-Gerbil) #10
  1. Vis portals. This relates to mitering but also to experience. If you don’t understand how the vis process creates portals for visibility calculations, find some tutorials (there’ll be plenty around) it is a very important aspect of Q3 mapping. The walls made by hollowing create vis portals in an unpredictable manner. When you make the 6 walls by hand, you have total control over the vis creation process, meaning that with experience you can learn how to make more efficient levels which means you can make higher poly counts in scenes which results in more pleasing eye candy for the end user.[/quote]
    You double checked that by looking at the portals created by both methods? They look the same to me. Dunno about whether it takes longer or not to compile, but I dread to think what kind of map was created where this became a significant factor.

And where’s all this stuff about mitering coming from? That only needs to be used on CONVEX corners / edges (and not even then if one end is caulked). Concave it doesn’t matter in the slightest, and a basic hollowed cube is nothing but concave.

:???:

(bsimser) #11

Don’t want to get too off topic here, but regardless of the complexity of the shape of the room, mitering is the way to go. You don’t need to mitre if the room is simple, you should mitre wherever you can.

As for hollowing a room and mitering taking longer? I don’t see where walls don’t line up when you hollow. The hollow command uses the grid size to determine the width of the walls it creates. Then you just go in and mitre as normal along the grid boundary. Either way (hollowing vs hand-bomb) mitering is done the same way (clip on the grid boundaries).

Anyways, the principal is that you tend to end up making boxy’er rooms with hollowing out vs hand creating them. Personally I work from floorplans so I always build walls and add roofs and floors later in the process so I’m not “boxed” in but it’s personal preference. I don’t see one technique over another having any techical difference.

(damocles) #12

The walls don’t line up because the ends of some walls overlap others. This causes light leaking on thinner walls.

And mitering should indeed be done. Unless of course the outer faces aren’t going to be seen in game in ehich case you can just caulk them, but mitering would still help with lighting.

I mentioned mitering of simple walls, because if the area you are trying to encase inw alls is a very complex shape, sometimes the mitering of every corner can cause more polys than it removes. Although the lighting point always stands.