Home
Was Microsoft's motion technology demonstration really all just smoke and mirrors? André highlights where the company may have overpromised.
During E3 2009, Microsoft released an incredibly well-received, promotional buzz trailer for the then-titled Project Natal that would later become the Kinect motion-detection system. The trailer debuted to incredible interest and applause at the amazing feats depicted in the video (and that’s not sarcasm -- people assumed this would be the next level of gaming and were thrilled to see it taken a step higher).
Project Natal boasted an honorable assortment of features and functionality compared to its predecessor, the Xbox Live Vision Camera, which featured picture quality on par with the Game Boy Camera. Seriously, Daguerreotypes had better clarity than that thing.
Finally unveiled as Kinect the following year, we all simultaneously wondered, “What the hell just happened here?” We just saw a kid transmute his skateboard into a video-game object and now a year later we’re in some kind of freaking circus act? All we get to see was a white guy dancing like a white guy and a young Asian girl giving a handy to a tiger? What the hell went wrong?!
Under U.S. law, I'm not even supposed to be showing you this.
One year (give or take a few months) after the official release of the Kinect, we are nowhere near coming to the grand acts displayed and showcased in the Project Natal trailer. Here’s what the original preview trailer promised us, and exactly what they failed to deliver on.
Project Natal can't...
1) ...hold a conversation with you.
What you see in the Natal trailer: Ian, a scene kid with a prominent hair helmet, approaches a nondescript martial-arts fighting game wherein a generic kung-fu guru taunts the gamer with, “You’ve come back for more?” Ian responds, “Let’s do this!” And the game subsequently laughs at him and responds with a mocking tone. Ian then kicks Master Wong straight in the dong.
"I'll break your hip, old man!"
The Kinect reality: At first glance, this just seems like a generic transaction of comments (assuming you’re speaking to somebody in real life), but Ian is speaking directly to a character in a video game who can apparently acknowledge and respond to answers more complex than a simple “yes” or “no” (and we’ve been screaming “no, no, no” at bad games for decades now). The game scoffs at Ian’s rebuttal to its challenge and therefore actually shows some level of thought depending on the answer given.
The pseudo-Japanese game shown in the trailer is also exactly one Hot Coffee mod away from being an ESRB nightmare.
No Xbox title presently available offers this level of linguistic understanding. Yoostar 2, a game where you reenact scenes from popular movies, comes closest by judging your performance based on a provided script, but the quality of judgment is so poor that it’s hilarious, especially if you’re playing this in a party setting and you’re surrounded by ambient noise from other guests. You could hire the original actors themselves to repeat their lines, and the game will still tell you that you have the acting talent of Howie Mandel.
2. ...detect the kinds of movements shown.
What you see in the Natal trailer: A family is sitting down to play a formula-one racing game, where the daughter mimics the motions of driving and her family follows suit by leaning side to side with the turns the car makes. Her father stands up and replaces a tire on the car during a pit stop. In a second hypothetical title, a kid is seen playing a Godzilla-type, city-destruction game and swings his arms around to wreck buildings; he then makes a roaring sound and assumes a pose while the monster breathes fire.
Natal trivia: Three of the four people in this picture are contemplating suicide.
The Kinect reality: Kinect can detect motion. I’m not going to say it can’t because that’s the entire focus of the peripheral, but I will say that while it does indeed detect movement, it definitely doesn’t do it at the caliber required to sense rapid “Z-axis” movement akin to replacing lug nuts on a tire or to see when you’re making a goofy roaring face to tell Bizarro Godzilla to give macrophiles a shameful and awkward erection.
If you want the Kinect to see you do something, you have to move in very pronounced motions because it seems only large movements can be seen by the stereoscopic cameras on the sensor. This works great for exercise games where you have to run in place or otherwise be full body active, but when all you’re doing is making a pushing motion, the Kinect can’t see this and doesn’t do anything. And if it does actually do something, then its detection is on par with the Nintendo Wii in terms of the likelihood that you’ll suffer first-world-country ragequit and kick in your plasma screen T.V.
Additionally, the Kinect doesn’t like when there’s a bunch of movement taking place all at once. The entire family moving side to side while the daughter drives? Doing that will cause your Kinect to overheat and explode. No, seriously, it’s just too much input and it will simply not detect your hands or the steering motion. It’ll be like driving a Toyota.
