Microsoft's Project Natal (360 attachment thingy)

[The Deco]

Up till now I watched to the pistol part.

Sony is definitley THE empire, although the economical crisis will probably hold everything back, on all companies.

Anywho, I thought about it and when it reached the pistol on the Sony video they should give you some kind of pistol attachment to the controller so it will feel like you are actually holding a gun. The question is though, how does he move forward? It could be cool if there was a way to move forward when u walk but physically u dont really move, just doing the walk action, I think that MUST be done with a treadmill though in order to succeed.

They should really give some kind of attachments to the controller, it looks pretty good for a start production, im surprised that this technology exists, I thought about for 10 years at least for some quality **** sensor in an affordable price(so it won't cost like the console itself)

I see him moving his all body around the area in order to move, but if he needs to go forward, how will he do it without clashing to the TV?

 

[Kasey]

The Sony thing didn't impress me at all. I'm still skeptical about whether or not the Microsoft Project Natal will work as well as it says but the Sony one reminded me a whole lot of this thing I have called a Wii. Has anyone ever heard of Wii before? It looks like its an Eyetoy wii thing and if I remember correctly the Eyetoy was hilarious. It reminds me of one of those games that you can buy and plug right in to the tv without a console. Plus I saw something a while back about Nintendo making the Wii with better graphics or something. But anyways SONY=FAIL. The only thing keeping Sony alive imo is that they still release PS2 games.

edit: [COUGH][ame="http://www.youtube.com/watch?v=pgZVQN7SuqQ"]Wii Motion Plus[/ame][/COUGH]

 

[Akation]

I found the whole Milo situation a complete gimmick. I just found it odd that entertainment can be gained from interacting with a virtual person behind a television screen.

At E3, the dialogue looked scripted, there's no way you can throw out any random comment and have the AI interact as expected.

 

[J-Dude]

I don't trust this stuff at all, especially if the supposed state of the art camera only uses ONE camera. If it were two cameras placed far apart from each other for parallax and depth perception, then MAYBE, but I've toyed around with NON-realtime motion and camera tracking, and there are a MILLION ways that the program can interpret motion wrong or see objects wrong because of the mentioned lack of ability to judge depth. If you are against a blank background that isn't a similar color to a part of your apparent body, then it might work, but even then, every time you start you're going to have to calibrate the program so that it can track you properly, so it knows that your feet are your feet, and no your elbows.

And then there's the inevitable intersections that you're going to do by putting your hands or arms in front of your body, where it they might blend with your body and make the motion tracker lose it's targets and need to recalibrate, or more likely, try and say that another part of your body is the one it no longer sees, and screw up that way.

And as mentioned, I find it really really unlikely that something like a skateboard will be trackable, because skateboarding REQUIRES that your feet stand in front of each other, and the degree to where it could misjudge a motion is going to be so frustrating that you'll wish you HAD a controller with which to throw in frustration, and then instead take it out on the neighbor's dog.

I might believe this if there were two, or better, three cameras positioned around you which together could generate a rough 3D model of you which constantly updated fast enough to give an accurate picture. And better still, there could be little remote trackers you can attach to your arms and feet to help it's limb recognition, but unless I see these things, or am proven otherwise, lack of parallax dooms this thing to being buggy, not half as capable as it's supposed to be, and ultimately NEVER used in any serious games.

 

[harSens]

I don't trust this stuff at all, especially if the supposed state of the art camera only uses ONE camera. If it were two cameras placed far apart from each other for parallax and depth perception, then MAYBE, but I've toyed around with NON-realtime motion and camera tracking, and there are a MILLION ways that the program can interpret motion wrong or see objects wrong because of the mentioned lack of ability to judge depth. If you are against a blank background that isn't a similar color to a part of your apparent body, then it might work, but even then, every time you start you're going to have to calibrate the program so that it can track you properly, so it knows that your feet are your feet, and no your elbows.

And then there's the inevitable intersections that you're going to do by putting your hands or arms in front of your body, where it they might blend with your body and make the motion tracker lose it's targets and need to recalibrate, or more likely, try and say that another part of your body is the one it no longer sees, and screw up that way.

And as mentioned, I find it really really unlikely that something like a skateboard will be trackable, because skateboarding REQUIRES that your feet stand in front of each other, and the degree to where it could misjudge a motion is going to be so frustrating that you'll wish you HAD a controller with which to throw in frustration, and then instead take it out on the neighbor's dog.

I might believe this if there were two, or better, three cameras positioned around you which together could generate a rough 3D model of you which constantly updated fast enough to give an accurate picture. And better still, there could be little remote trackers you can attach to your arms and feet to help it's limb recognition, but unless I see these things, or am proven otherwise, lack of parallax dooms this thing to being buggy, not half as capable as it's supposed to be, and ultimately NEVER used in any serious games.

They use a time-of-flight camera from some company Microsoft toke over a while back (http://en.wikipedia.org/wiki/Time-of-flight_camera). It measures the time it takes for the light to come back to the camera (or rather, the phase difference between the light coming back at the camera with the modulated light send out). This way you get reasonably accurate depth information.

I wonder if they can pull of the stuff they show in the scripted demo. It features both perfect speech recognition and good mocap from a relatively simple camera, which is hard to get right even in lab conditions.

 

[The Deco]

Hmm, still, maybe if they used 2 cameras they could get faster response time.

Saw it on Jimmy Fallon, there is a delay...

I don't like it you don't have controllers though, like Sony's motion controller, for me the feeling of something physical in my hand is important, and it will also improve my accuracy in case of bow and arrow and such

 

[Suh Dude]

As for this Natal. I'm not buying into the hype at all. I doubt Microsoft can make this "vision" affordable, or possible for that matter, though it would be cool if it did.

Considering it's basically a sensor bar and a "controller" built in, the price will be around $50-$75 approx. If any higher I won't even consider buying it. I feel ripped off buying the wireless adapter already.

 

[L]

No offense intended Kasey, but I believe you should stop being a butthurt tool.

Lith I think you should consider it if there are going to be any good games or uses for it, I know I will. To put it into perspective for you, people paid $500 for a PHONE, turned around and said that a monster home multimedia system from Sony for the same price was too much.