Wireframe

The lost art of handheld demakes

By Alan Wen. Posted

Traditionally, handheld gaming hardware has always been a few steps behind its home counterpart, arguably the same way home consoles used to lag behind arcade boards. The arrival of the Nintendo Switch, however, became a significant step in eroding divisions between the two. Nintendo’s hybrid is, of course, the most underpowered system of its generation, so there’s still a kind of marvel when porting houses like Panic Button and Saber Interactive deliver ‘impossible’ ports like DOOM Eternal and The Witcher 3, respectively.

A cross-platform release, in fact, is becoming applicable to both home and portable platforms, as high-end smartphones mean games like Fortnite, Genshin Impact, and eFootball (formerly PES) ship with the same features, even capable of cross-play and cross-save, as their larger counterparts. Cloud gaming, on the other hand, means any game can technically play on any screen regardless of what’s under the hood.

Platform parity is the utopia mainstream gaming has always been trending towards, and who wants to go back to playing the inferior handheld version of a blockbuster title that was meant for the big screen? In modern parlance, you could refer to handheld conversions as demakes: stripped-down versions of the original game, in some cases with the perspective shifted down from graphically intensive 3D to 2D. 

As we discovered when we spoke to several developers who cut their teeth in this era of demakes, though, it was also a thrilling creative frontier as they followed the late Gunpei Yokoi’s philosophy of applying lateral thinking with withered technology.

Advanced tricks

The Game Boy Advance was a 32-bit handheld console, but it wasn’t designed with 3D games in mind. Not that this stopped many plucky developers from making the attempt – indeed, one of the GBA’s launch titles was Tony Hawk’s Pro Skater 2, in which Vicarious Visions managed to take Neversoft’s original PSone code and transfer it to the handheld, faithfully recreating the core experience, including the skaters and levels, in isometric 3D.

Most developers would downsize a 3D platformer into a 2D side-scroller for GBA, but for Velez and Dubail, it was go XXL or go home.

“That’s why people still to this day say that it feels like it did on the PlayStation – it’s because we’re actually running the PlayStation code on the Gameboy Advance,” explains Matt Conte, lead programmer on THPS2 and part of Vicarious’ handheld division, which called itself Team Baisoku. (This was an in-joke based on the one untranslated function from Nintendo’s Game Boy Color documentation, which could set the CPU to double-speed; Vicarious Visions considered itself an underdog that had to work twice as fast as other studios.)

The Tony Hawk port was all the more miraculous, as Nintendo had told the Vicarious team that the GBA could only run C code, not C++, the language Neversoft’s code was written in. “I did all the low-level hacking to get it so that C++ would work on the platform, gave that back to Nintendo, and they published it on their developer website,” said Conte. “We were a little hungrier back in those days and weren’t afraid to take some risks. Even when Nintendo, the platform owner, was telling us we couldn’t do something, we went ahead and did it anyway.”

Matt Conte, once lead programmer at Vicarious Visions’ handheld division Team Baisoku. These days his challenges lie in VR at Oculus

Even the assets used were based on the originals, albeit scaled-down, while the environments had to be reinterpreted and rescaled to fit an isometric perspective. Perhaps the biggest challenge, however, was animating the skater. “This renderer was basically a size optimisation,” Conte explains. “A lot of Vicarious Visions games up until that point had characters rendered in 3D but as a 2D sprite. But we couldn’t do that from an isometric point of view because the skater can go in literally any direction, so the storage space would have just been almost infinite.”

Indeed, Conte estimates that THPS2 had about 5000 animation frames, while the sequel went up to 7500 – all of which had to be stored in just a few megabytes of memory. “Memory was always an issue,” he says. “On THPS3, the challenge was, if I can’t get this optimisation to work, we’d have to cut a level out of the game. Those are the sorts of hard decisions and cuts you have to deal with in a super memory-limited environment.” It’s nonetheless a testament that this GBA port was a success – one that led to handheld treatments for subsequent iterations, and it was even used to adapt Sega’s Jet Set Radio in a similar fashion.

It may lack multiplayer, skater customisation, and a level editor, but THPS2 on handheld really did retain the core Tony Hawk experience

Slimming down Sims

While handheld teams might feel like the B-tier of game development, having to make do with less powerful hardware, these smaller, often outsourced studios were also avenues for entering the games industry in a time before indies became prominent, not to mention an opportunity to work on big IPs without requiring the same big-budget resources. This included Daniel Collier, who was a producer and programmer at Exient, an Oxford-based company which had made many handheld ports for EA in both the GBA and DS era, including a port of The Sims 3 to the latter in 2010, a year after the PC version released alongside other console platforms.

Michael S. Lee was producer of all the Call of Duty DS titles at n-Space from 2007 to 2011

“Whenever you port something to a handheld, you’re going to be accused of crunching down a game,” he says, all too aware of the prevalent attitude that many handheld games are cheap lazy tie-ins that could be considered officially sanctioned knock-offs. That certainly could have been said for a few entries of The Sims on GBA, which really were just a series of minigames completely divorced from the core series. “That was one of the reasons for The Sims 3: we wanted to make a proper Sims game because we didn’t want that criticism.”

EA not only greenlit the port, but provided Exient with plenty of resources, including the largest ROM size available for the DS game cards, a year of development time (which was “a lot back then for a handheld title!”), and even the source code, though the latter was just for reference. What the team did take from the PC version of The Sims 3 were the models and animations, which then had the details crunched down so that it was manageable on the DS. “I think the artists actually did have to take the model and scale it down, removing the polygons from it to make it workable, so there was a human process,” Collier adds. “I remember writing a tool to sort of crunch down animations, which the artists would also go in to tidy up.”

The dual-screen was useful for displaying a map and equipping weapons, but also an opportunity to include minigames such as bomb-defusing

Not all features made it across, notably the absence of babies as well as The Sims’ famous woohooing – though it’s unclear whether that was down to technical limitations or because the DS version had a lower age rating – although having an outside world you could explore, new to the series, was something they replicated. “We were also going to just have a flat 2D map that you click to move to different places, but in the end, we were able to add a 3D world you could move around,” said Collier. “There were a lot of decisions to be made about what we would bring over, but generally we did try and get the essence of it brought across.”

While still a far cry from a full-fat PC experience, the Nintendo DS nonetheless marked the first recognisable iteration of Maxis’ life sim on handheld. The system’s dual-screen also proved useful for handling real estate, as well as providing the ability to drag and drop objects and buildings. An improved version was even made for the 3DS later, though this one wasn’t handled by Exient.

Daniel Collier, a former programmer and producer at Exient, an Oxford-based studio he joined in 2003

Beyond the Call of Duty

If there’s a great example of trying to make an action-packed triple-A blockbuster for a platform it was never intended for, it’s surely the Call Of Duty series. The prospect wasn’t quite as impossible as one might expect, though. After all, before Activision’s military shooter became an annual tent-pole, it even had a game on the ill-fated Nokia N-Gage, while Florida-based studio n-Space had previously developed a first-person shooter engine for the DS with a port of GoldenEye: Rogue Agent. But though you could refer to Modern Warfare on DS as a port – it even shared the same box art as the other releases – producer Michael S. Lee corrects us: “There is actually nothing ‘port’ about it,” he says. “You can’t take anything from a console and put it on a DS – it just doesn’t work.”

The handheld version doesn’t even follow the same story as the mainline version. In fact, Lee recalls there was something of a strange secretive dance the developer had to engage in with its publisher at first. “Activision basically said to us, ‘We can’t tell you where the mainline [game’s story] is going. We can tell you ‘It’s kind of going to be the Middle East’, and we can give you general things like that, but you guys have got to come up with your own storyline’,” Lee continues. The story would also need to be approved or steered in another direction in case it touched on anything that the main studio, Infinity Ward, was doing at the time.

With just enough choreographed setpieces and explosions, you’d never know that CoD was painstakingly crammed into the DS.

This resulted in n-Space opting to set its games up as ‘companion’ titles, as Lee refers to the handheld’s characters (and effectively acknowledging n-Space’s role) as the ‘B-team’. While the idea of playing as a secondary unit whose actions support the mainline scenario may also set expectations for what a DS game would achieve, the team certainly didn’t want to settle for a half-baked experience that would short-change fans of the Call of Duty brand.

There was still an onus on creating ‘wow’ moments, a dramatic sense of scale, and of course loads of the big explosions that the mainline series is known for. Due to the DS’s limitations, from only being able to fit a maximum of 2048 polygons and its meagre memory space, levels had to be painstakingly mapped out almost like box rooms – all while giving players the illusion they were roaming an open, realistic environment.

“We had to be very decisive in how we did things, to be mindful of every coded feature, from texture to audio to characters on screen, the number of AI on screen – all that takes up memory,” explains Lee. “We sat in so many meetings whiteboarding out just where memory was being spent. We had to know everything that went in that game, and we had to account for every byte just to squeeze absolutely everything in. But if we could gain just 10kB back from one feature code-wise, we could add this cool explosion to counteract that. We were sort of robbing Peter to pay Paul in the end.” 

The whole cloth demake of The Sims 3 on Nintendo DS.

These developers’ handheld adventures were more than a decade ago now – many of those involved have since moved to other fields. Lee is no longer in game development, while both Conte and Collier now specialise in VR, a relatively new frontier with its own limitations to consider. Still, they all have a sense of pride in the handheld games they made, even if history does tend to gloss over them. Describing THPS2 as the favourite project of his career, Conte says, “I think there’s always going to be a place in my heart for handheld consoles. We had so many problems, and we just ran at them head-on.”

Collier even regards the Switch as a bit “dull”, despite its obvious plus points. “The DS and GBA in their time were new and exciting handhelds,” he says, “but because people understood they weren’t anywhere near console quality, you also had a bit more leeway in reimagining games, and new ideas can come out of that.”

“What I liked about the DS was the idea of not being able to just throw a ton of money or polygons at it to solve a problem,” says Lee. And while he recognises the development landscape has changed – with huge advances in portable tech, especially in mobile – there are still new hurdles to overcome. “The challenge is different, not on how to keep it afloat like we did – it’s more of how do we make it look as real, as authentic, as graphically innovative? How do we push the boundaries?”

From Wireframe store

Subscribe