The location: Stanford University. Sony's R&D mastermind Dr Richard Marks is once again showcasing motion control technology with a variety of ingenious technical demos very much along the lines of those recently seen by Digital Foundry. The difference this time is that the date is January 21, 2004 and the host console for the impressive tech is... PlayStation 2.
Marks has been waiting a long time to bring depth-sensitive "3D" motion control to the mass market. Even longer than you might think. His original camera-driven demos for the PS2 have been carbon-dated to before the turn of the century, with his original swords-and-sorcery style demo actually receiving a public outing at the summer ECTS Show at the Islington Design Centre in 2000.
There's even coverage of his work in the launch issue of Future Publishing's Official PlayStation 2 Magazine. Rare's Nick Burton recently talked about their prototype Kinect demo called "seagull" where you flapped your arms and flew around the environment. In that self-same issue of the Official PS2 Mag, you can see pictures of a very similar demo put together by Marks' team using the prototype camera that would one day become EyeToy.
The Stanford University lecture in its 75-minute entirety can be streamed directly, and while the final PlayStation Move hardware is immeasurably more precise, accurate and user-friendly, the precursor to the new motion controller available to PS3 gamers later this month clearly has much in common with the final design.
Back then though, there was no actual motion controller as such - just a series of props, including a long wand-like object with a familiar-looking sphere on the top. The camera and the PS2 does most of the hard work, but curiously there are many echoes from today's Move. For a start, calibration is required. Back in the day this was achieved by taking the sphere and filling an on-screen circle displayed by the PS2. Voila: now the device can track the inanimate object in 3D space. While the implementation is undeniably primitive compared to the final rendition in PlayStation Move, crucially it works rather well.
Augmented reality (or "enhanced reality" as it was known a gaming generation ago) is also covered off with the same style of super-imposed 3D objects overlaid onto the webcam video. By the time of Marks' 2004 presentation, he was already experimenting with 60FPS video, satisfied that the doubling of temporal resolution helps to make the device far more precise and responsive. Clearly, many of the thought processes behind PlayStation Move have been a long time brewing.
Nintendo may well have been first to market with the mainstream-friendly Wii, but the Sony R&D team run by Marks has been consistently ahead of the curve. The combination of motion controller and camera was the subject of a patent that dates to the time of the Stanford talk. And there's more.
In the self-same 2004 presentation (around the 58-minute mark), you can also see Marks' direct experimentation with what is then known as the "z-cam" from Israeli company 3DV: a depth-sensitive 3D camera that can map the human body. Over five years before Microsoft announced Project Natal (latterly Kinect) and acquired 3DV's tech, Sony R&D team was already evaluating the sensor, and Marks is clearly fully aware of the potential for gaming.
The reality of the situation must be rather sobering for Sony. It could have owned motion control gaming, but instead Nintendo embraced the concept and left its rivals for dust.
Knowing how long this journey has taken, the functionality, refinement and polish inherent in the final PlayStation Move hardware becomes more understandable. Watch Marks' more recent demos, or better yet, buy Move and a copy of Sport Champions, and hopefully you'll get some idea of why we think that from a hardware perspective, Move is the pick of the bunch of the motion controllers from any of the major platform holders.
However, getting the most out of what Move offers developers is a real challenge, and it's one that we can only hope that Sony and the third party publishers fully embrace. Aside from its precision, where Move can really make a difference is in the way it performs as a true 3D controller. This in itself opens up a colossal array of new game concepts. Let's take a look at a couple of Marks' more recent tech demos in full direct feed glory.
First up, the manipulation of objects in 3D space. You can reach out, grab things and move them around in 3D space. While you might think that Wii MotionPlus would be capable of the same sort of thing, in actuality the applications are more limited. While the motion sensors within Move may well operate with a similar ballpark precision to MotionPlus, the Wii peripheral has no camera lock and the data is prone to drifting. Cover up the glowing sphere on Move and the same thing happens. Motion sensors on their own are not accurate enough.
That being the case, it's fair to say that only Sony's motion controller is consistently capable of achieving stuff like this:
Dual Move control is producing the closest thing yet seen in the games arena to the famous Minority Report user interface. While Kinect may well be the real deal in terms of controller-free interaction, its somewhat infamous inability to track hands and fingers at standard range makes the dream of "that" interface impossible to replicate in real life. Just the process of pressing a button in the games seen thus far is somewhat lengthy and can even be annoying.