GoldSource Mapping Tips Created 18 years ago2006-01-02 09:43:42 UTC by Kasperg

Bump for goodness.

Great thread this, looking back. We should take all the (useful) tips and put them together on a page somewhere on TWHL.

An all-round mapping tips tutorial?

Well, the idea of this thread was to make an ever-growing list of tips derived from personal experience. Not all of them are rules that should be applied in every single case.
I think we should leave it as just a sticky thread for now.

And a little tip. Thanks to gravity, we usually move on a horizontal plane. But do game levels need to be always like that?
Think of the cliff levels in HL1 or the way to the monastery in Loast Coast. The main gameplay surface of reference became a vertical one...
I haven't seen many examples of playing in the ceiling (maybe crawling through ventshafts in HL), but it's something worth considering!

CTRL-Selection-Boxing
(A short, but painfully drawn-out tutorial)
by rowleybob

K. We all know how to CTRL select/deselect verticies in Hammer, but did you know you can also CTRL-Selection-Box them as well?!!1 I didn't until today, and I can't believe I never noticed it before!

You know how when you're trying to select a bunch of verticies, but sometimes you can't select them all in one swipe of the selection box-- without grabbing unintended ones?

Well my friend, you needn't worry about this sort of problem if you have CTRL-Selection-Boxing in your bag of tricks!!1 Here's how it works:

1. Select an object/group of objects with multiple verticies. Notice our example... it's on an angle which makes it impossible to select half of it's verticies in one swipe. Here, we're missing the two groups of verticies on the right and the left.

2. In this one, we can't make the selection without grabbing those center ones, which we absolutely don't want.

3. So, why not let's try our new method?! Start moving the selection box around as many verticies as you can--without selecting ones we don't want.

4. Now, hit <ENTER> to select them. Next, while holding down CTRL, draw the selection box around the remaining verticies.

5. Hit <enter>, and you've done it!!1 *win

You can do this as many times as you want to select/deselct verticies. Also don't forget you can adjust the selection box in the other 2D windows, to further help you refine your selections

Here's one decent tip that every aspiring mapper shold know:

• Whenever making stairs, steps, or raised platforms, go 8 points high and not the common 16 points high. 16 point steps look blocky, old, ugly, and unrealistic. You can also half the size of the step texture you use to fit properly.

• When making secret rooms, make sure you have enough time to work out what the heck you're gonna throw in there! Lolz.

Naked ladies works very well.

Hahah.. Agreed.
fy_porn is one of the most downloaded maps @ cs-maps.org.

Oh god...I knew there had to be a CS porn map...

Texture-Application Tips for Noob
By rowleybob

Texturing is a little weird at first--I still dont really understand "face" or "world" lol, but like anything, you improve as you go along. A couple of tips I can think of:

?When scaling the texture, set the x and y values--under the Face Properties window(shift-a)--to what is listed under "Size". (also in face properties.) In most all cases, the size is also the intended scale for the texture.

?Use the "Justify" buttons also under face properties. They are very useful, and will make your texturing life much easier. For the most part, don't use the "fit" button until you know more what you're doing.

?If a texture looks funny, even after all this stuff, try checking the "Face" box for the heck of it. Sometimes that's all you have to do, but I must admit, I dunno why.

?Clicking the "Hide Mask" button makes darker textures easier to see. Use the mask to make bright textures easier to see.

?If you like the way the texturing on one wall looks and you want to transplant those exact settings to another wall: select the texture and while holding the ALT button down, RIGHT-CLICK the wall you want to look the same.

?Experiment with checking the "Treat as one" and "fit" buttons together. I use them all the time, but I can't think of any solid rules for them, but they can be useful.

Once again, the more you play around with all the Face Properties options, the better you'll get. Don't be afraid to try new stuff... it could make your mapping life much easier

I believe the World and Face settings work something like this:

World will stretch the textures to make them fit across multiple brushes well, even if they're at different angles. This makes it a really great setting for making landscapes, but if used incorrectly (like all the way around the corner of a rock face), it will stretch out the texture. You probably shouldn't use this setting for normal architecture, because it's more rough way of texturing.

Face will basically just align the texture most accurately to the individual face, making this the best setting for normal architecture. It doesn't take into account the faces around it very well, even with "Treat as one" checked. The upside to this setting is that the textures aren't stretched out when you shape a brush at an angle.

Remember, this is just what I've found from using it. This could be entirely wrong, but it's always worked for me like this.

Precompiling:

Before big compiles--or any over 5 minutes really--, I always now "precompile" before doing a full VIS and RAD go. By "precompiling", I mean I run fast VIS, and no Rad.

Doing this catches any really nasty coplanars/disapearing brushes, or really any other important performance flaws, before the Full compile.

Oh how it sucks to wait for an hour compile--or longer god forbid--, only to find you have disappearing faces, and have just wasted that hour for no good reason.

Precompile = win

Sounds about right, srry.

'World' will align a texture to one of the side-views, e.g. if you looked at the brush in one of the 2D views, you would see the texture unscaled - if the 2D view had a textured mode, of course.
'Face' will align the texture to the direction of that face itself, and therefor gives no scaling artifacts when looking straight at that face.

World works fine for square maps and for some lightly sheared walls, and gives less misaligned edges. Face works better for more organic structures, but can be hard to align. Using the Alt texture trick here helps, but it can only do so much when you're dealing with complex objects.

Sounds good, the precompiling. Happened to me a few days ago, although it was just a 15 min. compile (yes, full compile).
Nice Ctrl+selection box trick. Do you have to hold the Ctrl button all the time or just press it while hitting enter?

HL2 - When adding in a stairway using handrail props, add the handrail prop and a couple of steps (if not all of them) of the desired width and height before finishing the room or adding the floor to which the stairway leads.

If not, the odds are you're going to have to reposition the entire stairway or upper floor to make the top and bottom of the handrail appear correct.

Just wondering, but i've made REALLY complex maps and the compile times never went anywhere above 20 minutes.. what systems are you guys running??

Anyway.

Mapping tips... yeah..

I map with a laptop, and as annoying as it it, i learned to utilize my hotkeys well

The D key zooms in, whereas the C key zooms out. the [ and ] keys change the grid size. i no longer use my mouse to navigate the 3-d view, as W, S, A, and D work perfectly. (Just mess with the acceleration settings in the options) Um i also changed my default entity to a light instead of an Info_player_start, because im usually placing more lights in my maps than player starts.

and if you're doing a lot of node placing. You can change the default to a node and then just click everywhere in the 3-d view to place them. saves a lot of time and it comforms to any fake-displacement you created.

If you make big, wide-open complex maps, without func_walling overly werid structures, or having your textures scaled very small over a large area, you'll run into HUGE compile times.

So basically, if you map the right way, you shouldn't really run into huge compile times imo.

Take my dm_hothworld map-.rmf is included. There are 3 ATATs outside the hangar, in a big open area. They're pretty big, but I scaled the textures x4 to try to keep the compile down. Still, I found by not turning them into func_walls--they're fairly complex for VIS-, vis time streteches from 5 minutes to so long that I had to abort the compile because I wasn't willing to wait that long!!

Captain P: Make the first selection normally, and then you only have to hold CTRL down while you make the second selection box. Once the second selection box is in the 2D view, you can let go of CTRL and rescale the selection box at will--without "losing" your inital selection... Try it!

It's nice to see you around btw

Thanks rowley.

Big area's simply result in a lot of faces to be lit, and complex structures make the VIS process (often needlessly) complicated, thus taking a long time (often without much gain, because small details hardly ever block other area's from being visible). Trying to keep the world brush architecture relatively simple is a good idea, not just for the sake of face splitting.
I wouldn't call this 'mapping the right way'. It's simply using your knowledge of the inner workings of the compile tools for your advantage. As if a compile program or process is always designed the right way... and that's exactly why it's so usefull to know how they work.

@espen180: give the vertex tool a try. Pretty usefull in situations like these.

When you're making turns in pipes the regular way, without any fancy curves or anything, it's easiest if you keep the brushes intersected before you clip them on their ends, like this:When you're all done with your set of pipes, you can clip them all at an angle, and it'll go really smoothly. If you had clipped the ends before you had finished sizing all the pipes, you would have had to go back and realign all the vertices to make sure they matched up, because they would have been stretched out.

This is just one example of the many ways you can make it easier on yourself by planning ahead while mapping.

Use clipping to make cylinders smaller than 16 units wide, or make a big one and shrink it.

Bad idea. resizing larger cylinders will almost always result in off-grid vertices, and your brush won't end up as you see it it the editor. Any vertices that are off-grid are snapped back to the nearest grid point during the compile, so it won't always look exactly the same.

Your other suggestion would be a little difficult, although it is another way to make cylinders if you have no other way. The fact is, beyond a simple eight sided "cylinder" of sorts (it would actually be octagonal), most people won't be able to form perfect circles using only the clipping tool.

Srry you reminded me of an example.
If you DO happen to have a brush that is clipped at an angle on one side, but is flat on the other side, instead of stretching it using the handles, Vertex Manipulate it.
Drag a box around the outside white boxes of the other end, and drag it as you wish.
Quick and painless.

editing...

More Useful Map Vault Comments
By rowleybob

K. Map comments are great, and nobody likes them more
than I do. There are however some easy ways to make
your comments better.

In general, explaining WHY you like/dislike something
about the map is extremely more useful to the mapping
guy than not.

Sometimes it's hard to explain why you like/dislike
something, and you only "feel" a certain way. In that
case, play the map again or think a little bit harder
about it. If all else fails, simply tell the author
you can't put your finger on why at least.

Some tips to help the n00b and 133t h6x20rs alike:

THIS MAP IS AWSOME!! YOURS ARE ALL INTERNETS

While very encoraging, it doesn't tell the mapping guy
anything he can use to improve his map. Even if you
can't find anything wrong, or anything that needs improving
iyho, at least describe WHY said mapping guy is winning
all teh internets.

Nice Map

A variation on above. Such descriptive ratings
also have less merit. I shudder to think that I actually
used to rate maps like this DON'T DO THIS!!1

This maps sucks. 1 star for you. *fail.

Perhaps the worst, most annoyingly useless rating/comment
you can give someone. Give some well-thought out reasons
why you think the mapping guy fails and his map are suck.

If a map is unusually bad I don't even bother rating it.
What does anybody gain by rating somebody's first map 1 star?

Some Damn-fine Good and Better examples:

Good:
Your textures suck. Scale them down a little and they'll
look better.

Better:
Your textures are awfully scaled, especially the ones in the
long south corridor. Scaling them much above twice their
normal value is usually a bad idea. Check out this tutorial
at...

Good:
You map is blocky as hell, and under detailed...

Better:
Your map is blocky and boringly detailed. The main hangar
area is essentially a box and the trim and light fixtures are
also big boxes. Try smoothing out the edges of the room with
wedges or other structures, and reduce the size of some of
your detail stuff. I can build some example props for you
if you like... pm if interested

Soe, as you can see, it's all very easy to be a better
map-rating guy. The mapping guy benefits as well
as the rater--having that warm fuzzy feeling in your belly
after giving a nice set of comments, among other things I
hope!

That tip are teh suckage, nub.

Lol. Only kidding. I hate it when reviewers come out with stuff like that as well. It's not constructive critisism. It's just shit. Nice tips.

Although we don't usually make long singleplayer mods, I think there's an important aspect that level designers often forget.
In a time when there are countless singleplayer games out there, you need to make yours special. Some games do this by adding addictive and variated gameplay (strange weapons, nice effects) but most important, they keep the player motivated. If a game has no surprises, it becomes predictable and boring. I've seen it countless times. A lot of Unreal II and the parts of Quake IV I've played suffered from this.

On the other hand, you have other games in which you are motivated to follow the story somehow. For example, since the moment you see the Citadel in HL2 after leaving the train station, you are subconsciously thinking that you will eventually make it there. You get far away from it in parts of the game, but when you return to City17, the Citadel is closer and closer as you complete the game, giving you a sense of progress and never thinking: When will this killing-the-bad-guys end? (think about the middle levels of F.E.A.R...)
If there's a reason for the player to progress, something to look forward to, your game/mod will be much more entertaining and interesting.
From what I've played so far of PREY, they motivate the player by showing the brief encounters with the rail-transport thing carrying his girlfriend around the ship. Other things such as portals, antigrav walkways and wondering what has happened to the other humans in the area also contributes in some way.

I'm not exactly sure how we can make use of these principles, but I'll try to do something for my minimod. I can't compete with the high-tech style games out there but imagination is sometimes better than more polygons and high-res textures.

Concerning your comment regarding they way the citadel was used in HL2 I think you hit the nail right on the head there. The same can be said for Episode 2. The only difference is that you're trying to get out of the city and cannot actually see your goal but the Citadel is always behind you, hanging over you like a shadow, reminding you to GTFO.

I would love to be able to create this sort of tension in my mods and now that you've pointed that out I know exactly what to do.

Cheers. Great tip.

Good point.

The same applies to multiplayer, although a bit differently. I prefer to spread items across the map so, that it balances against each other and against the advantages and disadvantages in the map itself. This gives players a good reason to keep moving from place to place in order to stock up, or to find combat, etc. In other words: I try to create a map that continually gives players small goals to work towards to. Using landmarks to make area's easier to learn, and therefor allowing players to learn where they can replenish certain needs best, is a good idea.

On the other hand, it is possible to make a mistake by giving the players a big view of what they are about to do.
Sometimes, in several games I've played, there has been a big area where you can see what your goal is but can't reach it. That's OK if you don't know what you're going to have to do in order to get there (some levels in the first Tomb Raider work that way) because it can make you want to get involved. IF the steps needed to accomplish your task are very clear, it can produce a contrary effect because the player already knows what he's going to see in the next minutes, sometimes even the enemies he will have to fight.
As I said somewhere in this thread, the key is to tell the player he is in point A and has to get to point B (motivation), without telling him what he will find on the way.

When in Texture Application Tool Mode, you can instantly select all the faces (the FACES, not the brush itself) by holding the right shift button while left clicking the brush. This is usefull when you need to select a large number of faces (but not all) of an VM'ed brush, or a cillinder to do texture allignment.

Hold the left Ctrl key to add face selections.

You saiid "ass face"!!

LOL!

I always misstype the word "add". :/

"ass" is much more funnier!!

I tried the texture thingie (the left shift key works for me as well).

It would be nice to use this on a grouped brush or entity, forgoing the need to ungroup it or hit IG, selecting it, and then opening Face Properties: Your way is faster

Ok enough!

Some mod edit my post, please. %)

SRSLY
LOL
MOAR

Interestingly, using the method that Muzzle described actually doesn't select all the faces automatically, it just reverses whatever mode of selection they were currently in. So if you had any individual faces selected before you shift+left clicked the brush, those faces would then be deselected, and all the previously deselected faces would be selected. It can be pretty frustrating, at times.

No, srry, you're doing it the wrong way, i geuss. Try this.

• Make a brush or two
• Go into TATM (you know)
• Hold shift and left click a brush. All the faces of that brush are now selected.
• Hold shift and left click the other brush. You selection of the first brush will be lost, but now you have all the faces of the second brush selected.

If you want to add to the selection, just hold Ctrl while you shift + left click the other brush.

srry: I concurr with muzza (I just tried it again lol). Maybe different versions of hammer are the cause? Do you use 3.5? (I do, an I think muzza does.)

ZombieLoffe: BRAHHAHAHAHAHAHA!!1~

Hey... That's odd. Maybe I just got in the habit of using shift+ctrl+left-click all the time, in which case what I described would happen. Silly me.

Being able to reverse your selection like that is pretty neat, which is in itself a tip. Way to to Srry.

Yeah, you're right. Like, if you were making a crate, and you wanted the top and the bottom of it to be textured differently from the rest, you could apply one texture to those faces, and then use the shift+ctrl+left-click technique to select the rest of the faces. Another example would be to stagger textures across faces.

Okay, so I can't really see how you could get much use out of this, but it is kinda neat.

A tip, or rather a habit I have when making a map with a light_enviroment:

The first thing I do after deciding the theme of the map, is deciding the angle of the sunlight. You might think it's an artificial way of working, but things turn out much better if you already know how light is going to flow into a room. What parts will be dark and will need other types of lights, and how the shadows will spread across the surfaces. When you know the direction of the light, you can actually place other elements to achieve the desired shadow across a certain floor etc.
It might seem this method takes away the possibility of making versions of the map at other times of the day. However, if you are mapping for Source, for example, it will be much easier to keep track of which faces you can have a more detailed lightmap grid scale!

You should already have a good idea of that by the time you finish your abstract, and test it out when you box the basic geometry.

Also, I found a new way of plotting levels. Draw lines that first represent the flow of a map, then box it in with a quick layout sketch. I think scribbling a few lines just plain works the best. just blot our some lines and connect them as you see fit naturally.Good way to spawn ideas quickly.

Better for DM rather than Single player though right? Maybe SP combat areas would be cool done that way though. Sort of non-linear.

Repetition & uniqueness
Ok, so we know that when we are mapping, using repetitive structures or texturing is usually considered bad practice. On the other hand, it's one of the most important ways used to build real places, and mainly a fast one.
So let's say you are in a big room, full of pillars everywhere. If they look bad, the fact that there are many of them only makes it worse. But what if they look good, or they are slightly different from one another? Then you've probably found a formula to make a reasonable and cohesive place with a minimum amount of effort (most will be copy/paste).
Now imagine you've just finished the architecture for that room. It might still look boring. That's when you would introduce other parameters for uniqueness: You could, for example, make the place really dark, and add very focused lights in key places. What at first was just a bunch of pillars in a room, might look like a mysterious forest of concrete, in which being below a lamplight and far from it makes all the difference.

In other occassions, we are forced to make things that look alike. Think of a city, like City 17. All the buildings have a similar style. But are they the same? Nope. Some are broken, some have a set of awnings, some are taller, others are completely in shadow, etc.

The point is. Even though real-world structures and architecture is full of repetitive elements, as level designers our duty would be to change that feeling since it only gets worse with a game's simplifications (such as using tiled textures for big surfaces). Decals, sunlight, artificial light, props... you name it. Just give a player a reason to be there, to make him/her forget the place was made by you, and instead it exists on its own.

To do a landmine without the trigger_once and env_explosion, you can just do a func breakable with a high explosion magnitude and have the touch or pressure flags checked. when done in a hole it looks like a crater when blown up.

^^ This here is VERY simple and I've never thought of it! Nice one, TJB!

I don't have sound on my computer, so it doesn't matter to me.