Xatrix level designers (Q2)

6 December 1999 on www.gamedesign.net

Dan = Dan Koppel aka Daniel Koppel (Technical Lead and Senior Level Design)

Alex = Alex Mayberry (Level Design)

Mal = Mal Blackwell (Senior Level Design)

Q = ?, Uploaded by Oirfid

Q: Tell us a little bit about yourselves and Xatrix? How did you guys get into the level designing business?

Dan:I've always been into computer games and before that it was strategy/wargames. Until the arrival of (classic) quake I was a graduate student studying biophysics studying 3d modeling of proteins who just toyed with modifying levels for descent and doom2. After quake, I became completely absorbed with making levels. For quake, I made 4 user levels. A couple of my levels were noticed by the people here at Xatrix and I left school to join the great team here at Xatrix. I joined Xatrix in September of 97. Also, I'm married and love baseball (big Cleveland Indians fan).

Alex:I had been working in law enforcement for almost five years when I realized that I'd made a bad career choice. Taking a risk, I decided to quit my job and go back to school. While my wife worked two jobs, I studied computer science at our local community college. Everything was going great...I was getting good grades, I was completing the courses that I needed for my degree, and finally I thought I knew what I wanted to do in life. Then this little game called DOOM was released. I found out about level designing, and I started to dabble a bit in my spare time. It wasn't long before I was neglecting my classes and spending all day building levels. To make a long story short, Xatrix saved my life by hiring me as a level designer; I'm sure if things had gone on much longer they'd have found me murdered in front of my computer with a half finished DOOM level up on the screen. Lucky for me that my wife has a lot of patience :)

Mal: Well, I started out as a Graphic Designer for textiles. I had owned my own business for about 2 years before getting a computer (to do work with of course). At the time Wolfenstien 3D was a pretty popular game. As you can guess I had a copy of Wolf, and thought it was the best thing since sliced bread. Being new to computers I also had that swell America Online Subscription, through which I heard about the "Next Great Thing" the soon to be released "Doom!" The night the shareware was released on the net, I logged on, started the down load and went home. Without knowing it would be the last full night of sleep I'd ever have again. Well, the next morning I got in, unzipped the file and started up Doom. Wow! That was it, I was hooked. I ordered the full version the same day and played till Doom was burned into my retinas. Soon a neat little program surfaced on AOL, Deu. A Doom map editor, a new reason to stay at the office till' 4am every night. Doom2 was soon released as well as Due2 and finally Deth. I spent about 12 hours a day building levels and 2 or 3 doing work for real money (suffice it to say my business went under a year later). At about this time I happened to start a correspondence with another level builder on AOL, Dr. Sleep. We would exchange tips for building and critique each other's work. John, being more involved with the Doom community than I, soon found himself on a number of work for hire projects like Blood and Unreal. Thanks To John Anderson (of Ion Storm) I too was contacted by Q Studios (Blood) and Epic MegaGames (Unreal) and was able to work for a time on both of these projects. Unfortunately working from home part time was starting to strain my day job performance, and I was unable to stay on with either of these projects. These experiences though, solidified my desire to find a full time position in the gaming industry. Just before getting onboard with Epic John and I were to start a Doom2 TC known as Eternal Doom. Though John and I never really did anything on Eternal, John stayed in contact with several Eternal designers. At any rate, Alex Mayberry (of the Eternal Team), our Lead level designer at Xatrix happened to be looking for a Designer in the Los Angeles area and John was happy to drop my name (again) Thanks John.

Q: Did all of you work on Redneck Rampage before the mission pack? If so, how does level design for Quake2 compare in terms of difficulty and time?

Dan: Alex and Mal did, the q2 mission pack is my first professional project.

Alex:Both engines have unique problems and challenges to deal with. Build levels tended to be more what I'd call "handcrafted," as literally everything had to be done by the designer in order to achieve the greatest degree of realism. Quake2 on the other hand has automated many things for us, and obviously given us the awesome power of true 3D. This of course does not come without a price, as the engine forces us to deal with r_speed issues, long compile times and other limitations. On average, Quake2 levels take longer to build, but the end result is naturally worth the extra effort.

Mal: I'd have to say that both the Quake2 and Build engines have their own difficulties to deal with. In Redneck I had to work a lot harder to put together really interesting environments without the help of a true 3D engine. With Quake2 I'd say the lighting is the hardest thing to master. Lighting can make or break a level. There's also the r_speed limitations to deal with. The one thing I miss about Build is the instant gratification of building a level and going into the game to see it without waiting for a Bsp to finish.

Q: From Redneck Rampage (2.5D Build Engine Game) to Q2, were there any stumbling blocks for you guys, or was it an easy transition?

Dan: Most of my difficulties have come from my relative lack of experience. Making a unit (5 levels) that has a single theme is not as easy as it sounds. After about 3/4 of a level I'm ready to move on to something different. But for a unit you want the player to feel he/she is in a real place and that things go together.

Alex:For me it wasn't an easy transition. I'd never worked with true 3D before, and at first I found it very alien to me. I had to forget everything I knew about constructing levels, and learn to think more 3 dimensionally. Fortunately, Mal had had a lot of experience with 3D, and he was able to help me over the rough spots.

Mal: For me Quake2 was easier than Rampage. I'd been working with 3D as a hobby for a couple of years, TrueSpace and Lightwave. I liked to model so 3D was more or less second nature for me.

Q: In your opinion, what is the most difficult thing with designing levels for a 3d game engine like Quake2?

Dan: Finding time to sleep. ;) Staying within the limits of the engine. I am constantly running into situations were I want to do something sooo cool but either the entities don't support it or the limits of the engine just make it impossible. As level designers we are always looking at r_speeds, if we don't get the feel or functionality we want under our r_speed limit we need to rip it out and try something else.

Alex:The most difficult thing about designing levels has nothing to do with the engine itself. It's about taking the time and putting in the effort to do things right. It's about looking at a room and knowing that there's that one extra thing that would make it really cool, and then taking the extra time to build it. It's about spending hours getting the lighting just right and aligning the textures. It's about placing the enemies and then testing and testing and testing until you HATE the freakin' thing. And then finally, it's about not being married to any of the work that you've done, and if needed, being able to trash it all in favor of a better idea or needed change.

Mal: My greatest difficulty is my obsession with detail and perfection. I can build and delete a room 6 or 7 times before I feel I am happy with just the base geometry. Then there's the detail. Balancing what you want to build with in the limits of the engine and the "minimum system requirements" has always been a struggle for me. Sometimes I have to tell myself that less really is more<not>.

Q: How would you describe your methods of learning the ins and outs of the entities themselves and how they work? Did each of you work out the details for yourselves? Did you get any help from id?

Dan: Until, the quake2 was released id's entities changed quite a bit and many weren't fully functional until the game was done. (Some never became functional, target_actor) Many of the rest were similar to (classic) quake so I knew most of the ins and outs. Some of the rest we worked out ourselves and others we did get help from id (area portals and hint brushes).

Alex: Once you get the hang of how things work in the Quake2 engine, most of the entities are fairly easy to figure out. If something becomes tricky, there's always the old trial and error method, or we could simply look at the original Quake2 maps as a guide. Of course, the guys at id have been a big help in answering any questions that we've had along the way, so learning the ins and outs of the engine for the most part has been fairly painless.

Mal: We did alot of experimentation at the beginning with some of the newer entities and id was always willing to explain the stuff we had problems with. Also, if I still don't get it I just ask Dan.

Q: Have you learned any neat new tricks within the existing Quake2 entities? (Do tell.)

Dan: One trick I've played with is using pathtargets with a trigger_elevator, by doing this you can get an elevator to go to a number of path_corners. The path_corner has a pathtarget which goes to a trigger_relay wich sends the elevator to another path_target. After it stops (or you kill the trigger_relays) you can hit another button which sends the elevator to a different set of path_corners.

Alex: You bet! There's this one thinnng thatnlkldld damn keyboard is stickkkkin......gjdjkdk damn.....dj ....oh well, guess you'll just have to wait and see...

Mal: Well, I have had to re-think the whole conveyor belt problem. Having funk_trains actually "be" the conveyor was not working well with the insanes I was dropping onto it. The insanes were getting stuck going around corners and the belt was slowly going out of sync with itself. So I devised a much simpler funk_train using trains with the clip texture on them to push the insanes from behind rather than try to conveyor them. It's also great because it doesn't effect my E or W polys.

Q: Are you planning on adding any new entities or major changes to others?

Dan: We have plenty of new monsters, weapons, keys, power-ups, etc. but not really any new geometry based entities. We already have some in mind for our next game but that's for another time.

Alex:We've definitely added some new things to enhance the game experience, but nothing that will drastically change anything that was done in Quake2. Rather, most of the changes we've made will fall right into place in the Quake2 universe, and should fit in seamlessly with everything that id has done.

Mal: We have a few new toys.

Q: How does the mission pack compare to quake2 in terms of gameplay? Are you guys going for a more interactive story line or are you keeping the same basic level hub system of Quake2

Dan: Id did such a great job with quake2 that it's hard to improve on and were making a mission pack so with that in mind the game play is quite similar to quake2. You are actually another space marine invading stroggos but we do have a couple of great twists.

Alex: The game takes place at the same time as the events of Quake2. You play a different marine that touches down in another region of the planet, your mission to locate and destroy a secret enemy moonbase. Gameplay for the most part will follow that of the original levels.

Mal: The Mission pack takes place at the simultaneously with Q2 and we've adhered to the same level structure as Q2.

Q: Do you plan out a level on paper before the actual designing begins or do you just have a mental image of it and go from there?

Dan: Yes ... This one varies from level to level and depends on what I'm trying to do. I usually sketch out my level to get the 3d placement and flow down but the look and feel are usually based off of mental images.

Alex: I used to plan out levels on graph paper when I first started designing. As I became more experienced though, and developed certain techniques and themes, I found myself doing this less and less. These days, I generally build free-form, and let the level take form along the way. Once in a while I'll sketch something out on paper just to get a rough idea of placement, but otherwise I just try to get a level started (usually from some sort of image I have in my head) and then let it flow from there.

Mal: I will usually build free form, with just a mental image of what it is I want to build. I will resort to paper and pencil if I'm dealing with really specific routs or complex geometry.

Q: What do you use for inspiration when building a level? Books, Buildings, Movies etc?

Dan: Yes, Yes, Yes. When you crank through a level in 2-4 weeks and then start another one etc. anything that can give you an inspiration is a godsend.

 Alex: Once in a while I'll see something that will give me inspiration, but for me level ideas generally just spring to mind. If I'm stuck, I'll start building brushes and moving them around, changing heights here and there, or clipping off an edge, and soon enough something cool will start to develop. Getting started is always the hardest part...but once you get that first room going, then the rest falls into place.

Mal: All of the above and then some. Comic books are also a great source for inspiration. Building levels all the time can sometimes strain your creativity levels and any kind of outside stimuli helps.

Q: In the screenshots released, I have noticed that there seem to be a lot of outdoor areas. What techniques do you use to create terrain and large, open, outdoor areas?

Dan: Since Mal and Alex did most of the preliminary work in this area, I'll let them answer this one ;)

Alex: We've found a fairly efficient method to creating organic environments, minimizing the amount of brushes used and keeping the r_speeds down. As for a detailed description of this, I'll leave that question for Mal to answer, as he was the one to first hit upon this method.

Mal: This may be a little tough without being able to illustrate some of the finer points and all my experience has been with Qe4 or QeRadiant but here goes. I got the basic idea for the canyons from a small set of rocks in Base1. What I did was build a wall of parallel brushes all 16 pixels thick and all of the same height, but I would vary the lengths randomly. After building each brush I would go into edge mode and pull one side of the brush out to create a jaggy sort of wall. The trick here is to keep the edges of the brushes parallel to each other. After that I would go into edge mode again and pull the top leading edge of each brush out say 32 pixels (you can vary this however you like but you need to make sure you do the same for each brush). After you have the lower wall done you select the entire wall, copy it and Z-flip it. Put it on top of the lower wall and do it again till desired height is reached. Perhaps when this pack is shipped Shane will allow me to do a proper tutorial.

Q: How do you guys feel about maps changing hands within a company? (Everybody working on every map, even if only for tweaking, lighting etc.)

Dan: For tweaking and touch-ups it's fine but for a partially finished level I don't think it's a good idea. When you design a level you have an idea of what you want to do with it an where you want to take the player this can get watered down and lost if the levels are past around. My feeling is if it's not working it's probably better to trash it and start over. (This is probably easier than completely redoing a level - which I have done, ugh.)

Alex: We did this quite a bit for Redneck Rampage, but it's really not needed as much with the Quake2 engine. Quake2 is so advanced that many of the detailing elements that would have been done by other people for us is now done by the engine itself. Not to say that maps don't ever change hands; indeed, sometimes when a deadline is near, one designer might help out another by doing some touch up work, or place enemies and sound, etc. Generally though, maps tend to stay with their respective owners.

Mal: Actually, it's very rare that someone other than the level designer building the map ever gets to work on it.

Q: Do you ever feel that a map is 'finished'?

Dan: When I first got here the answer was no. Now, the answer is well sorta. There is always another map I can do such and such on. So when I'm finished, I'm done and it's time to move on to the next map.

Alex: You can ALWAYS find something in a map that you'll want to tweak. If you're building levels as a hobby, then nothing ever is truly finished (nor does it need to be). When making a commercial product though, there does come a point when you have to walk away from a level and let it stand on its own.

Mal: This is a trick question, right? Sure, there comes a point when I feel my maps are done and ready to be shipped. Although there will always be something that I could of done to make it better. Little tweaks here little tweaks there, ya' know.

Q: Have you had much of a chance to look at any of the user-created levels?

 Dan: I've tried to play some of the quake2 user maps and have seen a number of good ones. I haven't had a chance for about a month now so if there are some great newer ones I probably haven't seen them.

Alex: Just a few, and surely not as many as I'd like to. I imagine when the Mission Pack is done, we'll start taking a look at more of what's out there.

Mal: Yes, I've seen quite alot actually. We've always got our eye out.

Q: Any advice to all us amateur level designers out there?

Dan: Try to keep the players on their toes. If a level gets me to jump out of my seat, I'm sold. I can overlook a misaligned texture or a loss of texture theme and even higher r_speeds (still not over 1000), but a boring level is a sin, even if it looks good. Also a very important trait for a level designer is attention to detail. The most anal people here at Xatrix are all level designers. ;)

Alex: First off, do things right. If you build an area, make sure to finish it. Don't leave the polishing work for the end, because by that time you'll be tired of the level and you'll get sloppy. Be a perfectionist, and polish as you go. Next, use lighting effectively. Don't over use colored lights, and learn to create areas with a good diversity of light and shadow (always making sure that your lighting is logical of course). Finally, build vertically. It's a 3D engine, so use it to its full potential. Try to make levels that double back and revisit areas from a different height. This will maximize the amount of space used in the level, as well as make the environments more interesting and enjoyable for the player.

Mal: I would have to agree with what Alex already wrote, take your time and do it right.

Q: If there is one thing that you think is most important for a level designer to remember what would it be?

Dan: Game flow. Often overlooked but extremely important.

Alex: I'll say it again, do things right. Don't cheese out and leave something incomplete. If a texture isn't aligned, then align it. If the lighting isn't quite right, then spend the time it takes to make it so. No one really notices when you do things right (after all, who walks around a level and says, "Oooh, look, he aligned that texture!"), but they sure as hell notice when you do them WRONG.

Mal: If there is one thing that is most important for a level designer to remember it's the end result, the quality of your map when you feel your finished. (see question 15.)

Q: Doom had the rabbits, Heretic had the chickens, Hexen 2 had the sheep...what farm animals can we expect with 'The Reckoning'? Remember…the sheep are taken.

Dan: You forgot that redneck had the pig ;)

Alex: You're kidding, right? Remember, we're the guys who did Redneck Rampage. We've butchered more virtual livestock than an Arkansas slaughter house. We gotta lay low for a while before we piss off the wrong animal rights group....

Mal: We had chickens, pigs, dogs and cows in Redneck Rampage, I think were farm animaled out!

Q: Thanks for doing the interview. Any kind of closing comments you guys would like to make?

Dan: The quake community rocks. I felt that way when I first got into level design and got great feedback on my quake levels. Now working on a quake2 mission pack I get to see all the excitement from the other end. Late

---