Blocking

[Serberuss]

I'm interested to hear a bit about your process when you start a new map, especially since I think this is a bit of a weak area of mine. In particular I'd like to know how you block in your maps and how much you do it.

Currently my process is kind of get a bit of an idea, draw it down and see if I like it top down (often not the whole map), open the editor and do some extremely rough blocking in with huge blocks and everything basic (pretty much no shape just cubes), get an area done such as a bombsite and then detail it and move on.

When you block in do you add detail to the geometry? I'm not talking lighting and texting I mean creating the actual shape. So say if you were blocking in a part where a building will go will you create how the roof and roof supports will look, windows and doors and steps etc or just keep it very basic until you come to detail the map with texturing? Also do you do a section at a time or do you block the whole map at the start?

[NihiL]

Yeah, good thread, this is always a tricky and hard part of the procedure I'd say. Everybody has their own style I suppose, what I do is I sketch out... well... sketches in paint or photoshop first. Then when I go into Hammer, I put down simple brushes that represent the playable areas themselves, the 'air' between the walls, floor and ceiling basically. Just to get an idea. And I already use measurements like 128 units for a single story wall (ie. inside, like in normal apartments), twice or three times that for 2-3 story high buildings, 256 minimum for openings usually etc. This allows me to then go ahead and delete a brush, put in the walls, floor and ceiling as world brushes with dev/graybox textures and get the right measurements and clean brushwork. Then move on to the next 'air' brush, remove it and do the same. And then you can prototype that layout you get, move things around etc. and when you're happy, edit->select everything, apply nodraw texture and start texturing, putting in detail brushes and props.

This workflow basically enforces you to think about two things (proper scale [measurements for textures] and optimization [separation between world and detail brushes]) right from the get go. That's why I like it. But I'm interested to hear about other people's approaches as well.

Edited June 30, 2015 by NihiL

[text_fish]

After the obligatory pen and paper sketch, I block out all of the floor brushes, then I do a quick compile and run around (without bots) checking that the timings and encounter points are pretty close to what they were in my head. Then I throw down some basic walls and ceilings where necessary, and test with bots. Once I have this version, I'll often set a massive round limit and just let the bots play each other for a while whilst I'm doing something else, just to get a basic idea of balance. I'll then probably start blocking out a few details that might end up effecting gameplay (windows, cover, etc.) and then do further testing with bots and if possible with humans. In an ideal world I would then do a first pass optimization, blocking out the 2d skybox and a few hint brushes etc. but in reality I usually end up cracking on with more detailing and lighting than I should at this stage.

[Skybex]

I dont generally use pen and paper as I prefer to see things in 3d space as height can play a huge role. But I always make sure to have a theme decided and enough reference collected before I do anything. Then I get an idea of what I want areas to look like. "Site A should look like these 2 images, T spawn should look like this...". This way when laying the floor plan I can get spaces that work with the theme and i don't need to fight with the layout afterwards to get things working visually.

Once I get the floor plan done, I check all the lines of sight and see where any problem areas might be. What can be solved by simple cover vs what is flawed at in the layout and take steps to fix it. Then play with bots to find any problem areas. Areas that might be uncomfortable to defend, areas I felt i was killed unfairly, areas where I can see the enemy but they cant see me (slopes or other height stuff often causes this) and make sure to fix all these issues at the very early stages. Will also make sure the timings to the chokes are what I had planned. 

I will do this many times over fixing all the issues I can see while building the map up in a dev block stage or with very basic textures (no detailing) until I am confident not only the layout works, but the spaces work for future detailing. 

During this time I test often, doesn't always need to be a 5v5 but get as many humans as you can and lets the bots fill the rest. This way you can get human timings to chokes with pre nades and flashes as well as smokes while taking a site. even a 1v1 can help a lot with this. Additionally what can help is to block of half the map so the players and bots can only go to one site, this way you can check the chokes, cover spots and generally get a better idea of the site. So often when playing with bots they all march one way and your bots march the other and the bomb is down without anyone having seen each other. 

Only once I am confident in the layout will I do any real detailing.

[Serberuss]

Interesting, it seems we all do things a bit different.

I think I might start adopting an approach where I try to do a bit more geometric detailing without texturing and see how I get on with it. So kind of 50% of the way there perhaps so I get a better idea of how it will look when I come to playtest it. I think I'm also a bit of a sucker for focusing way too much on one area of the map and I don't tend to flesh out the rest of it quickly enough. This can cause issues when later on I realise my map layout isn't going to work at all because CT rotations are way too short for example.

Once I get the floor plan done, I check all the lines of sight and see where any problem areas might be. What can be solved by simple cover vs what is flawed at in the layout and take steps to fix it. Then play with bots to find any problem areas.

Do you not run into issues of not really understanding where your lines of sight are since you don't have any walls yet?