Digital Foundry

Tech Focus: What Next for Motion Control?

Wed 15 Feb 2012 9:00am GMT / 4:00am EST / 1:00am PST
Digital Foundry

Digital Foundry on next-gen options for Move, Kinect and Wii U

The recent release of Kinect on PC demonstrates that Microsoft remains firmly committed to its "hands free" controller, and actually features a key hardware upgrade over the Xbox 360 - the ability to work with players at close range: essential for a desktop environment. It's the first step in what is believed to be an evolution for the Kinect platform, which will see its true "sequel" released alongside, or even as part of, the next-generation Xbox. Whether we like it or not - and a great many core gamers clearly don't - motion control is here to stay.

"Nintendo's Wii was the most disruptive product of the current console generation - to what extent will motion control define the next?"

As the Wii gradually winds down and its next-gen successor is readied for launch, it's important to step back and take a look at the diminutive console's achievements. Described by many rather disparagingly as an overclocked GameCube (although essentially that is what it is), Wii remains the most disruptive video game product of the current console generation. While Sony and Microsoft embarked on a hugely expensive technological arms race, Nintendo merely refined its existing, modestly successful hardware with minor architectural changes and a big boost in clock speed. The real genius of its proposition was, of course, the controller.

As joypads became ever more involved and complex, Nintendo pared its own controller back, reducing the principle form of interface to a remote - and everyone knows how to use a remote, right? The add-on nunchuck controller added the requisite analogue controller required to make core games feasible and the rest, as they say, is history. Wii succeeded because it was immediate, original, enjoyable - and everyone could understand it and join in. Pack-in title Wii Sports fails as a comprehensive game with lots of depth, but as a sampler for motion control, it is a work of genius.

Nintendo's E3 presentation for Wii U demonstrated once more that the controller is the focus, but its established Wiimote isn't the primary interface.

Were it not for the Wii, doubtless there would be no Kinect, and almost certainly no PlayStation Move, but the problem with these products is that they are add-ons to existing consoles that were never designed with motion control in mind, and the range of games that support each of them is limited. Obviously, in contrast, the vast majority of Wii titles (the likes of Wii Fit excepted) are built with the motion controller in mind.

Sony has faced the bigger problem here, despite having arguably the premier product in the field and the most talented internal R&D team who were literally years ahead of their time. To my mind, PlayStation Move is the most flexible, versatile and potentially exciting motion controller of the lot. It does everything the Wiimote is capable of and a whole lot more, and its level of precision is unrivalled. The genius of its design is all down to the SCEA group led by Sony's Doctor Richard Marks - who was experimenting with motion control over a decade ago, and actually exhibited a wand-style controller for PlayStation 2 at ECTS way back in 2001. Marks was even demonstrating Kinect-style cameras in 2004 during a presentation for Stanford University students, over a year before the Xbox 360 even launched.

Unfortunately, as good as Move is, it doesn't define the PlayStation experience in the way that the Wiimote did for Nintendo: its many qualities are never the primary focus for games developers, and while there have been some great implementations in first party titles, Sony never deployed its best development teams on Move-centric products. The approach is somewhat at odds with Microsoft's Kinect launch, consisting of almost exclusive titles and the positioning of Kinect as a new platform.

More Wii U hardware demos - here we see the ways in which Nintendo believes that its existing motion controllers can actually work in concert with the Wii U tablet in multiplayer gameplay.

But Kinect has problems of its own - perhaps not from a commercial perspective where it is generally acknowledged to be the most successful launch in the history of consumer electronics - but certainly in terms of its technical make-up. It performs brilliantly on a set range of functions which make it perfect for dancing and fitness games, but it is dogged by latency and tracking issues, and its reliance on a relatively large play space (something addressed in the PC update) limits where and how it can be played. Microsoft's attempts to justify the device to core gamers also appear to have fallen short: the Kinect support in Forza 4 isn't so impressive, the Ghost Recon: Future Soldier demo at E3 just looked outright bizarre and far more complex than using a standard controller, while its utilisation for voice support seems somewhat superfluous when headsets (with mics) are integrated into every Xbox 360 package.

So, going forward, where does motion control sit within the platform holders' next-gen plans? Nintendo, of course, has already shown its hand. For its debut HD outing, the controller is still the focus, but the company has combined virtually all major forms of input into its tablet-style offering: there's motion control, touch-screen and conventional joypad style controllers too. Wii U is backwards compatible with Wii MotionPlus and other peripherals such as the balance board, but it's that tablet that's the focus - and what will make or break the success of the product.

The challenge facing Nintendo is that the tablet interface is somewhat more complex and not nearly as immediate a proposition as the Wiimote was in its day. But it is clearly a highly versatile proposition and it's up to the company and third party publishers to transform some of the various Wii U concept work we saw demonstrated at last year's E3 into exciting, fresh and original gameplay.

"Kinect is a work-in-progress; marketed as a standalone platform and one that will expand and evolve alongside the Xbox brand."

While motion control strategy is still up in the air with Microsoft's next-gen hardware, details and rumours are starting to emerge. Unsubstantiated stories emerged last week that the console itself will fit into a tablet-style controller, but this is entirely at odds with basic logic - cramming next-gen silicon into a tablet aimed at a price-conscious mainstream audience simply isn't a viable proposition, for all the same reasons (and then some) that those wishing for a true next-gen leap for Wii U's graphical capabilities are likely to be disappointed.

Worthy of far more consideration are the reports that Kinect will ship as standard as part of the next-gen Xbox, perhaps in two SKUs: the first, a set-top box, the second a fully fledged core-gamer orientated console. These rumours carry considerably more weight bearing in mind that Microsoft apparently held a consultation with developers on the form and function of Kinect 2 at an event hosted at Disneyland in Los Angeles after E3 last year, and appears almost obvious bearing in mind that the platform holder has worked hard in establishing Kinect as a standalone platform, one that will expand and evolve alongside the Xbox brand.

In this video from a 2004 presentation, we see Sony's Richard Marks demonstrating Wiimote style controls... for the PS2. Marks actually had motion control demos on display at ECTS 2001 - that's how far ahead of the curve Sony's research teams have been. But will this kind of forward thinking translate into an actual successor to the classic dual shock?

In terms of how Microsoft can evolve the hardware, it doesn't take a genius to evaluate the current system's limitations and suggest some alternatives in overcoming them. Latency is one of the major issues that holds back the kinds of games in which Kinect can be utilised.

Developers have made some strides in overcoming this issue, firstly by devising their own ways of utilising the z-cam's depth data, specifically when it comes to re-mapping it onto in-game characters. It's generally believed that Microsoft's own libraries only re-map the human form scanned by Kinect onto its own Avatars using a computationally expensive skeletal model, and developers have adopted their own tech to do the same job more quickly, at the expense of the kind of flexibility Microsoft's libraries offer. But the real solution here will be hardware-based, and that means binning off the USB interface for something faster and with more bandwidth on tap. Resolving this issue also addresses another limitation - the ability to stream real-time video with Kinect data simultaneously - something that the PC version has no problem with owing to a superior USB 2.0 spec implementation.

The second issue is one of precision. In the past couple of years, various claims have been made with regards the detail the Kinect cameras are able to resolve, in particular when it comes to the hands and fingers. Kinect titles have concentrated mostly on major body movement, and precise movements with the hands and fingers have not been possible to track effectively - hence the overly strange Ghost Recon performance we saw at the Microsoft E3 presser last year. By the time the next-gen comes around, upgrading to higher precision cameras, perhaps even with a 60Hz refresh, should be possible.

Kinect for Windows may look very much like the Xbox 360 equivalent, but it contains upgraded optics with the ability to work at close range - just one of the enhancements we would expect to see in the true 'sequel' planned for the next-gen Xbox.

Perhaps the most exciting element behind a bundled Kinect is the fact that developers can address the hardware knowing that everyone owns it - utilisation of functionality should hopefully be a little more fully featured than some of the implementations we've seen in Xbox 360 core titles, and hybrid control systems using both pad and camera could be intriguing. A hefty reliance on Kinect tech would make streaming gameplay via Cloud a lot more difficult though, and it's widely believed that Microsoft is investigating online streaming - beaming inputs (even motion controllers) over the internet would be simple enough, but the upstream bandwidth just isn't there to sustain two camera feeds and the audio from a multi-array mix.

So if Nintendo has revealed its hand and we can make a lot of educated guesses about where Microsoft is headed, where does that leave Sony? Of all the platform holders, very little indeed is known about the successor to PlayStation 3, least of all the technical make-up of its control system. But of the three major console-makers, historically it has been the most conservative, effectively sitting upon the innovations brought forth by Dr Richard Marks' team, iterations of which have gone on to perform so well for Sony's competitors. Will we once again see the standard dual shock bundled with its next console, a reworked PlayStation Move, or something all-new from Marks' labs? I can't wait to find out...