Various mapping questions

[Merc248]

1.) How does one make a smooth road going left, right, or a smooth inclination? I have a test map that I set up where I had a bunch of wedge brushes edging the road to the left, but I imagine that it would just cause the r_speeds to go up drastically once the map is filled with all of those turns.

2.) What is the height and width of a player model in HL2? How much allowance do you guys usually add to the player when determining how high a door frame should be, or perhaps when the person goes into a crawl space, how much allowance do you give in that situation?

3.) When you guys create a map, what is the first thing that you try to mesh out in the map before moving on to creating detail, adding more basic geometry, etc. For example, in the test maps that I've created so far, I tried to get the terrain first and then the geometry. In other cases, I try placing flat brushes for the concrete surroundings for the buildings, and then place huge blocks of brush to represent the buildings.

4.) Would you much prefer to create a map model and import it in, or try to create it as a prefab instead?

5.) Any sort of fundamental tips that you guys can give to a newbie mapper? Anything common newbie mapper fallacies that you guys can point out?

Thanks in advance.

[Fletch]

2)I'm assuming the dimensions will stay about the same at HL1 so 72 high and about 24 wide. Learning proper scale is pretty much a matter of trial and error.

3) Every mapper has thier own "process" that they feel comfortable with. I usually always started a map by coming up with a concept (usually inspired by photos or textures) and starting to sketch out a layout on paper. This let me get a handle on a rough layout as well as understanding the kind of scope I'd be working with. As I sketched a layout, I'd try to make note of what certain areas would look like, thoughts about paths and detailing, etc. The more I had in my mind to work with, the faster the process went. Then I'd just start mapping. I prefer to map by making one area at a time. Others like to start with big things and then work down to small details. It's just a matter of preference.

4) Model

5) Your first map will suck. Period. Your second map will probably suck, too. It's all about learning your skills. Resist the urge to keep revising your first map over and over again. It's important to be able to say "This is as done as it's going to get, let's move on." Other than that, read everything you can and don't be afraid to do some trial an error testing. If you're not sure how a certain entity works, make a test map that is a giant empty room with a light, a spawn, and your entity that you are working with. It will give you a chance to test and compile rather quickly until you get it working just right. then you can replicate it in your actual map. It's a quicker solution than to post on a forum and wait for a reply.

[Duff-e]

ugh.....the main reason i quit trying to map was because i either made a museum sized doll house or a doll house sized museum......those were confusing and lonely times

[JynxDaddy]

textures are normally the right size for their use if they are scaled 1, .25 or something in hl2.

eg: Door textures are how big a door should be, vent cover textures are how big a vent should be.

[Izuno]

indeed your first map will suck..and second, third etc. but you will get better if you keep at it.

more than anything, as you learn to map, getting feedback from people does help. learn to take criticism as feedback.

most of all, have a good attidute a don't be affraid to experiment. you will learn more by reading tutiroals and doing trial&error than by doing anything else.

[Merc248]

Sweet... alright, thanks for answering the questions everyone. Heh, sounds a bit disconcerting to hear that there's a high chance of your first through fourth/third map being shit, but eh... that's fine.

[Fullauto]

Heh, try all your maps been crap for 6 years

Mapping's all about experience. You'll screw up a lot, even later on. I look back at the stuff I was doing 6 months ago and am amazed I was so stupid.

[Fakedawg]

First of all - Hi, I'm new to this forum, but not new to mapping - I've been out of HL mapping for a couple of years due to work/school (and general "fed up'ness with mapping and dod), but I've recently picked it back up to brush up while awaiting the release of HL2. I saw some familiar names here so I thought I'd sign up ...

Now as for your questions:

1 - Roads with bends - I reckon you could use the arch tool - the fewer sections the arch is made up of, the less expensive it will be. It depends on the area how many you can afford. This goes for more or less everything - you allways have to weigh the detail against performance. BTW in HL 2 r_speeds aren't used as such anymore. You will see the term "expense/expensive" being used instead. Bind +showbudget to a key, then press that key in game to bring up a graphical presentation of the "cost" of various elements in your map - the higher the cost the more impact it has on performance. The elements in question will be things like: the poly count of world brushes (kinda what r_speeds gave you in HL), but also things like, models (e_poly count in HL), lighting cost, displacements (terrain), physics etc etc ...

2 - Think this one was allready answered

3 - I usually start out with a sketch of the map I want to make - nothing detailed, but a general layout. Then I usually create the layout in Hammer, and start working on the various areas. I tend to put in placeholders for the larger structures, in the form of square brushes (houses, cliffs etc). I do this in order to get a general feel for how the vis blocking will be, so that I can identify possible weaknesses and rectify these as early on as possible. Then I work on the main areas seperately, but I don't add all the detail. As I get the areas finished I do full compiles to check my vis blocking. At this stage I usually time the various routes from player spawns to flags/objectives to ensure that the map will be balanced in this aspect. Then I start doing the entity work, and finally I add detail and optimize at the same time.

4 - Models are usually less expensive than brushes, and the use of models is a lot heavier in HL2 than it was in HL. So for performance issues models should be used for detail - brushes for larger type architecture.

5 - Don't loose heart as you fail miserably in the beginning - you will probably have a bunch of great ideas as you start out, but as you work on your idea the limitations of either your knowledge/skill and/or the engine hits you smack in the face and you will feel like giving up all together - don't ...

Everyone experience set backs every now and then - you just have to learn from it and move on!

Here is some random advice:

- Get to know the engine - it's possibilities and limitations alike

- As a follow up to the first point: As you understand the engine (and program) you are working in, map smart - work around the limitations, you will soon enough develop a "bag of tricks" ...

- Read tutorials! If you know what you want to learn how to do, great - but even if there is nothing new you need to learn at the moment, take time every now and then to do tuts - you will discover new possibilities and techniques.

- If you get really fed up with what you are working on, either start another project you had planned and go back to it later, or take a break and do tuts that has nothing to do with what you are working on ...

- There is still a lack of decent tuts for HL2 - while you await their creation - do HL tuts, a lot of the brushwork tuts still apply to HL2.

- Figure out what most entities do. Usually you can work around a problem by finding the right entity, or by using an entity (or several entities in combination) you allready know, in a different manner.

- Don't let leaks piss you off. Leaks will be frequent in the beginning, and it will usually be a bitch to find them. But as you get more experienced tracking them down becomes quite easy.

- Save often! Make backups - on several medias. Keep backups of several versions of your map, so that you can go back to an earlier stage in the project if something goes horribly wrong ...

- Learn to model. For HL2 and other newer engines modelling is essential. Only being able to do brushwork won't be enough.

- Learn to make textures. Most of the time you can find all you need in publically released wads/packages (wads aren't used in HL2), but at times you need something special that no one else made. Being able to make your own textures can give your map a more consistent look as well.

- If you don't allready know how to model or texture - don't focus on that until your brush work reaches an acceptable level (we are talking months).

- Learn to do lighting well! Lighting can make or break an otherwise great map ...

- Allways remember: Gameplay has priority over "prettiness"!

- Playtest, fix bugs, balance gameplay, optimize performance, then playtest some more - repeat this process until the map is finished - only then actually release it.

- Don't make too large maps with too many alternate routes. Try to channel the players into action filled areas. People want to kill eachother, don't run a virtual maraton!

- Don't make too small maps with too few altenate routes. People don't want to be part of an everlasting nade spamming contest at a chokepoint. Make sure people can flank!

- Whenever possible, study other peoples work in order to get new ideas. Either by running around compiled maps, or even better, have a look at other peoples vmf's. It's great when people release example vmf 's (like Zaphod did with his dod_zaphod), you will learn a lot from it.

- If you borrow ideas, or use textures, sounds, models etc that other people made - make sure you are allowed to use it, and give due credit!

- Participate actively in community forums. Even though you are not a veteran mapper, there are allways someone who know less than you and need help with something that you probably can answer.

- Help support custom map servers for your game/mod of choice. A lot of people give up mapping because no one ever play their maps. Do this by either running a custom server of your own, be part of a team that runs one, or at the very least play custom maps yourself and tell your friends about it ...

- Voice your opinion of other peoples work in their "release" or "work in progress" threads - but do this in a polite and constructive way. A lot of time people just state things like - "that looks great", or "that looks like crap" - that doesn't really help at all ...

- Don't expect to be hired by a mod team or a game company in the near future. Probably the best way to get into a mod team is to start out as a custom mapper. If you develop skills, servers might start running your map in their rotation - then people will play it, and you will have something to show the team when you apply for a job.

- Finally - have fun. Mapping is supposed to be a creative, fun, challenging and stimulating experience.

PS - My experience comes from running a custom maps site/server for dod - as well as making custom dod maps myself.

[kleinluka]

Fakedawg,

good to see you here sir

[Fletch]

the old guard is crawling out of the shadows now that HL2 is out.

[Izuno]

Fakedawg,

good to see you here sir

I second that.

To expand on FakeDawg's point about examing the engine, I suggest playing HL2 from start to finish at least a couple times to examine the architecture, level of detail, entities in action etc. Also play CS:S to examine how they ported the CS maps into the new engine. Once original Source MP maps come out for CS and eventually DoD:S and other mods, examine them too. The more you study other peoples works, the more you understand what it takes to be original. We love original ideas around here.

Good luck and have fun.