Saturday, February 16, 2019

1999 Shenmue Pre-Release Interview with Director Keiji Okayasu | Dreamcast Magazine

1999 Shenmue Pre-Release Interview with Director Keiji Okayasu | Dreamcast Magazine

We have translated a 1999 interview by Dreamcast Magazine with the Game Director of Shenmue, Keiji Okayasu, which took place one week before the release of the first Shenmue game in Japan. This topic was selected by the Phantom River Stone blog patrons via our monthly poll on Patreon.

The interview translation starts below. Some additional images have been inserted.


Keiji Oyakasu was one of the main AM2 leaders involved with porting the Saturn version of Virtua Fighter (1 and 2) from 1994 to 1995. With Shenmue, he has completely dropped off the radar these past 4 years. We track down what has been going on during the 4 years he went “missing in action”!

KO = Keiji Oyakasu

Q: It has been so long that there’s a lot we’d like to ask about. First of all, could you tell us about recent events?

KO: You might suppose I don’t have much to do that Chapter One* is completed, but actually I’m incredibly busy now [laughs]. That’s because Shenmue 2 [the working title] is already getting underway. The fundamental work has already started, and we’re putting together the foundations of the story, If this isn’t done properly, we’ll have trouble later with things like the software and the design, so I’m busy at the start.
*The first Shenmue game was called Shenmue Chapter One: Yokosuka in Japan.
However, up to now I’ve also been responsible for looking after the software program and other parts as well, but from 2 I’ll be completely dedicated to a director-like position, keeping an eye on it as a whole. I’ve also got to get stuck into working on one more game I’m doing, Rent a Hero.


Q: After you successfully ported Virtua Fighter 2 to the Saturn at the end of 1995, you vanished from sight, didn’t you. In the end, the reason you disappeared was never revealed…

KO: You can assume that I was already in the Shenmue team from that time [laughs]. It’s hard to say exactly when the start point for Shenmue was, but core research was already in progress at a stage long before it had begun to take shape. At the time, we had become quite proficient at moving people using motion capture, and the idea was raised of creating an RPG using this technology. Of course, the person who thought of it was our section chief (Yu) Suzuki, and it led to instructions to continue as-is with the core research. And since the technology looked somewhat promising, the idea came about of using it to actually make a game. That was after the Saturn version of Virtua Fighter 2 was completed. It was around half a year after the RPG project had been decided.

So at the beginning, Yu-san gave me two choices, at the time I was still working on the Saturn version of Virtua Fighter 2. Yu-san explained to me, “I’m planning to start on this kind of RPG project next; what do you think?” He said I could either choose to work on that, or else be in charge of porting Fighting Vipers to the Saturn [laughs]. At the time, part of me felt like “Please, no more fighting game ports!”, so I replied that I’d like to do the RPG project. That was truly the only reason.

At the time, Okayasu was working on the Saturn port of Virtua Fighter 2
Q: At the time, did you think it would be over quickly?

KO: Well, I didn’t expect it to finish simply, but in about the normal time for creation of an RPG. I thought it would be done in about a year or a year-and-a-half [laughs].

Q: Was the concept for it to be a game with 3D characters, motion capture, and fully-voiced?

KO: At the time there weren’t things like a well-defined concept or specification document, but we set off from the point of asking ourselves, “Just how far can we build an RPG in 3D?” Naturally, this was before there was any talk of the Dreamcast. And so I joined the team in around February or March 1996.

Q: From that time, you no longer showed yourself, did you. But various rumors reached us through our contacts in the industry [laughs]. Such as, “An epic RPG will be coming out from AM2 targeting the end of 1996”, or “No, there’s been a slight delay and it will now be 1997” [laughs]. I guess you could say it was a case of “where there’s rumor, there’s fire”.

KO: The team members at the beginning were all ready to make it on the Saturn, but in fact we’ve started over about twice. At first we started with the intention of making it just for the Saturn alone, but the scale of what we were trying to do grew so big that we realized that it would be impossible on the Saturn. From there, there was talk of requiring a RAM expansion or similar fitted.

However while we were moving forward on that, the next-generation hardware of the Dreamcast showed up… The thinking changed to being that if we’re going to be building it from now, we should do it for the Dreamcast. But we had intended at first to put it out for the Saturn, there’s no disputing that. That’s because we knew that the graphics we had made at that time had reached a level that was unbelievable for the Saturn, and we were confident.

So when it was decided to make the change from the Saturn to the Dreamcast, I think the ones who felt the most frustrated were the designers. Despite having created graphics showing “See how far we can push the Saturn!”, once you move them to the Dreamcast with its high specifications, they’ll be taken for granted, right? Just supposing they had come out on the Saturn, I think everyone’s reaction would have probably been “the graphics are amazing”. For that reason, it was frustrating that the hardware changed, and felt like a lost opportunity.

But, well, comparing them visually now of course the screens we made on the Saturn are outshone [laughs]. Having said that, it’s the techniques developed on the Saturn at that time that have made the images on the Dreamcast possible.

Q: By the way, from around when did Yu Suzuki fully devote himself to Shenmue? I get the feeling he began leaning towards it after completing the arcade version of Virtua Fighter 3 in September ’96…?

KO: By about ’97, he had already become to be totally immersed in Shenmue. Well, perhaps not totally immersed, but he was starting to get pretty serious about it. In particular, by around spring of ’97 he really meant business.

Q: What sort of size was the project team at that time?

KO: It was larger than usual. Having said that, I don’t think it would have been more than around 50 people.

Q: Speaking of that time, in February 1996, Final Fantasy VII was announced for the PlayStation, being released at the start of ’97. In a way, it corresponds to how AM2’s RPG project moved to becoming something fully-fledged. I suppose you would have been aware of Final Fantasy and others, at that time?

KO: Naturally, we were aware of it. What we attempted to do with respect to that, was to see how far we could go with the concept of creating something that was “movie-less”: whether we could produce high-quality images without using pre-rendering… that’s what was decided.

Final Fantasy VII for PlayStation, released in 1997, made use of pre-rendering for backgrounds and cut scenes.
Q: In the end, how far did you get with the Saturn version of Shenmue?

KO: The scenarios were broadly completed, but there were still a lot of places that weren’t well-defined. That’s due to the particular method we used to make it. What we do here is, firstly, we put aside for later things like the connecting parts in the game and the “in-between” parts, and start with the key parts, you see. So if the quality of graphics for one part improves, we raise the quality towards that level in the other parts too. Then we repeat that work process cycle. So even for those of us at the studio, a sense of when the game would be released didn’t really hit us.

Screen capture from footage of Shenmue on the Saturn.
Q: Even so, you gradually increased the number of staff, didn’t you?

KO: That’s true, we did increase them, but only the one time. After that there was a long period where we continued to make it with a fixed structure of members. What was roughly 20 members at the start, then doubled in size and stayed like that during the move to the Dreamcast, then suddenly shot up again.

Q: In contrast to that, at that time you had started keeping a low profile, hadn’t you?

KO: That’s because we ourselves were doing so. Although I was in AM2, I stopped looking at what the other teams were working on because I was so involved with this project. Strangely enough, our project itself was carried out with a high level of secrecy. Even within AM2 itself, at one time we wouldn’t even let staff from other teams see what we were up to. That’s how immersed we were.

Q: What area did you work on?

KO: Regarding the story and so on, Yu Suzuki, our section chief, worked on it the whole time and was very particular about it. Yu Suzuki looked after things like the story and dramatic direction, while I did the game part, roughly speaking.

Q: In the meantime, the Dreamcast was announced in May of 1998. What were things like at that time?

KO: By May, I think we had already started using the Dreamcast, and were producing pictures. We were busy testing the Dreamcast’s ability. By around summer things began to take proper shape, and we had started producing something for showing at the end-of-year premiere event.

Q: There were loads of events even after the premiere, so that must have made development difficult, did it?

KO: It did have an impact, yes. After last year’s premiere, there was a stream of various events, about one every two or three months, and there was quite a bit of work needed for those.

Keiji Okayasu with Yu Suzuki at the Shenmue Premiere in 1998.
Q: Several times, a release date was announced but the date was then delayed again. Did you think they would be feasible at the time?

KO: Well, of course we thought we could hit them so that’s why we announced them. But, to be frank, we were probably overly optimistic. Our comments regarding the delays were not fabricated or anything. To be honest, we had packed in too many different elements, and in the end we weren’t able to finish adjusting them all in time. Time being needed for debugging was probably the biggest factor.

Q: That’s something I can understand, honestly, looking at the recently-delayed release of Dragon Quest. Huge games require huge amounts of time.

KO: It was so huge that too much of it wasn’t able to be properly estimated. For example, take the full voicing. At the time the decision is made to have it fully-voiced, you can’t gauge how much data volume it will need. Something like how many discs the game would have, wasn’t decided until right at the end.

Q: Amazing that it was able to fit onto 3 discs! [laughs]

KO: That’s what I think, too. [laughs]

Q: During development, I imagine there were several parts where the direction in which Yu Suzuki was thinking, and in which you were thinking formed a kind of “tug of war”. Was that the case?

KO: The toughest thing was that we’re a game shop, so “games” are common sense, right? However, what Yu-san considers to be common sense is what’s “ordinary” or “everyday”. How to portray that aspect in Shenmue was the hardest part.

So, depending on the player, many may come across a scene that makes them go “Huh? Was this kind of guy in the game?” Or wonder “Why is Nozomi here in this place?” – some players may have seen a certain event by chance, while I’m sure many more players will have finished their game without getting to see it. I think those kind of little discoveries in Shenmue are fun. Chasing the story and completing it to the end is actually quite simple. However, there’s a lot of fun to be found apart from that.
Not all cut scenes will be seen by all players.
Q: By the way, how far along is Shenmue 2, approximately? With Shenmue 1 being brought up to its current level of graphics, I can’t yet really imagine how far the next one will be taken.

KO: We’re reviewing various aspects. The other day, I watched the video clips from the premiere from one year ago, and there was a huge difference compared to now. [laughs] Even we ourselves were surprised at the difference. So Shenmue 2 is done in broad outline, and now we’re revising it to the level of Shenmue 1. We won’t just be using the Hong Kong graphics we showed at last year’s premiere as they are. Of course, we’ll make use of those parts we can from what we’ve made before though. There’s no doubt that the graphics will become even more amazing.

Q: Will it be ready in a year’s time?

KO: Er…. If all goes well, I have a feeling it will be, but we won’t know for sure until we try. [laughs]

Q: Thank you very much. Good luck with the development from here, we’re looking forward to it.

--- End of interview. Translation by Switch at PhantomRiverStone.com
Become a Patron!

No comments:

Post a Comment