Q&A: Square Enix's Murata Talks Crystal Tools, Unreal Engine Initiatives

Originally known as the White Engine, Square-Enix's Crystal Tools initiative has taken shape over the past few years as one of its key efforts to standardize cross-platform technology for its forthcoming titles, being used not only for Final Fantasy XIII, but also its forthcoming MMO. While the company has also licensed Unreal Engine 3 for some future development, Crystal Tools general manager Taku Murata has previously said that its internal engine was designed to cater strongly to the demands of Square Enix's developers. To learn more about the history of the effort, why Square has licensed Unreal Engine, and whether Japan is being held back by not licensing locally developed technology, Gamasutra talked with Murata, who elaborated on how the work behind Crystal Tools was an amalgamation of all the experiences gained on working on large scale productions like Final Fantasy. To begin, can you give me some of your history in games? TM: Since I joined Square, I've done Secret of Mana, and the second in the series of that -- Seiken Densetsu 3 -- I don't know if it's been released here. Also Final Fantasy Tactics and Vagrant Story. I was the lead programmer. I was also the director of PlayOnline, and technical supervisor for Final Fantasy XII. Why did you decide to change the name of Square Enix’s internal engine from White Engine to Crystal Tools? TM: Actually, the initial name that we used -- White Engine -- we just used it as like a code name. But in September of 2007, version 1.0 was released. At the time, we were trying to figure out the official name for that, and we wanted to come up with something that would reflect the company. "Crystal" is something that we wanted to use, because it many different colors, so we wanted to use that meaning too. That also sounds like our company too, Crystal. How challenging was it to build this engine? Did you build it completely from scratch? If not, what resources did you have previously? Or was it just completely designed from the ground up? TM: Actually, we didn't use any resources. We started from scratch. It's basically all based on our experiences. We kind of picked what we found was effective or useful, so we basically put all these good things in this one engine together. What were the most important considerations for Square Enix, in terms of developing an engine? How did you make the engine versatile for the type of games that Square Enix wants to make going forward? TM: Actually, we didn't consider making a base that can work for all titles. Not to cover everything, but to cover the base. Then we have the ability to use plug-ins that can accommodate other tools. Depending on the title, you can always just add something else that doesn't exist in this one. It's always a challenge to do multiplatform -- how did you address the multiplatform issue for contemporary systems? TM: Actually, there's SPUs for PS3, and multicore for 360 already [optimized]. What really matters is the size of VRAM. But the VRAM size can be handled by adjusting texture sizing, so that can be taken care of. In that sense, I don't think it's that big of a deal that you have to start with the PS3, technically. The PS3 is a very powerful machine, and it's very expansive. So as long as the converting process can be handled correctly, it shouldn't be a problem. A lot of companies in Japan have mentioned that they feel that the Japanese game market is falling behind in terms of technology, partially because PC development and advances in that sort of technology have been happening mostly in Western markets, and support is in English. This engine, though, seems quite high-end. It's very unusual within the Japanese industry right now. Do you agree with that? TM: Until the PS3 was launched, regarding the 360 and PS3, it is true. I agree that maybe Japanese game technology is behind because there were no previous experiences, or a base in that PC market in Japan back then. Since then, things have changed. Now, I think it has come to a point when the PC architecture and technologies have advanced, particularly in terms of the Japanese way of expression, such as, "In this game, we feel very strongly that we are very capable." I do think that in terms of the language barrier, yes, it still is a challenge. We do face those challenges in terms of communication, but in terms of technology, I don't think we're behind. We probably have conquered that problem...or probably we have gone across that point and advanced. Why did the company also license Unreal Engine 3? TM: We used the Unreal Engine for The Last Remnant because we had a particular Western -- European and American -- market in mind. I believe that the game engine reflected the culture and philosophy of the market, and the creator. Actually, I wouldn't say that we would use Crystal Tools for everything. To create Final Fantasy, yes, Crystal Tools is ideal and it's good. But for other titles, we could use other engines as well. In that sense, it's always an option to use another method. Also, I don't mean that the Crystal Tools can only create Final Fantasy. Could you elaborate more on how you say this engine is particular to the Japanese mindset? Is that in terms of the pipeline and workflow? TM: Maybe I shouldn't say it's very Japanese-like. That's probably too vague. I'll say it's very Square-like. All the experiences that we've gained from large-scale production of Final Fantasy is being input into this engine. Considering that Japan's developers have complained about Western engines being primarily in English, would you ever consider licensing out the Crystal Tools to other developers, like Epic does in America? TM: We have considered that option, but at this point, considering preparing the documentation and the support costs, even if we don't think it's impossible...at this point, we are not doing that. In terms of in the future, if you were to license the tech, you'd probably be the first company to license an engine in Japan. Do you think that this kind of thing would be accepted at this point? Because I know in the past it was very difficult to share technology in a public way over there. TM: Things have changed compared to before. Koei actually licensed the Dynasty Warriors engine for Dynasty Warriors: Gundam. Capcom's MT Framework, I think they're licensing that too. Things have changed, so we're not the first one. Why do you think this change has come about? It's all quite recent, within the last couple of years, I think. TM: It's only my personal point of view, but before, they used to just own their own technology without sharing, and that's how they kept their play safe. But recently, they have probably realized the importance of sharing information on technology between developers, particularly for the next-generation consoles. It used to be that the European and American markets have strong PC backgrounds, so they're much stronger in that sense. So probably Japanese developers have realized that as well. What do you see as the benefits of licensing an engine, versus building your own in-house? There's a lot of debate about that, even in the western market right now. TM: It's much quicker to buy something, to purchase something that's already existing. That's easy, and it's probably less expensive than to create your own. The benefit behind having an internally designed tool is basically... the Crystal Tools we have developed because we wanted to use all the know-how that we have gained from experience in the past with the Final Fantasy series. We wanted to get all of the good experiences put into these tools. That's the benefit of this tool. For existing licenses, do you feel it’s worth the trouble when you have to actually learn how to use it and fit it into your pipeline? It may be quicker to get, but it's harder to make it work right for you. TM: That's why the Crystal Tools are much more familiar for Square Enix people.