UPDATE NOVEMBER 21, 2008
The map is available in the downloads section HERE
or at FPSBanana HERE
Hello everyone! Since this is my first post on the forums I just wanted to thank everyone here first, because the community here is great, full of fantastic knowledge that has helped me build this map over the last year. So thanks!
Now for some background, this actually began about a year ago, a month after the Orange Box was released. TF2 quickly became my favorite game and due to there being so few maps released with the game, the creative wheels began to turn! Soon after I downloaded the Source SDK and began playing around with hammer and drawing out some sketches of ideas. The truth is I had NO idea how much work was involved in actually laying out the game area of a map was, so my first concepts were simple. My original idea was intriguing to me and, and oddly simple: a perfectly straight, VERY long map, capture the flag gamemode, with pillars spread evenly across the length. The idea of two teams charging head-on into one another seemed so chaotic and action-oriented I thought "That's gotta be gun! Hell, I want to play it!" After learning the basic skills and techniques to map-constuction I produced this:
Now it doesn't take a game designer to realize that TF2 maps are all built around a central theme, whether it be a Goldmine like pl_goldrush, or a granary, not surprisingly, like cp_granary. So when I considered my initial concept, and tried to think of a reason why my map was so long the idea hit me right away. TRAINS! What if I made it like a train warehouse, or garage, or whatever those things were called! I considered modifying it to be more of a train station, but the industrial construction/resource harvesting corporation theme didn't quite fit. Thus my map now had it's central theme: it would be some kind of train shelter/station in the middle of a massive trainyard! Now I don't know if any of you have seen one of these, but they are actually pretty cool. Here's a shot of one via google:
Moving along I now had a great theme to build around. The first thing I wanted to do was decide where the intelligence should be located. At first I had each one located in a separate room directly at the end of the respective sides. Simple, straight forward. Then realized I needed there to be more than one way to enter the room. How about a vent that you could drop down from? Great idea! But I needed a way to get up to a vent in the first place. How about a second level? Done!
But then after sitting on the idea for a few days, I got bored. I wanted more, and this silly "straight long map", which I had lovingly named "The Gauntlet" just simply, well, was too SIMPLE. Looking at all the Valve maps, I realized if I was going to build a map, it was going to be as close as I could get to their quality! So I scrapped the layout and began redrawing concepts. I still loved the trainyard idea, and kept the long length of the map. But it needed more...verticality. Rethinking the intel room placement it, didn't take me long to think of a great place; how cool would it be to have the intel room be on an upper level, overlooking the map!? I put it on what I considered to be the 3rd Floor, high above where the general action would be.
Of course, now I needed a way to get up there. When pondering different paths I could make, my attention strayed towards the center of the map for some reason, and I put the intel room on hold. I now how a fixation with placing a central room on the 2nd floor, where the two teams could meet in the middle for some action. I began with a square room, with 4 entrances, one on each side.
Along with this idea, as you will note in that last shot, I had another idea, probably my favorite looking back at the map. I had this vision of the entire 2nd level being a metal catwalk hanging from ceiling, made up of a series of narrow paths with railings. The image of ragdolls hanging and falling off the edge down to the main level, and being able to watch from that ground level seemed so cool! Now at this point in my mapmaking career I had just befriended a person on Steam, now good friend and my mentor at the time, Sick Spider. He noticed that I was in Source SDK all the time, and asked what I was up to. As it turned out he had a lot of experience in mapping, and offered to take a look at what I was doing. He was APPALLED! Not by the map of course, but by my brushwork and texture work. I will remind everyone this was my first time ever using a mapping program, and I was learning it all as I went. So he took me under his wing and showed me the ropes, introducing me to concepts like optimization, CRUCIAL to any mapper. One suggestion he made was to texture everything in NODRAW, a texture I was not aware of, that way I wouldn't render the things players wouldn't see. A brilliant concept! Now, with his help, I could really dig in to the construction. Going back to the catwalks, I soon produced this:
NOW I was starting to see my vision, and with my new-found skills and techniques courtesy of Sick Spider, I really started to work fast. What used to take days now turned into hours. Sweet. I started to get down some more detailed layouts.
At this point I had realized that a trainyard would be nothing without trains, thus it would be necessary to have trains speeding through the map a la cp_well. Thus the gates and traintracks seen above. I also began to move on to other areas. The sewers, a 4th level, came into play.
My initial concept, while a bit primitive and simple, stuck, and is generally the same now, besides the exit stairs.
I also began work on my spawn rooms, which became my first detailed room. I wanted the spawn rooms to feel and look like locker rooms, a look I don't think Valve captures very well in their maps. The first thing I think about when I think of a locker room, is the showers, so they became a centerpiece for me. Initially they were going to be stalls like so:
However I failed to find a TF2 prop for a showerhead, in fact the only prop that showed when I searched for shower, were these weird things that are supposed to be sanitizer spray things.
Don't ask me, I don't know. At any rate it just wasn't working, so I went with another, more risky idea. There was actually one other prop that came up in the search, which was perfect! Except it was a Half-Life 2 model...Well I took a risk, and figured (hoped) that anyone playing TF2 MUST have HL2 installed as well, and thus will not get a missing model error when playing the map. Here's to hoping!
Now at this point I had finished most of the layout and begun texturing certain things, getting excited at how my first map was turning out.
Everything was coming along. And then a bomb got dropped on me. Sick Spider, after looking over all the work I had done, took that time to inform me that my brushwork was incredibly messy, and had no cohesion in a lot of places. He also took that opportunity to inform me that APPARENTLY you are supposed to work in specific gridsizes, usually 8x8 or 16x16. I had been working at the smallest, 1x1. After walking me through my mistakes, we both concluded that it really would be best if I rebuilt the entire map from scratch. Disheartened but more motivated, I set to it, and amazed myself. It had taken me about 2 months to get to where I was at that time, including learning everything about Hammer. It took me 4 days to do it all over again. Now I was really pumped, knowing how much better and more efficient I had gotten. I reached a point where I started adding lighting, something I initially had trouble with. My first work with lighting was too dark, unrealistic.
I couldn't figure out how Valve and so many custom mappers were doing it. Then I found you guys, TF2Maps.net, and discovered a thread that had all of the Valve maps decompiled. AMAZING! Now, after browsing through their maps I could see exactly how they constructed, well, everything, including lighting.
I also began to have massive problems getting the cubemaps to work. Apparently only a couple of the Valve TF2 skyboxes work properly, producing correct cubemaps that fix any reflective issues (the purple/black checkerboard).
It took me months but I eventually got it fixed, settling on the drowsy sunset of Badlands (I initially wanted to use cp_well).
At this point it was the beginning of the summer of 08, and I was soon leaving home to go work for the summer at a kids camp on an island, far from my PC, and wouldn't be back for a few months. I was desperately trying to get a working beta out before I left, however it was not meant to be. The layout was complete, but I had not begun to build the trains yet, and I was having some major optimization issues. I had no idea how to use Hint brushes or Area Portals, so the map had none. In a more closed off map this may have no been as much of a problem (not that ANY map should be missing those key brushes), however my map had a SERIOUS design flaw that in retrospect I wish I had avoided. Because it was so long, there was a lot for the game to render when a player looks across the map from one end to the other. This caused some serious lag, rendering the map unplayable on most computers. After I returned from the summer job, I began struggling to find an answer to the problem. I tried Area Portals, but I was building them incorrectly and thus they did not work, causing more problems. I was out of luck, and my first calls for help online went unanswered. For the time being I decided to begin work on the trains. MAN are those things bitches to make. I plan on writing a tutorial to help anyone who wants to build them like Valve, because it really was a long and torturous process that I had to teach myself, just by looking at how they built them using the decompiled cp_well. Initially I wanted to had the trains be 3-5 cars long with flatbeds so people could board them. I had come up with this brilliant idea of making a vent that led from near the intel room to the main (ground) level.
The kicker was I wanted the exit point of the vent to drop right down on the train tracks, thus if a player timed it right, they could board the train for a quick getaway!
I finally got the trains working, except when I tried adding extra cars there was far too many output/input mixing, and to keep things short, it didn't work. I will continue to work on that for a later update, but for now, one train is enough. And the escape route is great!
So the train was built, and I finally fixed things up enough that I compiled a playable version, albeit a laggy one.
Initially, there was only one way to get up from the ground level to the 2nd level (catwalks). On each side (remember, it is a CTF map, so each team has a symmetrical base) near the center there was a staircase on the right that went up to the center room I mentioned earlier.
My rationale for only having this one route (technically two, but both staircases lead to the middle room. Was that because the map was so long and had 3 levels (not including the intel room), I worried that it would be too easy to get the steal the intel and get sneak back home. This would funnel all of the action to that main center room.
After playtesting, I received a number of complaints about this and realized a second route would be necessary. I decided to add another set of stairs leading up to the catwalks, close to the spawn rooms. To compensate for having a route so close to the home base, I placed the stairs on a raise platform, and in order to reach it you must jump on top of a number of boxes. This slows the player down and gives the enemy a chance to recover.
Another balance issue I had earlier in development was a lack of a sniper roost. Snipers can be difficult to balance for, because they need a certain amount of cover and enough distance to warrant their long range. And though I had actually initially designed the map with snipers in mind, having it be so long, as production moved ahead I had not given them their own designated place to be most efficient. A playtester suggested adding two roosts outside the intel room window on either side, and it ended up fitting perfectly.
Now last month I finally got into contact with a wonderful guy from FPSBanana named Lost, I'm sure some of you know him. One of his specialties is optimization, and he agreed to take a look at my work and see if we could get the lag problem figured out. He was ever so courteous enough to fix some of my brushwork to fix a few leaks we were getting from Area Portals, and after he added a few of them and some Hint brushes I was finally able to finish up the optimization and after a lot of trial and error, was able to, for the most part, fix the lag, no longer receiving a memory error. Part of the problem though, was that on top of that central room, I had this big open platform players could reach by either rocket/sticky jumping onto, or jumping on to of a few barrels and crates from inside the room.
Now in theory this was great, because it gave players the high ground and snipers another area to utilize, and it was cool because you could see the whole map. However that last point was a major problem, because now that the whole map was visible it was being rendered. I concluded that I needed to close it off, but I really liked the extra level, so just altered it a bit, as seen below.
So now I was able to optimize the map even more! So here I am, near the end. The only things left now were some detailing with overlays and decals, as well as a 3D Skybox. The overlays were easy, and I think I did a decent job using them to make the map more immersive. In later updates I'll refine them even more. The Skybox was difficult though. After I figured out how to build one, I played around with a few ideas. I had planned to have most of the outdoors be a rolling plain with farms and hay bails littering the fields, but I had trouble getting the distance and look right.
So far now, I've settled on an urban look, where buildings fill the players view. It's not the most spectacular skybox, but it does its job, and I will pretty it up for later updates.
And so now I've reached the end of production, and I'm VERY close to releasing my first Beta of my long-ago-dubbed map, ctf_fu_trainyard. To explain briefly where the FU comes from, I would have used ctf_trainyard but the name is already being used by another map. I also had been making the map mainly for the server I frequent, The Maximum PC/PC Gamer server found at 188.8.131.52:27015. I have been playing TF2 competitively for a team called Team Future US, founded and organized by various editors from the two magazines, MaxPC and PCG, thus I dedicate this map to that team, and the company that helped kindle my love for PC gaming. I good point to make now, because it happens to involve the last issue I am having before I release the Beta. On each Intel Room desk I have included a magazine for each team, a PC Gamer cover on the Red Team's,
and a MaxPC cover on the Blue Team's.
I cannot seem to properly embed the custom textures into the BSP using Pakrat however, so I'm currently at a stand still. [THIS IS NOW FIXED]
But now, after a year of work and learning, I now present to everyone my first map which I am deeply proud of. I've made plenty of designing errors, but I've learned from all of them and have no regrets! So please, if you've enjoyed what you've seen here and are looking forward to trying out this symmetrical Capture the Flag map (when you're not playing hours of Left 4 Dead like me heh), please keep an eye on this thread and leave me some feedback, positive and negative. I'd love to hear everyone's criticism, and look forward to adjusting the map to make it fun and playable for everyone! So cheers, and again, keep an eye on the thread. Thanks everyone!
For some more screenshots, old and new, check out these albums. Beta Production Old
I don't mean to sound like an asshole, but this theme is done perfectly by valve.... twice! CP and CTF versions of well. I will never understand why you would want to compete with that. And not even those are very popular ... so good luck!
The catwalks look great, though i am concerned about optimisation issues due to its linear nature. As well as the use of HL content, which may be an issue for those who don't have pre-orange box games.
I'm not sure what else to add, i'm kinda hung over right now, ha ha. Textures seems nicely alligned and appropriately chosen.
The development shots and annotations were nice but it seemed a bit much considering you're probably only after feedback?
3d skybox is adequite. But i think it needs more considering it's gonna be a large thing in the map.. with players flanked by it constantly. Though it is one of the hardest things to get looking right/nice.
The wood on brick doesn't make much sense, though i understand you're trying to stick with the themes. mixing the wood and brick textures is hard to pull off, though it can be done; it doesn't look right here. You're probably better off not utilizing the wooden textures unless you can really pull it off with brush work in later releases.
A Boojum Snark: Don't listen to grazr he is probably drunk