Your Mapping Process?

[Tobias Greenich]

Alright, so I'm sitting here at work (and working wonderfully hard, I assure you), and since I can't get my mitts on Hammer Editor, I thought I'd ask something that I've been wondering about.

What process do you use to take a map from thought bubble to completion? Where do you start, how do you plan? What have you found works, and what doesn't? Do you sketch out whole layouts, or just key areas and wing the rest? How do you approach putting together the basic map? What keeps you on task and keeps the map coming together?

Any advice for newer mappers trying to work on/through their first real maps?

 

[YM]

Think
Sketch
Think some more
Alter sketch
Orange map in hammer
Upload to server and test
Adjust orange map
Test again
Start textureing and doing displacement/ammo health/ important props and gameplay elements
Test once more
Revise any areas that don't quite work
Add lighting and details
Test many many more times
Make any last changes
Build cubemaps
Release
Relax
Celebrate
Find a random server with your map on, go on and play and be happy

(optimisation should be spread across all of the stages that aren't tests, even the 'think' stages

 

[Snipergen]

Couldn't think of a better way youme just said.

 

[Fyfey]

Ah, very informative... cheers Youme. I'm heading in the right direction... apart from putting lighting in near the start!

If I miss out the light_env, will the map be light enough to navigate?

 

[YM]

ahh, sorry forgot to say that 'basic lighting' is all part of the orange mapping process. You add basic lights so you can tell if an area is going to light the way you want, Its more important to do this when you want a certain lighting effect in an area like a spectacular shadow (see the upcoming black mesa mod with some of raminator's shadows here and here)
of course for various early tests you may want to turn lights off all together (no VRAD in compile) but thats up to you.

If you have no lights at all, including a light_environment the map will default to fullbright, which makes it plenty bright enough to navigate.

 

[Scotland Tom]

For some reason I discovered that Youme's method (which is probably the most common) doesn't work quite as well for me. Instead I take a slightly different approach.

1. I still sketch out my idea. I may do one sketch that I'm happy with and then map it out and iterate it from there, or I may do a few sketches before I'm satisfied.

2. I jump into Hammer and block in the first major area. For my current map that meant blocking in the first battle area up to the first CP of my dustbowl style map. Blocking in this first area, or any other area, involves optimization. From the first steps I take in Hammer to the last I'm designing thinking about areaportals and hint brushes. I'm ensuring everything that can be a func_detail is made so. If you don't design your map from the ground up with optimization in mind, it will come back to bite you in the butt down the line.

3. I do a quick compile and run around to see if the thing is worth continuing. If so, I move on. If not, I may go back to step one and sketch some more, then try again.

4. Assuming I didn't go back to step 1, I pick a point to start from - usually a spawn area because it's typically the first thing at the far end of a map. I block it in, texture and light it, add details... the works. This is probably where I differ from others the most. I get one area to a near beta-like state before I move on. Again, using my current map as an example, my Attacker spawn point is in a nearly complete state. Otherwise, however, I only have the first CP blocked in with dev textures and I haven't even begun blocking in anything else.

5. I repeat step 4 over and over again, progressing through the map. I'll complete a section, compile and run through it (possibly with a few friends), complete the next section and do the same.

6. The map should be in a state which is near-complete. I'll compile a beta version and try to get it up on some test servers. I'll get feedback and iterate as necessary while continuing to finish details and fine-tuning the thing.

7. I'll complete the map, release it for the public, and enjoy the product of my hard work.

I should mention that during most of these steps I'll compile the map to test out various features, lighting and textures. I may invite a few friends onto a private server in order to do this. So even in the Alpha and pre-Alpha stages I'm testing various iterations. The earlier I catch problems, the easier they are to fix.

So that's my basic process. A bit different, but it seems to work well.

 

[SaerisHawkeye]

hmmm, many of these steps get omitted depending on the map, but:

brainstorm - usually a theme or gametype will hit me and then i have to think of how to build a map around it.

research - very rarely, for a specific theme i'll go do some research and collect pictoral references to work from.

layout sketching - also very rarely, i'll sketch out the map progression before i open hammer.

open hammer - first things i do in every map is create one floor brush in the approximate size of the entire map, then in the center i place 2 props, a CP and a Heavy, for scaling references.

place some blocks - to get an idea of how the map is going to be laid out, I place giant block brushes in all the areas where there will be buildings in the approximate size they will be.

start cutting shit up - next i'll start making this blocks look more like buildings, give them floors, rooms, windows, doors.... all very roughly though.

color code - next, i'll go over my map and apply solid color textures to everything according to what that object is. this not only helps me identify objects which will have similar textures later, but for playtesting these base colors are enough to play with and see lightmaps with.

place the mechanics - from here, i'll start placing all the entities that will make the map work, like spawn rooms, flags, and CP's.

compile and test - at this point the map is complete enough that i should be able to play a round on it and win. i'll keep tweaking and testing until this works exactly the way i want it.

makin it look pretty - now i'll go back and do all my custom brushwork and lay out some props to flesh out the till now very empty map. i'll also usually add all the lights during this step aswell.

playtest again - now i have to go back and see if what I've just done is causing my map to lag a lot or if it impairs movement, allows for exploits, etc... usually this portion of the test is where i want to play with others and get feedback.

make it look prettier - so by now, people are getting settled in with the layout of the map and the shit in it, most of the brushwork is finalized, and so now it's time to go back and texture everything. this step i hate the most, because texturing frustrates me to no end. i'll also go back and tweak all the lights in case there were things I didn't like about them earlier.

add health/ammo - now i'll go and place health and ammo packs around the map, because, after the next compile, people are gonna be playing the "release candidate" of the map, any changes i make from here on are probably gonna be minor.

test! test! test! - get the map out there and have as many people play it and give feedback on it as possible. hopefully there won't be very many major changes i'll need to make, and preferably i'd rather not make any at all, lol...


So far, I haven't completely finished a map... I'm like that with a lot of things. Mostly because the further along in the process I get, the more bored I get with working on the map.