Level Design

[Zarcath]

Anyone here with level design experience that could chime in with what basic steps you do for planning?

I'm wrapping up the first map I started working on in Unreal3. It was pretty much inspired by 2fort, there's a bridge in the middle, a pit under it that leads to a basement, basement leads to base central area. I didn't want to recreate the entirety of 2fort with the extended basemap intel room, so I terminated the level after the central courtyard, just added a side room with a flag cap point (like Badlands intel point). The map is largely designed as a CTF PVP map...

My next map I'm planning to do a Deathmatch PVP map. Thinking a circular multi-tiered level. Going to draw it out first this time, instead of winging it within the editor. After that I'd like to put together a singleplayer PVE map with some environmental puzzles...

I'd like to get more experience level designing, the more the better. But I get the feeling that U3 maps will go largely unnoticed (can they be converted to be used with other games that use the same engine?), and I'm really not entirely familiar with the U3 gameplay style (as opposed to TF2 or even CS:Source), infact I've never really been a fan of the Unreal playstyle (quick and dirty deathmatchs).

So I figure if I do a TF2 map I could get some good feedback/experience. The last time I tried to use the Hammer editor I only managed to make a box, where as with U3Ed I had a much easier time putting together an entire level. I haven't touched the Hammer editor since...2.0? Is the newest edition as user friendly as U3Ed?

[TwiNCannoN]

I've got going on 3 years experience with Hammer now so if you need any help I can assist, although I haven't tried any TF2 mapping yet. Interlopers.net - Half-Life 2 News & Tutorials is your best bet for learning all things hammer related though.

The editor itself is very friendly both in usage and of course the fact that help is always easy to find. It hasn't had any major changes since it's conception, though (actually it hasn't had many big changes since it was called Worldcraft, really). The style has always been the most intuitive to me however, but it may suck coming back to it from a UE editor since yeah, the way you construct maps is basically backwards (although BSP makes way more sense to me than CSG).

[Zarcath]

What game have you been working on?

[TwiNCannoN]

Dystopia. A Cyberpunk Game for the Source Engine. New version should be due out in a month or so as well

[Zarcath]

Ohhh sweet. Yeah I've played Dystopia on and off, what maps did you make?

[TwiNCannoN]

Fortress and Undermine

[Believe]

Hammer isn't very user friendly, but once you get the hang of it, it isn't that bad, but still takes a lot of getting used to and messing around with.

Can always watch tutorials, but I don't ever really do that. I just fuck with it until I figure it out.

[Zarcath]

Nice! Did you do the original Fortress too? Fortress and Undermine are two favs. Those two and Assemble, Silo..Broadcast..Vaccine. Wow I want to play right now lol.

If you didn't do the original Fortress, how did you go about with the remake? What sort of planning did you do for Undermine?

[TwiNCannoN]

Fortress started out as a 3rd party map and I got on the dev team through it (amazingly, it was really shitty back when it was 3rd party :P). Fortress was pretty much winging it, it went through a lot of changes and basically I consider it my "first real map" even though I've been mapping since way back in doom with doomED (the duke3d BUILD engine remains one of my favs).

In the original design for it it was a lot more like it's homage - 2fort, with the inside area being much tighter and smaller and basically a carbon copy of 2forts (I originally chose to base it on something I knew to make the learning process a bit easier). I ended up scrapping the entire inside area and redoing it for the public 3rd party releases and then once again for the v1 release.

Undermine was much more thought out and planned and because of that took much less time to make. I still have my notebook full of undermine designs. It still needed a bit of redesign in the end but that's the nature of the beast; for the most part the very first sketch still resembles the actual layout quite a bit.

Given that, planning stuff out is a really huge thing IMO. Especially so for Dystopia since the maps are absurdly complex being both objective-based and the fact you're designing with two worlds in mind (cyberspace). I have a bunch of other maps down on paper and once you sketch it down (and are comfortable with the editor of course) it's a weekend job getting it made in dev textures which is really nice since you can see right away if the map is going to be balanced and such. Which is another big thing I'm sure you know - block the map out first, detail only what you know is good to go. Even a simple mistake like gettinga room's proportion wrong can ruin hours of detail if you're not careful to pace yourself and make sure to check the map out periodically.

[Zarcath]

Which is another big thing I'm sure you know - block the map out first, detail only what you know is good to go. Even a simple mistake like gettinga room's proportion wrong can ruin hours of detail if you're not careful to pace yourself and make sure to check the map out periodically.

Do you follow any sort of mathematical guidelines when making areas or do you base it mostly off the feel of it? When I was working on my 2fort remake, I started out with the bridge/front area, and then pretty much based the rest of the proportions off of that. What "felt" right when I was running around.

Is that acceptable though? in your experience mapping? Should other experienced/good mappers "get" what you're doing?

Since I've played FPS's for so long and so many different types of them, I'm not sure if i'm meta-designing them too much (choke points, sniper nests that can leave you exposed from certain angles, corridor/door alignment for firing arcs). I tried showing it to someone who wasn't into FPS games and they didn't get it at all and were more interested if it looked cool or not.

[TwiNCannoN]

Well, there is certainly a lot of "meta-designing" you can do before even starting the map in Hammer. Most of it having to do of course with what you're designing. In Dystopia you're going to want to avoid a chokepoint that can be spammed with grenades and such too much, in TF2 you're going to try and implement some nice turret spots for engineers, etc. A lot of very minor things someone wouldn't notice can make a huge deal, for example aligning spawnpoints the wrong way and making players have to turn around every time they spawn.

I tend to be a perfectionist with numbers and lining up textures/geometry, which can be both good and bad. Although sometimes it can feel like a waste of time, it can also save your ass and make design changes a lot easier (i.e. one small example, in the EU3 area in undermine the ceiling is 32x32 unit squares aligned the to grid, and in the version coming in v1.2 there's a new vent exiting from the roof there with a dimension of 128x128 - since it all adds up I didn't even have to realign any textures or anything). It can especially save your ass with geometry since a wall being even one unit off can cause a leak in your map (generally the best thing to do with geometry is try and use as big a snap-to-grid value as possible, don't ever use 1 unit snap to make big pieces of geometry or turn off snap to grid).

I will say that getting proportion right is definitely a "feel" thing when starting a map rather than a math thing. One thing I'll do is make a bunch of "props" (basically model detail in hammer) and use the player models as the graphic - they'll be in the default arms-outstretched pose, but this way you can easily see how they line up with the geometry. Obviously this is a non-issue for a map for like HL2 single player since you can just use NPCs but for multiplayer maps it's a pretty useful method for assisting in getting proportion right.

Sorry for random rambling, just trying to spew out any tips I can think of.

[Zarcath]

Thats cool, not rambling at all, it's what I was looking for.

Do you have something new you're working on now?

I wanted to remake molten core...is it bad that I can remember the layout by memory? :P it's actually a pretty simple design.

[TwiNCannoN]

I'm putting the finishing touches on undermine for the v1.2 release at the moment. It's received some pretty heavy updating just like fortress did in v1.0.

I would've liked to get another map out with the release but none of the ones I've been working on here-and-there are close enough. I've got two fully blocked out in dev textures atm - one has punks assaulting an office building from a high floor, stealing data and escaping via the roof; another has corps raiding a desert rogue punk base and reclaiming a piece of high tech equipment that was stolen. Maybe they'll see the light of day eventually. Both are pretty hefty jobs art-wise though and I've been putting mapping off a bit lately to learn modeling.

Speaking of molten core I actually have an unfinished HL1 deathmatch map from way back based on Blackburrow from EQ1, only it's like a science facility or something. Always thought that layout would make a kickass DM map though.

[Cloud9_]

Originally Posted by TwiNCannoN

I'm putting the finishing touches on undermine for the v1.2 release at the moment. It's received some pretty heavy updating just like fortress did in v1.0.

I would've liked to get another map out with the release but none of the ones I've been working on here-and-there are close enough. I've got two fully blocked out in dev textures atm - one has punks assaulting an office building from a high floor, stealing data and escaping via the roof; another has corps raiding a desert rogue punk base and reclaiming a piece of high tech equipment that was stolen. Maybe they'll see the light of day eventually. Both are pretty hefty jobs art-wise though and I've been putting mapping off a bit lately to learn modeling.

Speaking of molten core I actually have an unfinished HL1 deathmatch map from way back based on Blackburrow from EQ1, only it's like a science facility or something. Always thought that layout would make a kickass DM map though.

I have such fun memories of Blackburrow Having a deathmatch map based on old EQ levels does sound interesting though

[splok]

Naturally, the better you plan before you start, the less trouble you'll have later. How much is worth doing depends on how complex you are making your level imo. It's never bad to break out the graph paper and plot out your rooms and major elements. Of course, you can go as rough or as detailed as you feel is needed, but no matter what you put on paper, the better you understand every detail of your level before you start, the fewer problems you'll have. Of course, it's possible to take it to an extreme that might be counterproductive as well. Last I had to spend a week doing an appx. 50 page level design doc for a level with 5 minutes of target gameplay and that I only have two weeks to build. Of course, my school really stresses documentation.

As far as actually building the level once you have it concepted (however you go about it), I do something like this:

First, block out all of the spaces that you'll be working with. Nothing fancy, just the dimensions with a single (or default) texture. Then load it up and run around it a bit. Make sure your room proportions feel like you think they should feel.

Once you're happy there, block out some of the major elements that you know you're going to need. If you need a railing, just but a solid block where it should go. If you're just dropping in models from the editor, do that for the big, essential pieces. Basically, anything that would seriously affect the player's path as he runs through the level needs a representation, even if it's just a box. Then run around the level and make sure it flows like you want and that the proportions feel right now that it's populated a bit. Also, make a habit of ducking and jumping near your objects... Getting the hang of making things the right height can be tricky.

Next, go through and add all of the gameplay elements, mobs, scripting, whatever. Imo, the safest way is to get the entirity of your gameplay in before you even touch the art. I do it the other way sometimes, but you'll be more reluctant to make changes to your structure once your art is done. This way, the lost work when making changes is minimized.

Once you're entirely happy with the way everything feels and the way it plays, then start on the art pass. Build whatever complicated BSP you need. Do an initial texture pass. Drop in your models. Etc. I would only do a basic pass at first to make sure everything actually looks like it did in your head.

Next add your lighting. Do this afterwards because which textures and models you pick can have a pretty huge affect on how the lighting looks in your level.

Finally, keep going back and making tweaks, putting in more details, and adding polish until you're sick of it.


Concerning the editors you've mentioned, we're currently working in Hammer, and Unreal3 is far more user friendly than Hammer. There is some debate over whether the LD's will get to their final solo project in Source or Unreal3, and if we have to do it in Source, there may be a revolt. Another option to look at if you aren't a fan of UT3 is Gears of War. However, making a single player map is typically far more work than a multiplayer for whatever that's worth. Radiant (Doom3, or CoD4 for example) is also preferable to Source imo.

I tried showing it to someone who wasn't into FPS games and they didn't get it at all and were more interested if it looked cool or not.

Sadly, that's all 99.9% of people will be interested in. That's the curse of the designer. When everything is perfect, it's utterly transparent to the vast majority of players. Its when they DO notice things that it's a problem, because most people only bother to notice design aspects that piss them off or frustrate them. I'm sure you see that in practically every discussion on this forum, and that's likely among the tiny segment of gamers most likely to have intelligent discussion on design.