I was wondering about how other people go about mapping.

I'm sure we all know that planning is great idea, but what do people do after that?

I personally make the outlay of my map (the streets basically), then add details (starting with interiors then window frames etc). That's still a big job though so I make it in areas making sure each one is sealed before making the next one joined on to the previous one.

Another thing I do is compile often. So after an 'area' is complete I'll compile and run around it. However this is one practice I'm wondering about. Compiling is taking a long time on my first source map. So I'll do a day's work, compile and then I've got to spend another day waiting for the compile. It's really slowing my progress as I don't get that much time to map. However I'm loathe to stop doing that as the last map I did 'broke' after I did a load of work without compiling and then I had to go back. Is compiling often a common thing, I might be worrying unduly?

I was also wondering how long compilation can be expected to take under source. How long would the official maps take to compile?

All my best mapping is done at random, without layout, I know a layout is a fundamental part of level design but I very rarely have enough creativity in my to actually take this step, mapping at random is suitable for me.
Sometimes I will get a good idea for a portion of a map, like building placements etc.. I open up hammer and throw down a few nodraw brushes then actually make the houses in that position later.

I know of at least 2 people that draw their layout, scan it and turn it into a texture, they then import it and lay it down in hammer as a brush and base their mapping on that, kinda the same concept as 3D modelling programs/.

for my map I found an ariel view of the area I wanted to map and turned it into a material then used that to place nodraw brushes where the structures were and then added the hedges and trees.

if you don't have a scanner you could always plan your map out in a paint prog then create your material from that

I've never drawn any kind of plans for any of my maps. I just "wing" it when it comes to design and layout.

I compile quite often. 90% of the time with vrad turned off, to save loads of time.

:)

Sometimes i will draw up a quick layout on z peice of paper then start mapping from that but changes always occure

I've tried to do it in the most logical way possible; put the streets down, then add blocks for buildings. Then once the map is all even (takes same time for both teams to reach certian flags) I redid the buildings I wanted to be empty and such, but it got so broing.
Now I find it a lot funner to start at one end and slowly go through everything then add details later (Although now I have to redo all the ground as the displacement maps are all jumbled together since I didn't bother lineing them up to sew)


Hey Trav; You from NZ mate? I see a map of Auxs in you siggy there, you near there? I'm up north :p

Auckland born and raised, spent my first 18 years in aucks then moved to sydney, been here for 2 years~

patriotism ftw.

Usually I layout the roads and sidepaths, then legoblock the buildings. Once thats done I go back and start one area and complete it to my satisfaction detail wise and try to keep to that standard through the map.

I noticed no one even said a thing regarding Kid-A's compile times, so I'll just comment on that...

a day to compile = something wrong. I just can't believe it would take that long. Either you're running hammer on a 10 years old computer or you haven't done any optimization.

Either way, optimization should be done while making the map (in my book) so you're sure to get it right and don't get stuck as you are waiting forever for your map to compile. I'm already fed up when mine takes more than 2 hours, so I couldn't even think of waiting 24 hours....

Yes that's the same mistake I made / problem I'm having, I shall certainly not do it again.

As for planning I already have a vague idea of the general layout of my map as I based it on a town in France that I have been to a few times. I drew it out on a piece of A4 paper and started working from that.

I've been doing my buildings in separate files, then I have the main layout map with the "lego block" placement buildings. I'm planning on dropping buildings in there once they're completed separately.

Originally posted by Furyo
I noticed no one even said a thing regarding Kid-A's compile times, so I'll just comment on that...

a day to compile = something wrong. I just can't believe it would take that long. Either you're running hammer on a 10 years old computer or you haven't done any optimization.

Either way, optimization should be done while making the map (in my book) so you're sure to get it right and don't get stuck as you are waiting forever for your map to compile. I'm already fed up when mine takes more than 2 hours, so I couldn't even think of waiting 24 hours.... Thanks. I thought something might be wrong.

What are things to consider for optimization? I'm pretty sure my map is simple at the moment, I've got a few displacement maps and a tree model but that's it. Other than that simple brushwork...

I suppose I might have copy and pasted a brush over itself, could that be a problem?

edit: The problem seems to be PortalFlow, all the other processed finish in seconds...

OK I found the problem. I converted some complex brushwork to func_detail and it's done in 2 minutes...

Geeze a day! Have you got any brushwork errors? I'm guessing you know how to run "check map for problems" on hammer. If it reports such errors its usually best to fix them or the map won't compile, hang, or just take a long time to sort itself out. If not, it may be worth clipping your complex brushwork into sections as well so vbsp/vvis handles the planes differently.

Also, remember to apply the "nodraw" texture to the faces you can't see.

Its also best to not have a skybox around the whole map, or if you do at least put a hint across the highest point players can see in the map (or just below it if this is the only high point, it all depends)...Its usually good to put the skybox "between the gaps" if that makes any sense?

Make sure brushes don't overlap frequently. Clip them off if they do.

Never use the carve tool!! I know its easy, but it can cause so much hassle!

When making a brush func_detail, also remember that it doesn't block any vis...so if theres much behind it (ie. if its a wall) then don't make it func_detail as everything behind it will be rendered when it doesn't need to be, and can cause fps problems further down the line.

When it comes to creating a new map. I always draw it out on paper first and show a few of my friends that play the game, as well as my dad to see what they thing. I write in what each room is (eg. kitchen, bedroom, storage, empty, etc and the style of each house if its a town based map). I don't see the point in poncing around with scanning the image in and converting it to brushwork from a texture, etc as I find that a bit unecessary. I could just build it all visually from having the paper next to me.

The 'brushwork' was a market stall, I think it should actually be a model or something really. However it's in the middle of an open area which would probably mean lots of leafs (leaves?) would be created by it. numportals went from ~4500 to ~1500. I also made any wall with windows in func_details. Obviously this doesn't matter as anything you can see from the window is best rendered even if you are behind the wall with a window in it.

There are some market models in props_italian called:

anzio_market_table1.mdl
anzio_market_table2.mdl
mkt_table3.mdl

and also some nice baskets:

mkt_container1.mdl
mkt_container3.mdl

I have the basic idea,...but don't commit it to paper.
I build based on photographs of the area, or structures I want, but scaled to fit that same idea.
While I always end up with a useable / playable version, it's never exactly how i invisioned it, as once the idea begins to take on 3 dimensions, you quickly realize that something about it is impracticle, or improbable, or just plain doesn't or won't work in a game environment.

Re: optimization, do skybox brushes block VIS?