| |
|
|
||||
![]() |
||||||
| |
|
|||||
|
Leaping Lizard’s Centipede 3D
Last year
I was quite dismayed to see that Alfred Hitchcock’s Psycho was
being remade. It’s one of my favorite films and I couldn’t understand
why anyone would think it necessary to remake it. How could a remake possibly
approach the brilliance of the original? But then I noticed that Gus Van
Sant, a filmmaker whose work I admire, was heading the project, so I thought
the remake might have some merit. Still I wondered, what could have provoked
him to tamper with such a classic? The irony is that at the same time
I was dubious about the prospect of a Psycho remake, I was working on
a new version of The Task at Hand With plans for the PC and Sony Playstation as intended platforms, Hasbro Interactive was very clear in explaining that they wanted Centipede 3D to be true to the spirit of the classic Centipede. Hasbro Interactive had a commercial hit on its hands with Frogger 3D, but at the same time was listening to complaints about the game. It seemed that many people enjoyed the first few levels of Frogger 3D the best. It just so happened that these were the levels that most resembled the classic Frogger. Because of this, Hasbro Interactive wanted to make sure Centipede 3D was closely tied to the original Centipede. Leaping Lizard had already spent some time developing its outdoor 3D engine, which up until Centipede 3D was used exclusively by a flying combat game called Raider. Hasbro Interactive saw the technology and thought it ideal for their new version of Centipede. A prototype using the Raider engine with Centipede game-play components was created by Leaping Lizard in record time, and once delivered to Hasbro Interactive, the deal was soon signed. Eager to work on a project with the challenge of creating a new incarnation of Centipede, the team at Leaping Lizard was concerned with exactly the same goal as Hasbro Interactive: creating a distinctly modern game that still captured the feel of the classic. At the end of 1997 work began, using the existing Raider engine but without most of the Raider game play in order to make a new version of Centipede. San Francisco art shop Mondo Media had impressed Hasbro Interactive with its excellent work on Interstate ‘76, and it was brought in to do the cut-scenes for the new Centipede. Hoping to match the art style of the cut-scenes with the in-game art, Hasbro Interactive and Leaping Lizard decided Mondo Media would also do some of the animated gameplay art. The core Centipede game play demands that there be a large number of monsters and mushrooms on the screen at one time, and as such, a very low polygon count was required for all the models. The Raider engine utilized a level of detail (LOD) system, which swaps in lower polygon versions of models depending on the game’s frame-rate and a given object’s distance from the camera. Mondo Media, more experienced with high-polygon work, at first found it extremely challenging to stick to the 90-60-30 polygon limitations for the monsters’ LODs. But as the project progressed, Mondo Media adapted to the polygon limitations and produced some great work, creating more than 20 animated character models for the game.
Designing a Remake At Leaping Lizard, work began immediately on the project. Programmers started porting the Raider engine over to the Playstation, while designers and artists dove into the first world of the game, Weedom. Six levels were produced rather quickly, with the first two levels attempting to mimic the classic game play as much as possible. Hasbro Interactive didn’t think these levels were close enough to the original, however, and by March of 1998 asked that we start work on a completely separate "classic game," turning what had been a one-game project into a two-game endeavor. As we studied the classic game by endlessly playing our authentic, coin-operated Centipede, we not only developed the mock-classic game, but started to rethink the design of the modern game as well. Up until this point, the modern game levels had not been very reminiscent of the original Centipede, and as we studied the classic, we began to see why. Of the six original levels that we had designed, only two survived without major retrofitting of the landscapes, while one was completely overhauled and three were discarded. We then spent approximately three months working up three new levels, refining and balancing them until they were fun and reminiscent of the original Centipede. With these first six levels relatively finalized and armed with a better idea of what was going to work well in Centipede 3D, we spec’d out the rest of the game and designed and implemented 23 more levels in only three months. By working initially on only a few levels until they were actually enjoyable to play, we were infinitely more prepared to design the remainder of the game, and hence had to do almost no reworking for the rest of the project. The Playstation Version Around April, both Leaping Lizard and Hasbro Interactive became concerned that Leaping Lizard wasn’t going to be able to get the PSX version of the game running by the holiday season deadline. This was largely because of difficulty finding an available PSX specialist, and a number of PSX programmers that Leaping Lizard had hoped to hire had backed out at the last minute. As such, both Leaping Lizard and Hasbro Interactive thought it would be best to bring in another team to handle the PSX conversion, preferably someone who already had a suitable engine running on the PSX. That team turned out to be Real Sports Games of Elgin, Illinois. Real Sports Games had a very nice looking Playstation engine up and running its then-in-development Jeff Gordon XS Racing title, and thought they would be able to get the conversion done for the all-important Christmas deadline. The deal was finalized at the 1998 E3 Expo, and they started work on assessing and porting the title immediately.
Real Sports Games was given an extremely challenging goal: porting a game which was not yet complete. This made it more difficult to fully assess the scope of the project and to see how it would fit within the constraints of the PSX. Further complicating matters was the fact that, in order to get the game to work with their engine, they felt they had to rewrite large sections of the PC game’s code from scratch. Though the decision was made to use the exact same level files as the PC version, the project became less of a straight conversion and more of a reworking of the PC version of Centipede 3D. Real Sports Games ended up neither doing truly concurrent development with the PC team, nor porting a completed game. Following completion of the PC version in October, I was flown out to Elgin to work with the Real Sports Games staff on wrapping up the conversion and to make sure all of the correct design and game-play elements were functioning correctly. My stay was initially to be for two weeks. Once there, however, I observed how many of the game-play elements had yet to be implemented, and I began noticing how different systems in the game, though they might work on earlier levels, were not going to be adequate in later levels. My stay in Elgin stretched to three months as I started working on the coding of the game, and got all of the game-play elements working correctly. During this time, other members of Leaping Lizard were flown out to help finish the project, including programmers Chris Green and Gary Skinner, artist Jane Miller, and project manager Elaine Albers. In December, the game entered beta, where it stayed for four months, going through a particularly long and arduous quality assurance process. Fortunately, after December I was able to work remotely on bug fixes and to fine-tune game play using the excellent StarTeam source control system, which worked seamlessly over the Internet. Working on Centipede 3D was at once an extremely gratifying, yet terribly confounding experience. The game’s design seems to have worked out particularly well, with play that instantly reminds people of the original game. Getting that design implemented on the PC side was a challenging, yet rewarding experience. But then making that design function on a system significantly less powerful than the PC proved mind-numbingly difficult and unendingly frustrating. The entire team was disappointed when the product didn’t make its Christmas ship date. Looking back on the whole project, there are some things that worked out gloriously well and other things that only cause us to hide our heads in shame.
|
|
|