Q&A: Crytek's Seeley On CryEngine 2's Asian Invasion

At the recent Gstar international game expo in Korea, Gamasutra had the opportunity to talk to Crytek's engine business manager Harald Seeley, who was there to spread the word about his company's CryEngine 2 to potential publishers and developers. Seeley says that Crytek has successfully licensed both CryEngine 2 and its previous incarnation in Korea, and he recognizes the importance of giving more game professionals in the region to actually have hands-on with the title. But establishing a presence in the burgeoning Korean game market isn't a simple matter, especially for engine companies, as they cope with issues of localization, finding the right talent base of coders, and expanding the niche console market in a landscape dominated by MMO gaming. In this in-depth chat, Seeley discusses meeting the needs of the Korean language base and becoming more MMO-friendly: So are your tools fully localized into Korean, or do you have Korean support? Harald Seeley: We don't have any localization of the tools in any language, except English. A lot of companies are having trouble right now in Asia with game technology, because their local technology has fallen far behind. So they're having to adopt English-language technology, which is also extremely difficult for them, because they're not natively speaking English. HS: Neither are a lot of the people at Crytek. We understand the problem, and we're definitely looking into doing localization of our documentation, for example. First we have to complete it for the English version. We've been working very hard on that now that Crysis has shipped, putting together tutorial videos and step-by-step instructions together. Once that step is completed, then we will probably look and see if we can get the documentation translated over to other languages. Korean is definitely very high on that list, and some of our existing licensees have offered to help us with that task, because they definitely want to see that happen. It seems like the first proper engine that localizes itself into an Asian language will have a significant lead, because pretty much everyone I've talked to on the Asian side -- like from Japan or Korea -- is having these kinds of technology problems as they fall behind in game development because of translation issues. HS: We understand the problem, obviously, since we're a multicultural company with 26 of our own. We know that English is our common language, but at the same time, if we could find a local partner, we could work together with them to help us make the support here more local, both in terms of access and in terms of language. I just noticed that Gamebase is a local partner for Gamebryo. They're doing it through a third party -- would you consider that? HS: What we consider is not something I can comment on right now, but one of the reasons why everyone is here right now and not just the engine licensing team is to seriously look at different companies here in Asia, to see if in the future, if we decide to expand here, who would be the right partners. That's how we got into our existing satellite studios right now. It was through a similar process. Do you have a separate engine team? HS: We have an R&D team who is managing the engine and the tools. They're separate from the development teams. Do you have a support team, or is it not quite there yet? HS: We're building a support team. Yes, we have several support people here today. Support involves not only answering technical questions, but also providing on-site training, which means that people get hands-on, one-on-one consultations with issues with the engine as they run into them. What is your perception of the Korean market right now? I've been asking a lot of Korean companies this question too, and they have varying responses. HS: I think it's getting really crowded, and I would not be surprised to see a shakeout, in terms of the number of companies who actually succeed to stay around, and those who fall by the wayside. We think that using an engine as good as ours will ensure that a company that comes out with a product three, four, or five years from now -- some of them take that long to put together -- will ensure that they have a product that still is viable and not dated and worn out by the time they actually get it to market, which I think is a big risk for a company. If they're going to invest the time and money it takes to build such a big project over such a long period of time, they don't want to get to the end of the road and realize that they backed a horse that now looks four or five years old. The CryEngine 2 was not created with MMOs in mind, was it? Obviously it was created to support the game. HS: It was created to support the game, but it was also created with the idea that it would become standalone middleware which was capable of supporting more than one kind of game. MMOs were definitely always figured into the process, but in the rush to get a game out, you always take some shortcuts that later you have to go back and do more thoroughly. That's an ongoing effort, because we have so many MMO licensees, and we have good relationships with them such that we don't want to see a lot of replication of work. We want to make sure that everybody who has a good idea and makes a change to the engine and makes it more MMO-friendly... that that gets shared with the rest of the community. So we foster that kind of interaction between them. How are you absorbing those ideas without making it too bloated? HS: Well, we have a team that's going to review each of those ideas, and that's the engineering team that built the engine to begin with. Since we have our own efforts internally to support for multiple studios now and multiple genres and platforms, it's important for us, internally, to make sure that the engine is as generic as possible, because otherwise, each team will end up with their own version of the engine. And I've worked at publishers where that's the case, where in every project, they have an engine that's just for that project. It's really hard to leverage advances in engine technology across multiple teams when you do that. That's definitely a trend that I'm noticing a lot still, with people working on the same issues separately, which is a serious problem. So you're keeping your licensees all in the loop on the developments that happen? HS: Yeah. Some licensees are more involved with us than others. Some are still in the very early stages, and still figuring out what their designs are, running the editor, and building their first content. Others are much more advanced, because they've already got an existing product that they're trying to integrate with, so they're focused on those particular issues right now. With those people, we treat them as they're one more studio that's using the tool -- no different from any of our internal or external studios. What kind of things have had to be added in order to make it more MMO-friendly? HS: Well, the first thing that had to happen was that we had to make bigger distinctions between the different aspects of the engine, and decouple them further. We made a big effort when we first designed this engine, to keep everything very modular, but occasionally, hidden linkages creep in through macros and through other means you didn't expect, and as you're going through and cleaning up that code, to find those issues, and piece them apart. One thing that I've heard from the developers here is that it's extremely difficult to find good coders that, first of all, actually want to be coders, and second of all, can do a really good job, and not just what they're told. They can actually go out and think about certain thoughts. HS: I think the answer to your question is self-evident. We've got some really, really good coders, but coders come in all flavors. There are those that are very disciplined, and write really well-structured and designed code, and then there are those that think out of the box, but tend to write a new version of the code every couple of weeks in a completely different manner. Those are a little bit harder to manage, because they're really creative, but they can be really disruptive, once you have a very large codebase like we do. We've got a very strong R&D manager we've added to the team in the last four or five months, and he's made a huge difference, because he understands the issues technically, and can mediate between those who are looking for very firm, rigid guidelines so that everything is very clean-structured, and those who want things a little more freeform, so they can maximize the performance. Specifically in Korea is where they were talking about that sort of thing - a dearth of "true" programmers. HS: The game industry is still relatively young, and that's part of the problem. You don't find people who have decades of experience coding games. I think it's also an issue of maybe getting some really good university programs together. Yes, I think they're trying that. HS: I've been talking to developers about who they would recommend to us to try and partner with, in terms of schools and universities where we can teach our tools and technology and provide it to them, as a means of teaching future programmers and future designers. It's hard to find any particular one that everybody agrees that really stands out as the up-and-coming game academy, if you will, of Korea. There's a lot of smaller schools, and a lot of bigger schools that have a program or two, but nobody really stands out as being a leader in this industry yet, which is amazing, considering how big this industry's become in such a short time here. The console market here has yet to take off, obviously. It's been very late getting started. The PC bounds here have really driven the PC market very heavily. It's not only Korea -- it's expanded into the many local Asian countries around here as well. That means that we think the CryEngine, once it's been developed a little bit further, people are going to start seeing some progress on the MMOs that have licensed our game engine technology. Once they start seeing the screenshots, and start getting some reviews and previews of gameplay, it'll obviously open up a lot of doors. We're not in a big hurry here. We're not trying to suddenly flood the market with a lot of engines. It's a very pricey engine, and it's of very premium quality, and it takes a big investment in people to train them to produce the kind of assets you see in our games and our demos and our videos. Companies that are willing to step up to that are few and far between. It seems like training students up in your engine would be pretty useful for you, because then you've got people who are used to your technology want to do that when they go ahead to the gaming world. HS: We're also trying to foster the self-taught hobbyists. We've gone really far out there to provide them early access to the technology before the game even shipped. We've got several mods underway right now. We've put the sandbox out there along with the demo, along with a lot of assets we kind of snuck in there for people to find and have some fun with. We figured that would give the demo a lot longer lifespan, if they could turn it into their own demo. It's really happened, and we're really gratified by that, but it's also given the self-motivated individuals a chance to learn our tools and technology. A lot of them have aspirations to become professionals later, so this is a very inexpensive and very productive way for them to do that. And we have people on the Crytek staff who started out as FarCry modders. We hope that we can expand on that to make it useful for all of our licensees, as well as ourselves. It seems like a good way to get reliable help that you know you can count on, because they taught themselves your engines. HS: And they're also producing...it's a very active community, and it's a very outgoing community. A lot of them are producing assets to teach others how to do things. We've only got a limited staff who can only work on a certain number of tutorial videos at the same time, or documenting certain areas at a time, but the fans are just great. They've been creating all kinds of new content for us to distribute on our crymod.com website, to teach people how to use the tools. Even our commercial licensees are finding crymod as a useful resource. The FPS market here is pretty good. All of the competitive gaming things... I don't know how long they'll be around, but it seems like it's made for that. Okay, I think that's about all I need to ask, unless there's anything else you want to mention. HS: The thing about CryEngine 2 is that we made a concerted decision early on to build it in a way that was future-proof. We don't say that lightly. We could have taken the CryEngine 1 -- which was a solid engine and looked pretty good -- and we could've continued to evolve it and improve it, and call it CryEngine 2, but instead we decided to throw away some existing approaches and work strictly by the optimum visual quality we could get from the newest hardware, with the idea being that's what's going to scale the best in the future. If we had done something focused, for example, that was very low-stress on the video side, like texture-streaming for example, and put several years of effort on making the world's best texture streamer, that's something that doesn't leverage all the extra transistors in the video card, so it's going to fall behind in an approach where people are using those transistors more effectively. Similarly, if we had just gone along and improved CryEngine 1 using the same techniques we were doing and just modified it with the new hardware coming out, it might have a faster approach. We would've had a strong engine, but it wouldn't have separated it very much from any other engine of that triple-A quality out there. We made a concerted decision to do things the very best way they could be done with the hardware that's available today, and then scale it down to hardware that was two or three years old as best as could be done. And I think we've done a really remarkable job on that, given where we've started from. We think now that we've done that, now the hardware will catch up, on the performance side. Then there will be no more excuses for why people have to settle for anything else than an engine like ours can offer.