Map Makers - What's your approach?

[nymets1104]

To the Veteran and even newer map makers, I have recently installed the SDK and I am finally taking the plunge into map design. I have gone over the basic tutorials and read through danh's helpful guide. I am curious as to how others approach the level design.

Mainly:
What do you design / create first? Terrain? Spawns?
Do you use the native terrain editor or import from third party program?
Do you make changes to the elevation before laying out a material for the terrain?
Is it best to copy/paste from stock maps or pull each item from the content browser as you go?
Do you layout static meshes before pylons/lighting or later?
Does any of this really matter or is there only one good approach to making the level?


Seeing how the editor gives you a blank slate to start, I can really appreciate that we have so many well created and well playing maps. Any input or advice would be VERY much appreciated.

 

[Moskeeto]

I'm not a map-maker, but I do believe it's best to start with basic terrain. Or, if you want to go extra detailed with the terrain, you could use Terragen to create something nice, export it as a height-map, then import into the SDK.

 

[LATTEH]

for this game research about the battle and try and copy some of the basic terrain of the battle (if you can with aerial photos.) but for other games i usually do this:
http://insurgency.gamebanana.com/threads/195243

This was for my map Kandagal which you can see from beginning to end there.

 

[Twrecks]

Check out these sites:

http://www.worldofleveldesign.com

http://www.hourences.com

Both have free articles and also books for sale you might find useful.

I start with a quadrille (graph) paper and sketch it out several times, using any reference mat'l handy, like Reallife(tm) maps. Gives an idea about spacing objectives and spawn points. Most maps have about a 150m approach to caps (50uu = 1m, so 150m is about 7500uu, or around 7-1/2 grids set to 1024).


First thing that gets plunkered down is the terrain. It's going to decide "how big" and "how detailed" your map will be. Smaller grid (50 or 64uu) is good for maps because it lends itself to the stock trench meshes. However, a large map with that gid size will suck the FPS down and make the file size really large when light mapped.


Adding a dominant directional light and some sort of sky dome sets the mood. Adding a spawn volume and hooking it up in Kismet will allow you to start running around. I start with a commander on each side in the world info, and build from there.


I've made around 6 maps this way, none were released/finished. Yakovlevo probably had 4-5 false starts, before balancing size:FPS.


remember:


The 4 Fs of Game Design:
Fail Faster, and Follow the Fun


http://www.darklorde.com/the-4-fs-of-game-design

 

[Sgt.NightFire]

What do you design / create first? Terrain? Spawns? I like to start with the terrain. Because the best maps are made as how the real battlefield was if you make it after real location or a location from a war movie, i found those to b the most fun maps to play. Darkest Hour had allot of these kind of maps and all were great.
Do you use the native terrain editor or import from third party program? Native
Do you make changes to the elevation before laying out a material for the terrain? Got no clue what that means, too tired.
Is it best to copy/paste from stock maps or pull each item from the content browser as you go? Depends on what you want, i rather just load in a map with a static mesh i want and put it in the editor fresh.
Do you layout static meshes before pylons/lighting or later? Always before.
Does any of this really matter or is there only one good approach to making the level? Nope!

 

[Raven1986]

I also do sketches on paper, currently I am re-working StalingradKessel, so it's not a 100% own work. Yet I will add a large amount of new stuff to it, so it will look quite different than the stock map.

I have tried out doing terrain a bit but as Wrecks told me, my 8GB RAM are not nearly enough. Meanwhile I think I will go for 32GB to wreck havoc on that vile SDK

 

[Catalavos]

I'll echo most of what's been stated here so far and add a few things:

-I like to add some basic stock meshes to the map to get a sense of scale. It's hard to visualize how large or small your map is until you add some elements of known size.

-All of my map ideas have been inspired by real events and real places but early on I realized that staying true to life can present problems for gameplay. Be flexible with your ideas and leave some wiggle-room for changes that will be necessary as you build your map. These changes may not show themselves until your map is in a beta-complete stage and people have had the chance to play it a few times.

-Stay on the grid!!! This becomes especially important if you want to have buildings with basements. Staying on the grid will make it easier to hide unwanted or unneeded terrain blocks. Also, using the grid snaps makes it very easy to place repetitious items like fences or windows.

-Worry about things like lighting and pylons well after you've laid out your map. All pylons really do is tell bots where to go (although you do need to place one over each PlayerStartVolume). Lighting will require several attempts to get the look you want. I usually just copy/paste the DominantDirectionalLight from a stock map to get started then tweak from there.

-Don't forget to disable all the CastShadows settings on your sky dome. It'll give you strange lighting and you'll drive yourself nuts trying to figure out why. This is usually the culprit.

-Brush-up on your photoshop skills in your spare time. You'll need them once it's time to make your overhead map.

-Don't get discouraged when something goes wrong. The SDK is a fickle b***c and she will poop on you; a lot. In many cases, however, there will be an easy fix to your problems so ask lots of questions.

-Add some mappers on Steam. You'll have a deep pool of experience to draw from at almost a moments notice.

-Have fun!

 

[nymets1104]

Thank You all for your advice and insight. I have taken something from every post here and have already begun to apply it to my project. Twrecks, loved the links and have been watching several Tutorial videos on the WorldOfLevelDesign site. I will not hesitate to ask when I hit a bump in the road. Now excuse me while I return to the rolling hills of North West France....

 

[nymets1104]

How important are "enterable" buildings? I have created several Solid BSP Buildings (Think Spatanovka C & D Caps) but wanted to make a couple that could be entered possibly with multiple floors. Do any of you think making enterable buildings is important or is it a waste of time?

 

[Moskeeto]

I personally love having buildings that can be entered.

 

[i8pptu]

I think enterable buildings are always interesting but there is a lot that goes along with them.
The first thing that stands out in my mind is consistency. Can I make my building meet the same standard of quality as other non-enterable/pre made enterable buildings.
The second thing is what is the point of the enterable building. I believe that everything in a map should (more or less) have a purpose. If you have a building that is enterable but it serves no purpose, to me, it would seem like a waste of time.
There are other things to consider like optimization and whether or not you should be able to mantle windows(i.e. cover node related things) etc. This all adds up to a lot of time spent creating something.


One of the important things for me when creating a map's design is to think about purpose. I might want to put something cool in my map, but if I ask myself why is it there? does it make sense? And I can not answer those questions, then I wont spend time on those things.


I guess I try to focus on the things that I want in the map for sure. Like the way the terrain generally is laid out and specific building/structures. Other than that, listening to the community feedback can help solidify what you think the map should be/needs.

 

[nymets1104]

Fewer larger capzones or more smaller cap zones?

I have gone back and forth on this a few times and for the moment, all I have to base this on are bots. The large caps fall very quickly when playing with bots, but I am thinking that a map feels more balanced when there are more smaller caps? any opinions?

 

[Catalavos]

Fewer larger capzones or more smaller cap zones?

I have gone back and forth on this a few times and for the moment, all I have to base this on are bots. The large caps fall very quickly when playing with bots, but I am thinking that a map feels more balanced when there are more smaller caps? any opinions?

I think it kind of depends on the scenario and the object or location being captured. All the maps I've played have some combination of both large and small cap zones.
If it's a specific building, like the Spartanovka church for instance, I think that the cap zone should only include the interior. But if it's a large building you might want to break it up into floors or maybe the north/south sides. If you're attacking something larger and less defined, like a farm, you can get away with several cap zones. Make 1 for the field or pasture, 1 for the barn, 1 for the chicken house, etc... What ever you choose, make sure that capturing the objective also carries with it some other, greater purpose. So that the "Chicken House" is also the building that houses the radio.

For combined arms maps I think it's kind of cool to have objectives that can only be taken by infantry but if you do this you'll need to make sure to put them closer to the beginning of the map. I had an issue on Little Saturn where the attackers were down to 1 tank and the defenders were down to a few riflemen. The final objective was a single building and the cap zone included only the inside of the building. The tank couldn't enter the last objective and the riflemen couldn't blow up the tank. Perfect stalemate.

 

[StmRtn]

-Don't get discouraged when something goes wrong. The SDK is a fickle b***c and she will poop on you; a lot.

Agreed, 101 %