2D or 3D, which format is the preferred format for a fighting game? Well it all depends on the origin of the title. Games that began their legacy in 2D still play best in that format. Some began in 3D and have always worked best in that format. Games that began in 2D and tried to make the switch with 3D never really caught on with gamers. Things such as animation, control and character models did not make the transition smoothly. Street Fighter was not the first to experiment with the cost-effective switch. We could still learn a lot by studying how these games were put together and what they taught the studios.

Characters from 2D games looked awkward when presented in polygon form. Their polygons were large and blocky, textures dull, animation stiff and effects limited. Part of the reason for this was because studios like SNK and Capcom did not invest heavily in 3D fighting game R&D during the early 90's. By the time they decided to make the transition in the late 90's they had already been passed up by rival studios. Only those that had a strong legacy in 3D technology, like Sega and Namco, really possessed the tools, staff and experience to make memorable characters and retain a visual impact in 3D.

However this was not the end of the 2D fighter, nor did it mean that 2D games could not eventually make the transition. Street Fighter IV was one of the new format, dubbed a "2.5D" fighting game. That was, a game rendered with 3D technology but whose entire control scheme and presentation was meant for a fixed 2D perspective. It preserved the control aspects that were partly the reason for the genre's success. It was not the first in the 2.5D format but certainly did learn heaps from previous efforts. Namely the problems and solutions that other studios experimented with while first trying to move the 2D experience into 3D.

I hold Street Fighter EX in very high regards.However when it first came out I was not impressed with it. It looked so crude compared to Virtua Fighter III, moved sluggish and didn't play nearly as smooth as Street Fighter Alpha 2, which were also in the arcade at the time of its release. Despite its flaws and limited features I grew fond of it. The characters and canon created for the series were fresh and complimented the SF universe. In the long run I had more fun playing EX than IV. So what were its flaws and what were its saving graces?

Part of the reason the new characters worked, especially when presented along with the classics, were because ARIKA had the forethought of making their 2D mechanics work in 3D. They did not want to create another Tekken or Virtua Fighter with the SF cast, but rather make the mechanics as close to 2D while rendering the world in 3 dimensions. They didn't force the entire game to stay on a flat plane but would turn around the z-axis during grabs and other special moves. They did these things sparingly, just enough to show audiences that they were mindful of breaking third dimension but respectful of the best way SF was meant to be played. They also had a number of unique visual cues that, often seem overlooked in most 2.5D games. Characters in sprite form are mirror images whether they are presented on the left or right side of the screen. That is, Ken and Ryu always have their front exposed to audiences in every version of SF, save for the EX series. Instead when they moved to the opposite side of the opponent, the character would turn and show players their backs. The same thing applied to other characters. Although their striking ranges remained consistent regardless of which side they were on, like classic SF games, this detail also showed audiences that the models and game were true 3D.

One of the biggest detractors to the game were the concessions that ARIKA made in order to make the gameplay consistent with sprite-based titles. There is an inherent problem with trying to recreate the classic SF formula while using 3D models. The ranges for strikes, counters and combos were thrown off in 3D. To most players they felt off anyhow. The opposite turned out to be true. 3D hit detection boxes were very accurate and strikes only registered when there was actual contact. In order to make EX feel more like a sprite-based game many moves were modified to create the feeling of 2D. These things included having fists and feet swell to enormous, cartoonish, proportions during dash moves, and other similar silly visuals. They did make the game play more like SF and less like Virtua Fighter, but the cartoonish animations broke the consistency and tone they had worked so hard to establish. ARIKA was locked into a possible solution for the 2D gameplay / 3D graphics dilemma but one which didn't sit well with gamers.

A "better" solution wouldn't present itself until more than a decade had passed. The editors at EGM previewed an early build of Street Fighter IV. They made mention of the buff characters but also noticed that the game did not seem to play quite like SF. The problem of using 3D hit boxes had again become an issue. Yoshinori Ono and the team had also noticed this problem and were looking for a solution, promising to have something in time for the public debut of the game. As it turned out the "fix" was 20-years-old. While indeed the game did feature a powerful graphics engine running the visuals, Capcom and DIMPS decided to use 2D hit boxes in order to recreate the ranges. This solution was nothing short of inspired.

However it was not simply enough to use 2D hit boxes. A number of concessions had to be made to the graphics and animation in order to keep the gameplay consistent. These were things that DJames Goddard had been talking about in his blog. The engine could create frames for every move in real time, running at a full 60 frames per second, however 3D models appeared to move strangely when the game was up at speed. It turned out that the problems in animation reflected 3D hit boxes, they were just too accurate for their own good. All of those extra frames of animation were distracting and didn't flow smoothly. Sort of like watching a high framerate film, anything above television's old standard of 29 frames a second gave audiences a weird sensation. These characters moved quickly yet remained in perfect focus. The animators began skipping frames on purpose in order to make the characters move in a way we had been accustomed to. In order to create a sensation of high speed movement (rather than high frame movement) the developers created blur filters to mask the sharp focus. Similar motion blur techniques had been employed by Sega since Virtua Fighter 2 in 1994.

The question that Capcom has to ask themselves is whether 3D or 2.5D are the future of the franchise. The question for gamers is whether or not they liked the bulky characters and artistic direction overall. Or would they preferred to have seen something more traditional? Would it have been possible to preserve the artistic vision while also moving the franchise forward and making it cost-effective in the current industry? I would say yes, absolutely, but the way of finding that balance had not always been so obvious. In the next blog we'll see how Capcom's perennial rival, SNK, gave the industry an entirely new way of approaching the genre. They found a way to bridge the gap between traditional animation style and 3D while making the development cycle competitive. Find out more in the next blog.