Prior to the current generation of fighting games I would have said that 2.5D was the only way to make a modern fighting game that didn't break the bank while preserving the look and feel of traditional sprite-based fighters. The format was in fact the future of fighting games. The groundwork laid out by Taito in Battle Fantasia and Sammy with the Rumble Fish had shown how closely polygons, with the use of textures and filters, could create characters that were almost passable for hand-drawn sprites. They had married the two worlds of traditional artistic vision with the programming ease of 3D models. For years I assumed that it was the only way to do fighting games proper in the current climate of cost-effective design.

If Street Fighter IV had tried to go down the road of traditional sprites, in high definition, then gamers would be waiting years for the release rather than months. The shareholders at Capcom would not have been happy, nor would the public, if the game did not top everything they had done in SF III. The relatively short development cycle and runaway success of SF IV drew a number of mixed emotions from me. I was glad that Capcom had decided to resurrect the franchise and happy that it was a hit at home and abroad. It showed that audiences still loved the genre and that Capcom could compete with other studios without breaking the bank. I wasn't happy that the development cycle seemed so short, that their new characters seemed pulled from rival games nor that they revised the classic SF II characters and canon in order to appeal to a broader audience. Seeing the game presented in 3D, with graphics that did not try to recreate the look of sprites through some sort of cel-shading or other filter effect, left me a little cold. As if they had just tried to abandon everything that made the previous titles great while rushing a new game to the market. This was pandering and I always thought that Capcom was better than that.

I was uncertain that it would ever be possible to return to the roots of the franchise, to make an HD sprite based fighter with more detail than previous titles without breaking the bank. It was 2.5D or nothing... or so I thought. SNK announced that their next King of Fighters title would be sprite-based. Not only that but they would be presented with more colors, details, frames and effects than any other game in franchise history. This was not solely an HD-upscaling project but a complete rebuild of the franchise. I was impressed with the concept and wondered how they were going to accomplish this within an acceptable timeframe and on budget.

How would their iconic characters get a facelift without making too many concessions to a 3D fans? Would Falcoon and the art staff at SNK try to make their new characters appeal to western audiences or would they stay true to their animé-influenced roots? Would these characters still be relevant despite the influence of mixed martial artists on pop culture? That is, would these sprites remain relevant to gamers?

SNK played it very close to tradition. They preserved the artistic influence and ended up creating some amazing sprites that were bold enough for modern fighting fans, yet retained respect for the classic titles. I began wondering how SNK would be able to create near 20 characters for the game at this size and with this amount of detail given the unusually high time associated with sprite creation. There was something that SNK was not letting other developers or audiences on to.

In the previous blog I mentioned that sprites require lots of work and pre-planning. The designer not only has to create the general look of the character but also draw them from multiple angles so that animators have enough reference material to recreate the figure performing a number of moves. If they wanted to change the scale of an arm or a leg, or move a fist the slightest bit, then they would have to redraw and repaint dozens of frames. How was SNK going to handle the pressure of making sure each character was in the right scale, and each frame near-perfect, in order to ensure no redraws?

More importantly how would they preserve the artistic intent so that each character was presented with a consistent look?

One way to accomplish this was to have the featured artist not only design the characters, but also create the frames of animation and paint them in. This is a tremendous workload with a lot of potential problems that have to be addressed before the first sprite is even rendered. If and when the sprites are ready they end up nothing short of breathtaking, well, breathtaking by late 90's standards. Such was the case for manga legend Range Murata for the characters he created in Psikyo's Dark Tenshi: the Fallen Angels and Atlus' Groove On Fight. The sprites were amazing and showed a level of detail and style that no studio had matched, or probably will ever match. Can you imagine the visual impact SF IV would have had if Ikeno and Akiman had hand painted every sprite?

Every designer places their own fingerprint on the universe. If you study the various representations of Ryu you'd begin to see differences of his appearance. Wether is was Capcom, SNK or Udon doing the sprites, each designer left their respective style. Akiman did not draw Ryu like Bengus, Kinu Nishimura or Alvin Lee. This influence, this aesthetic was very difficult to recreate in 3D. Which was why titles like Battle Fantasia and the Rumble Fish had so much visual impact when compared to other 3D games. They preserved the artist's original vision for the characters and universe. Would future Street Fighter games return to these roots, especially to help them stand apart from other 3D titles? Or would Capcom stick with what worked in SF IV and abandon their sprite-based legacy?

I'm certain that Capcom would not have abandoned the sprite if only they could have found a way to speed up the process and maintain a high standard.

There were ways to do this but most of them made the games looked sloppy and awkward. I'm referring to motion capture technology. The technology has sped up the process of animation but it ends up making stylized characters move weird, at least weird when compared to traditional animation. The second way of speeding up the animation process was actually a precursor to motion capture. It was called rotoscope and it involved a process by which artists pretty much traced over film of scenes, objects or characters they wanted to recreate in 2D. Rotoscoping was an important part of animation history it helped create the original Lord of the Rings animated feature. When used by an artist, rotoscoping could take concert footage and at only 7 or so frames a second and make that footage truly memorable. However it seemed to share many of the same problems with motion capture. Characters moved like people, rather than cartoons. The art of animation lay in the word art itself. Tracing a live actor seemed cheap, or rather to cheapened the artistic medium. Several studios went overboard on the use of canned animation, like Hanna Barbera and Filmation or relied too heavily on rotoscope, like Disney. Disney even copied their own frames on several of their animated features, further distilling the artistic importance of cartooning.

Animation legend Ralph Bakshi spoke at length about CGI animation, rotoscope and character animation. He said, and I agreed, that animation lies in understanding artform rather than the physical process. Anyone can create animation in either 2D or 3D but only an artist that understands the nuances or cartooning, motion and design, could ever make that animation come to life. Neither motion capture, nor rotoscope could be properly used to create games which were meant for a 2D experience. We've already looked at the lengths studios had to go to in order to make modern fighters work, things like hit boxes and frame skipping just to keep visuals consistent and create balance. How then would SNK be able to create hundreds of frames of new animation, in high definition, for KOF XII without hiring a massive animation team or spending a decade in development? It turned out that there was use for one of the animation shortcuts after all.

Rotoscoping might have very been the way that SNK was able to push up the development window. Rotoscoping might be the reason why King of Fighters XII has such amazing models. It was my understanding that SNK learned many things about the strengths and weaknesses of 2D and 3D fighters when they tried moving all of their franchises into one format. Samuari Spirits, Fatal Fury and the King of Fighters did not lend themselves readily to 3D. Problems with low polygon models, hit detection, framerate and animation kept most of these titles from being widely regarded. However this didn't stop SNK from trying to make 3D work while also applying these lessons to their 2D titles.

Development of KOF XII probably began immediately following the KOF Maximum Impact series. The models featured on this title were the closest the studio had ever come to recreating the Falcoon style in 3-dimensions. Their stances, animations and proportions were very well done, even if the game was not well received. Rather than dump the models I contend that they were used as the template for KOF XII. These models were used in a rotoscope capacity to help speed up the animation process in KOF XII. It makes sense when you look at how smooth and proportional the new animated characters were. Either the animators were trained highly in the style of Falcoon, or he created hundreds of character sheets that animators used as reference material. If that were the case then the game would have taken much longer to produce. Instead, I believe that these 3D models were given moves and ranges consistent with 2D presentation, this way artists did not have to imagine how a character looked while flipping, spinning, dashing, or even bouncing on the souls of their feet. They could reference the moves from a fixed perspective and then "trace" over them. They would thusly achieve realistic lines, shapes, highlights and shadows while also adding details like rippling clothing and "speed blurs" which were more stylized.

KOF XII seemed fluid because all of the frame skipping and hit box issues had already been addressed before rotoscoping. Moreover the characters were proportioned very much in the style of the designer, rather than being swollen in a concession to the west. Okay, to be fair some of the characters did get beefed up significantly. Terry Bogard and the Ikari Warriors Ralf and Clark were far thicker than they've ever been, but other characters like Kyo and Ash remained lithe to help make the contrast more significant. The same thing could not be said of SF IV. Models in that game were thicker across the board. By keeping the 3D models handy, animators were able to change proportions or add new moves without fear of losing the consistency between sprites.

Creating animations in 3D and then spending time redrawing them in 2D might sound redundant, and expensive for developers, but it made sense for the market. Fighting games had to compete for attention between themselves and against titles in other genres. Not to mention that for many studios the concept of the fighter was dead. Animated sprites were more striking now than in the 90's because 3D had homogenized graphics. It turned out that some fighting games, like SF and KOF have always worked best in 2D. Their visual appeal was in preserving the artistic intent and animating that style, while being mindful of the lessons that Ralph Bakshi had taught animators. The success of the best fighting game control, balance and animation had been in part due to the limitations of 2D technology. It was the imperfections of frame skipping, hand-drawn graphics, and the programming variables of hit boxes as much as the planned details that helped the games succeed. In a similar way the hand-drawn quality was making KOF XII develop into something truly special. But none of it would have been possible without the contributions of 3D technology and rotoscoping to shorten the development cycle.

The day is coming when a team of engineers, artists and animators will create polygon technology that is indistinguishable from the hand-drawn sprite. Until that day comes the best way of preserving the art of fighting games is by going the traditional route and creating sprites by hand. I hope that the teams at Capcom and DIMPS take these things into consideration when they begin working on subsequent fighting games, whether it's Street Fighter, Vampire, the VS series or something else. They have 3D models of many of the world warriors already, perhaps they have begun letting air out of their muscles in order to prepare them for a second life as a rotoscoped figure. Stranger things have happened...

I hope you enjoyed this short series, thanks for reading and commenting. Now please excuse me. I have some end of the semester work to catch up on. Take care!