• Hey Guest. Check out your NeoGAF Wrapped 2025 results here!

Engadget Kinect hands-on shows Kinect latency and "Reflex" demo

kinect_latencybakx2.gif

It would be cool to dodge some jokers tomahawk in CoD by leaning.
 
That lag is the reason why I say that the Touchpad/PS4-Eye & even the Sixaxis is going to be a lot better for secondary inputs than the Kinect Camera.

they should have pushed for a 60FPS Kinect.
 
Carmack knows what's up.

He said it's still better than an analog stick though, which made me grin :p

OT: that looks awful, clear delay (and the demo looks awful, who wants to lift their controller up like some kind of idiot for something a button press could do faster and better, or duck and weave about)

Still looks like about 250+ ms of lag, even if you assume the lcd they're showing it on is probably a piece of shit with 50ms of input lag, it's still useless.

After how much MS has been praising Kinect 2 tech and being adamant that its part of X1 to not sell a Kinect-less SKU, I was expecting much better performance. And these are tech demos not even real games that will have to process so much more. Actually PS playroom had better performance from the videos shown.
 
How the heck do you play the first game in that video if you constantly have to take your hand off the controller to tap your head? That looks like a horrible implementation.

You would have to do the same thing in real life with a rifle and night vision goggles.
 
After how much MS has been praising Kinect 2 tech and being adamant that its part of X1 to not sell a Kinect-less SKU, I was expecting much better performance. And these are tech demos not even real games that will have to process so much more. Actually PS playroom had better performance from the videos shown.

To be fair, overlaying a video feed with your games interface masks actual lag.
 
It's odd, with how advanced Kinect 2 is over the plain stereo camera's PS4 is using... but Sony managed to get a better Tech demo (play room).

I guess I personally give different degrees of credit for a "trailer-style" promotional video compared to a outsider hands-on preview video and write up, but in terms of "marketing" the feature, I totally know what you mean. I feel like word-of-mouth is actually going to have a strong influence on how each of those devices gets received during the early release window. I really liked that Sony video.
 
I prefer Kinect 1, it was more advanced and could tell what you wanted to do before even you did. #Downgrade

ba0c7231ee96ba55740e80czb6.gif

I'm amazed at how popular this gif still is :P

I worked on this "level", which, like a lot of stage demos and CGI trailers, was never actually meant to be playable, hence the actor "playing" the game.
 
https://www.youtube.com/watch?v=01y1vXBl1-8&t=1m18s
Look at how much time it takes from when he touches his head to when the Xray vision turns on.

Anyone saying that the lag is bad or "1 second delay" is just lying. This is really good and its only a simple demo. I have no doubt that devs will be able to do some really amazing shit with Kinect over the next 12-18 months.
Yeah that gif that's been quoted a billion times now is clearly the worst example in that whole video.
 
Look pretty good actually but that's not how I like to play my video games. I can see this being useful for my 8-year-old cousin though.

They definitely have improved the Kinect a lot, there's no denying that.
 
The lag is definitely noticeable but the more worrying thing is the amount of lean required to actually perform the move.

That is certainly configurable. Since the lean is either "on" or "off", there needs to be a point to decide if the user really did want to lean. They must have decided on a lean toggle because a person moves a little bit constantly and the screen would probably be swaying back and forth too often.
 
Can someone explain the difference in tech between the PS Move and Kinect and why the Move seems so much better with regards to latency?

I was seriously hoping that the Kinect would have been better for v2, but it still looks like complete crap for anything that would require any sort of precision.
 
Looked pretty laggy to me.

If that's how the final product will be then there won't be much immediacy in the gameplay.
 
https://www.youtube.com/watch?v=01y1vXBl1-8&t=1m18s
Look at how much time it takes from when he touches his head to when the Xray vision turns on.

Anyone saying that the lag is bad or "1 second delay" is just lying. This is really good and its only a simple demo. I have no doubt that devs will be able to do some really amazing shit with Kinect over the next 12-18 months.
Is that directed at moi?

I count 20 full frames when I drag that GIF into premiere. Not quite a second but well into "laggy" territory.

But that's not even the while story. The game or demo needs to recognize intent on the players behalf. So there will always be another layer of timing overhead.

Basically it just doesn't look fun or immersive. Look at how much work the Rift people are doing to kill that latency.

RiccochetJ said:
Can someone explain the difference in tech between the PS Move and Kinect and why the Move seems so much better with regards to latency?
My understanding is: Move camera tracks the glow ball @ 60Hz, and the wand sensors run at a much higher refresh even than that for inertia and direction (few hundred times per second), and these two inputs are combined for error correction. The way Kinect is going about it is much much more difficult and technically demanding.
 
You are probably posting that image to show latency. But that is a gesture that he is doing so it is waiting to make sure that a lean is desired before doing it. As you can see during the rest of the video, just general movement of the upper body is not translated on screen. There is either an "off" or "on" to the lean.

I actually hadn't thought of it that way. Makes sense that if it tracked your upper body one-to-one, the potential for too much swaying would definitely be more impactful than the slight delay in confirming a lean.
 
Kinect 2 has potential. I think Microsoft is very brave in betting on it and it may pay off for them.

Now the challenge is turning cool tech into fun games.
 

Keep digging. The spy footage isn't doing it for me.

Someone gif the part where he turns on the vision, that seemed almost instant. Is it possible the Kinect 2 starts the on-screen action when he reaches the full lean in that gif?

The lean animation is not going to start until the player finishes the motion. It's not a case of as soon as you start moving the character leans; otherwise with any slight motion the camera would jumping left and right constantly.


This, basically.
 
Can someone explain the difference in tech between the PS Move and Kinect and why the Move seems so much better with regards to latency?

I was seriously hoping that the Kinect would have been better for v2, but it still looks like complete crap for anything that would require any sort of precision.

Kinect is doing joint tracking and all manner of random things, PS Move is tracking balls.

Edit:

Seriously cannot wait for devs to play around.

DoubleFine get on it!
 
The lean animation is not going to start until the player finishes the motion. It's not a case of as soon as you start moving the character leans; otherwise with any slight motion the camera would jumping left and right constantly.
 
375ms?

I'm so tired right now so just assume I'm wrong.

That would be correct, but I'm curious what part he's counting those frames on (the lean bit, or skeleton tracking bit or...?).

I think a count on the skeleton tracking would be of interest e.g. No buffering for gesture recognition.
 
When I came into this thread, I was judging things solely off the posted.gif. But, watching the actual video, I'm not sure what the demonstrated problem is? It seemed fine. The tap head for xray wasn't instantaneous, but it doesn't seem like it was meant to be. The leaning was quicker to lean the screen in some instances versus others, but that seemed reasonable as well.

I haven't really read any impressions of Kinect Sports Rivals (though I'm sure there's a thead for it) so, I'd curious as to how this lag impacts actual real game scenarios.
 
The video is 23.976 fps, 9 frames of lag. Someone do the math.

Taken from displaylag.com:

A game running at 60FPS tends to have a minimum of 3-4 frames of controller lag (66.7ms) regardless of the display being used. Therefore, a game running at 30 FPS results in a fixed controller lag of at least 6 frames (100ms+).

Haven't Microsoft said Kinect now runs at 66ms (down from 90+)? So is most of the lag visible in this video from the display?
 
The video is 23.976 fps, 9 frames of lag. Someone do the math.

Worst case (motion just after frame x, motion is visible on frame x + 1, game updates just before frame x + 10, game is visible on frame x + 10): roughly 420 milliseconds.

Best case (motion just before frame x, motion is visible on frame x, game updates just after frame x + 8, game is visible on frame x + 9): roughly 375 milliseconds.

Completely disregarding fluctuations depending on the game. All "frame" mentions are those of the video rather than those of the game or camera, although those should also matter for obvious reasons, making the best and worst cases a lot different depending on those frame rates.

(Correct my math if wrong.)
 
That would be correct, but I'm curious what part he's counting those frames on (the lean bit, or skeleton tracking bit or...?).

I think a count on the skeleton tracking would be of interest e.g. No buffering for gesture recognition.

I was stepping through at the skeleton part. Definitely around 9 frames.
 
Top Bottom