The PlayStation 3 and Xbox 360 action title Dark Sector is the highest profile 'AAA'-aiming game thus far funded by Puzzle Quest publisher D3Publisher, and an absolutely key project of Canadian studio Digital Extremes, a company once best known for working with Epic Games on the Unreal series.

In fact, Digital Extremes, formed in 1993, has a long history of titles from Epic Pinball through aid on Unreal Tournament 2004 to Pariah and Warpath. But the long-in development Dark Sector, first unveiled back in 2004, might be described as a 'make or break' title for the company.

To find out how development is going on this ambitious next-generation, multiplatform project, and to learn about the challenges of developing a game on an original engine with intriguing new play mechanics, Gamasutra sat down with the game's project lead, Steve Sinclair.

How have you been able to spend so much time in development?

Steve Sinclair: We've been with D3 for a couple of years, and before that, Dark Sector's been something that's been on the backburner at various times in development. It's been two years in full production, and prior to that, we worked on Unreal Championship, and some of the Tournament games. Some of it was funded on our own dime, and some of it was with D3.

More of what I was getting at is like... how is it working with D3? They seem to be really patient, in terms of wanting to make sure that it's the best game it can be.

SS: Absolutely. They have given us more time, because the scope of the game is pretty large, and with the inclusion of multiplayer and that sort of thing, we've been able to put some things back in the game that we were sort of kicking ourselves for having to cut. Like the multiplayer game modes and that sort of thing.

Part of it is, "Do we have a choice of being a little bit later?" because of PlayStation 3 and 360... you know, they're not as similar as you'd like, and not as similar as we had initially thought when we were pitching the game. In some ways, we had no choice but to take extra time.

So you don't want to release one and then the other?

SS: No. They really want to do it simultaneous, and so do we, actually.

Yeah, it makes sense.

SS: I guess it makes sense with us, I mean, as far as we're told, anyway, it sort of makes sense in terms of you spending your marketing all in one go, and if you've got both out there, it's much easier.

Plus you also don't have people being like, "Well, I'll just wait for the other version," and then they don't, because something else cool comes out.

SS: And it drops off the radar. Absolutely. I guess that goes back into you spending your marketing at one time.

What phase of development are you in right now? Are you to polish yet?

SS: Yeah. We're in polishing the difficulty, the AI, the animation, and just playing the hell out of multiplayer and making sure it's cool and awesome.

_{D3 and Digital Extremes' Dark Sector}

I was talking to Harvey Smith a little while ago, and he was saying that...

SS: Harvey Smith. Yes, BlackSite.

Yes. Well, and other things, you could say.

SS: Deus Ex and so on. Probably one of my favorite game designers. When I started the game engine, I actually based it off of one of his GDC papers -- 2001, or '99 -- of the data-driven game development, so the whole engine we have is sort of based on his ideas of... the example he used in Deus Ex, where they had one door whose health was too low if you did the Men in Black explosion. You could break the level flow, so they had to go through all the levels and raise the health of all of them, which was a horrendous effort.

So he said, "What you need to do is have these driven from templates, and data-driven design fixes one place and it goes everywhere." That's been a big foundation of our game and our mission. And there were thoughtful unit designs that he talked about. Anyway, sorry. You were talking to him, and...

I was talking to Harvey, and he was saying that the last few months of the game development process are where the game just starts looking better and better every day, and that is the time when you really need to be able to focus on it. It's a time when you can actually add flair and tweak things and make the experience more...

SS: I think the last few months is when you determine what your review score is going to be, I think, because it's the make-or-break, where you have this mess of loose ends. That's what it's felt like, to me, in the last month. The game has turned a corner from being like a panic attack to being like something that makes me smile and giggle while watching the QA guys playing the boss fights and thinking to myself, "Shit, I didn't think we'd be able to do this good!" So, definitely the last week has been crucial.

To make a film analogy, not that those are necessary...

SS: Not like we don't make them enough in this industry!

It's stupid, because we've got a silly complex about film being better than us. But anyway, it's like taking the rough cuts and putting them into the final edit. It's a real big...

SS: Yeah. Absolutely. There are good analogies there, in terms of just editing down things, cutting out levels that don't work, cutting cutscenes or chopping them in half when they're brutally long, and things like that.

You guys seem to have a one-project-at-a-time sort of focus. What is the advantage of that, versus more small projects? Or at least, what is the reason you want to do it that way?

SS: I think that we are always trying to get to that two projects, staggered development thing. The thing that always seems to happen is, creatively you overreach, and you end up having to pull people from the mystical second project and get an all hands on deck sort of thing.

I suppose that's... you know, better planning, better scope control -- these are things that can help fight that, but in my experience, anytime we've done that stretch, we've contracted just to get it done and get it out at that sort of reasonable quality level. We're always trying for that, but it's hard.

It seems like it might be better to do a smaller, XBLA-type title as a "stepping stone" into two projects.

SS: Sure, like a smaller scope kind of thing. I agree. I think with this game, we were kind of idiotic, and we bit off three major factors and we should have only taken one, which is to change the genre a bit. Like, the third-person genre I horrendously underestimated in terms of how that impacts game design and level design. Sometimes it was like grasping in the dark and being astounded at, you know, weapon-switch times, and these things that you can't mess with for third-person, that you can just laugh and say, "Okay, like in Call of Duty 4. Switching between weapons is faster than reloading them." Try that in a third-person game and it looks like a Marx Brothers film, you know?

And then also starting the engine from scratch, all these things were major, major hurdles. We should've just done one. We should've picked one thing to try to do. I think that's also really limited our ability to grow. Although we've grown in staff, we're still working on this project, and this project only.

I'm hoping the next one is going to be easier, because we're going to have this base to build from, and hopefully it'll be a success for Dark Sector and the ability to make the next one really polished and refined and "the one we always dreamed of," and so on, with more established technology, and a more established genre.

Are there any similarities between the tech that you guys have done and the Unreal tech, since you guys were pals in the past?

SS: I think maybe in some ways, it might have influenced the programmers, that experience, [as well as] maybe some of our respect for the demands of console development and the demands specifically of resource tracking memory leaks and that sort of thing.

These are the things that we paid very close attention to in initial design because of our experience with putting Unreal on the Xbox. But I guess that's where the similarities end. There is a sort of sameness in the next-gen techniques that people are using, like normal mapping, and bloom, and the depth of field and the high dynamic range lighting. There is a little bit of convergence that happens there, but I don't know.

Yeah, that's kind of across the board. Have you played Unreal Tournament III?

SS: I have.

What do you think?

SS: It's probably got the most gorgeous levels I've ever seen. Absolutely.

I was wondering, since both Gears of War and Unreal Tournament III are out, is there an extra push to one-up them? I mean, just mentally?

SS: I hope that one day, we might be able to one-up them. I think us humble Canadians are content, yet driven, to just compete. Those guys have got it together, and they have an established base with which they grow from. They are definitely one of the competitors that we want the developers to be.

But I have the same kind of rage and lust for Naughty Dog titles and Valve titles, and the same kind of hero worship for Chris Butcher at Bungie and all sorts of things. But we look at all that big stuff with a careful eye, because we've worked with them in the past, for sure. And they have some really great techniques that they've done, in terms of the way they use instancing, and prefabricated models.

^{Epic's popular third-person shooter Gears of War}

I want to ask about the weapons switching, and other stuff. The thing that seems like it would take a lot of tuning -- because when I was watching, it looked very difficult -- is how long the [three-sided throwable player weapon] glaive takes to come back. Because when you're out of ammo, you're just waiting. How much work did that take, to figure out how long it should be, like how long it should be away, how far it should go -- that kind of stuff?

SS: I would say that one level you saw where you didn't have any other bullets for your offhand gun -- those two fights are the only fights where that happens, and then after that, you're constantly in this kind of dual-wielding mode. What that allows us to do is actually to make the return time for the glaive a little bit longer, so you are balancing the guns with the glaive.

That's just been something that's organically evolved, just nudging the values here and there -- that sort of thing. But once we did the dual-wield, we went back to the market and re-specced all weapons and added two more pistol types and pulled out two back weapons, because the pistols take a greater emphasis in the game now, because of the dual-wielding.

I know you're going back in and doing more multiplayer stuff now. Do you feel like... well, I've been feeling like, I should say, that multiplayer hasn't killed single-player, but it's almost supplanted it, as the thing that people focus on in many ways. It's like... it could just be the press, but I feel like people aren't spending as much time, sometimes, figuring out how to make a good single-player experience, as long as the multiplayer is really solid. I don't know.

SS: Right. It's been the inverse for us. We've had a lot of experience with multiplayer, working on those Unreal games. They were multiplayer games. Any single-player component to speak of was a way of rephrasing what was multiplayer. Again, that's where I think the studio stretched quite a bit, in terms of figuring out how to tell a story. For us, it certainly hasn't supplanted or even competed in terms of our focus.

It's definitely been single-player from the beginning, and in fact there have been points very late in development where we didn't think we'd be able to get multiplayer at all. That was pretty painful, because technically, it was not a problem. The engine has been a networked client-server engine from the very beginning. So for us, the multiplayer is cool, and it's uniquely Dark Sector, but it hasn't eclipsed at all the single-player. It's definitely firmly there.

There's a lot of the kind of modification of skills and abilities that happens. How much is too much of that? How do you decide, "Okay, people have a bunch of stuff they can do now." When does it start getting too complicated?

SS: I would say that one of my fears is that we're pretty darn close to that threshold -- that saturation point. One of the earlier concepts of the game that we had until like a month ago was this sort of XP system, where number of glaive kills equals points toward getting the next evolution. A huge problem with that ended up... for example, you saw the power throw, when you hit the gates and you slam them open. We couldn't do that, because we didn't know when you were going to get it, which meant that we couldn't present you with a challenge that you could prove your knowledge of it.

In fact, we found that when we focus tested, putting the two gates is like ten times better than putting one, because they'll accidentally do one, but by the time they get to the second one, they're kind of rocking. So we took out the XP system, and said, "Okay, you're going to get it here, here, and here." And that allowed us to pace a challenge with it.

What you saw, I think, was that we rapidly layer on the aspects of the glaive, and then you're going to play for about an hour and a half with that stuff, and then there's going to be a boss and you're going to get another simple power, but the complexity of the glaive is important, because that's our defining weapon in the game. We wanted to give it as much as we could.

^{Digital Extremes' Dark Sector, in development for the PS3 and Xbox 360}

Yeah. You seem very reliant on people liking the glaive.

SS: Absolutely. The dual-wield kind of relaxes that tension a little bit, and we did go back, like I said, and gave you a whole bunch of different pistol weapons. But yeah, if they don't like that, we're kind of screwed! (laughs)

Does it make you nervous?

SS: Yeah, it does. I mean, as soon as someone rips someone in half in slow motion with the aftertouch, I think they'll hopefully appreciate it.

The achievement stuff that you're doing on the PS3...?

SS: Yes sir.

Is that visible by other players?

SS: It is not.

So it's just for you within the game?

SS: It's just for you within the game.

Does it do anything else? Does it unlock anything?

SS: No. It is merely just to present you with a challenge. "Do 30 headshots," do this and that.

So it's more like a high score type of aesthetic.

SS: Absolutely. It's pretty simple, how it's implemented.

I was just wondering, because... I know some people who are like, "Okay, I've got two versions of this game. One of them's got achievements. I wonder which one I'm going to play!"

SS: Yeah, for sure. I guess it's just "Where is Home at?" and how much beta code can we stand to use. I would love it if that was integrated in there. I'm not sure. Again, I don't think we can stand to do another delay.

No. But you could probably, since you've implemented them, I bet you could maybe patch it?

SS: Yeah, if there's an auto-update and we add more maps and game modes. That would be awesome to squeeze that in there, for sure.

What do you think is the importance of story in these kinds of single-player experiences? As you know, I like story, as I mentioned earlier, but still a lot of the time I want to skip it. How do you make it so that people don't want to? Because sometimes I don't want to, like in BioShock and Call of Duty 4.

SS: Sure. How do you do it? From what I've seen, it has to be outrageously high quality. That's part of the thing we're trying to do, is to polish them up so that you don't want to skip them. But I think they're important.

I think at last year's GDC, they were talking about some Harvard research on gamers and paths, and one of the top ones wasn't graphics and so on -- it was role engagement. "Why am I here? What is my role in this world?" That seemed to be one of the highest.

Beyond multiplayer, and beyond sexy next-gen graphics, they wanted to feel like there is a reason and a compulsion to continue. So I guess stories are too old to smack you in the face of role engagement. I think there are probably other tools that might be better, but I don't know what they are.

I think those other tools are probably like, situational happenings that make you feel like you're in the universe and integrated there.

SS: So-called non-invasive cutscenes or whatever, right?

Right. Scripted events that make good sense, like the ones which happen in Call of Duty 4. The best implementation of story I've seen is probably BioShock, because you have to search for it. You have to find the story.

SS: You're talking about the little voice-overs?

It's cool, because you can ignore it if you want, but it's kind of neat.

SS: Creepiest moment in games is in System Shock 2, putting those recordings on. Did you play System Shock 2?

I actually didn't. I'm remiss.

SS: It's tricky, but I remember there's one part where you get the recording of the doctor who's gone insane, and you can hear it in his voice -- the straining and the insanity -- and he's talking about Nurse Bloom

And of course, just at this moment as he's speaking through the dialogue, you turn the corner -- and it's 1999 graphics -- but you turn the corner and you see this body on a table that's obviously female, and you feel a flush going up your face. Some of that stuff is brilliant for sure.

Yeah, that's good stuff. Have you anything to say in the Quebec versus Ontario debate that has recently...?

SS: Is there a debate?

Just a small one.

SS: Does this have to do with government funding and so on?

It does. It has a lot to do with that.

SS: Interesting.

Because you Ontario guys don't get it.

SS: We don't get as much. Well, I think Telefilm has now opened it up for game developers in Canada, which I think is going to be pretty helpful for us as well. I don't know. I think this probably for my boss to talk about more than me, because he has to pay everyone. (laughs)

Fair enough. Do you have any plans to sell your technology at any point?

SS: Well, maybe. But I think that the kind of people who I look up to, technically -- the Naughty Dogs and the Insomniacs [don't]. It's hard to say. I think it would be nice, but it would definitely be gravy, and it wouldn't be a company focus. I don't know. Is that part of their competitive advantage? I'm not sure.

In some ways, it seems like companies are selling a kind of supplemental business, in a way, and also to keep them going when they're not working on a project or stuff like that.

SS: Yeah.

But also, when people do weird things to their engine, they can be like, "Oh, look what they did. We can do that!"

SS: Sure. I think Epic sort of hit that perfect storm for market penetration, and those guys have... I think they've had to grow a lot to accommodate that. But they've always done a great job on their tools, as much as everybody likes to bitch and moan. Their stuff is nothing short of amazing.

It's true that not everybody wants to work on a tool that doesn't wind up getting used in the game that you are making.

SS: For sure. The f*cked up thing is, I think a lot of people will get these engines. You see a lot of slips, or you see massive teams -- even compared to ours -- where a guy comes back and says, "Did you realize that this team is like 60 guys with 10 people on programming stuff... and they started with Unreal 3 as well! What the f*ck are you guys doing?" You know? It's a little bit out of control. So, scope management's tough.