Old Discussion Thread for all 3 motion controllers

[patsu]

Is Rob Dyer a new hire ?

Anyway, Industry Gamers has an interview with him, Senior PR VP to talk about Arc:
http://www.industrygamers.com/news/ps3s-motion-control-approach-speak-softly-and-carry-a-big-stick/

Read it at your own risk.

 

[(((interference)))]

Some news about natal. Basically it's a bit unfair to the tech as lag seems is easier to notice than when using a pad. If the figures in this news are accurate Ms has already made a good (80ms to 120ms). In the end no matter the resources Natal takes CPU wise the most limiting factor for the take could end human perception this may push Natal toward 60fps and a minimal number of frames between input and output.

http://multiplayerblog.mtv.com/2010/02/22/project-natal-timing-the-delay/

If the average lag is 100ms, that's excellent as that's the same as the theoretical minimum controller latency at 30FPS.

If theyre getting 80ms at times, it should mean the software is running faster than 30fps,

 

[LightHeaven]

check the video ~30secs in, another (BAM, there it is!) moment where the avatar is all scrunched up.
This is a software issue but not *that* difficult to solve, so a bit strange its still occuring

This site is claiming that the new demonstrations were made with the same hardware from e3 (but probably more advanced software), but that Ms already has final prototypes that are more precise.

I'm not sure if its bullsh*t or not, but even a october release leaves Ms with almost the same time they had to improve the tech from E3 09 till now, so maybe that's still a WIP...

 

[Crossbar]

Is Rob Dyer a new hire ?

Anyway, Industry Gamers has an interview with him, Senior PR VP to talk about Arc:
http://www.industrygamers.com/news/ps3s-motion-control-approach-speak-softly-and-carry-a-big-stick/

Read it at your own risk.

Trying to filter out the PR fluff, this is what I found interesting in that article:

“From a third-party perspective it's easier to develop for, you can use the same code base that you currently use for PS3 or 360 or even the Wii in order to get a motion controller game out."

This indicates that the controller setup with wand and nun-chuck (left hand dualshock or whatever) will basically offer the same buttons and analog sticks as the dualshock, where the wand motion could replace one of the analog sticks.

 

[(((interference)))]

I wonder why all the press are still complaining about the Natal lag when they've shown it is as good as Halo 3 on average (sometimes even better) and much better than the input lag in games like KZ2 and GTA4.

 

[rabidrabbit]

But can input lag of a traditional controller be directly compared with the input lag of a full body motion sensor?
It might be approximately the same, but how it feels for the player is more important than what it is in paper.
When you see your "mirror image" responding off-sync with your motions is more distracting and disturbing than when there's a physical controller between you and your avatar, which isn't mirroring your movements 1:1.
The lag is something one might get used to, but it sure doesn't give a good first impression.

 

[Crossbar]

I wonder why all the press are still complaining about the Natal lag when they've shown it is as good as Halo 3 on average (sometimes even better) and much better than the input lag in games like KZ2 and GTA4.

When was this shown?

Sounds like a pretty stupid comparison as lag in FPS games is rarely a controller input issue.

 

[Silent_Buddha]

But can input lag of a traditional controller be directly compared with the input lag of a full body motion sensor?
It might be approximately the same, but how it feels for the player is more important than what it is in paper.
When you see your "mirror image" responding off-sync with your motions is more distracting and disturbing than when there's a physical controller between you and your avatar, which isn't mirroring your movements 1:1.
The lag is something one might get used to, but it sure doesn't give a good first impression.

It's not too different from a platformer where you press a button to jump and X ms later your character jumps. Or when shooting. Or turning in a FPS (controller latency here is particularly infuriating). Heck with Wii, I've noticed considerable controller latency for motion controls. Haven't tried the new 1 to 1 controllers, so not sure if those have been improved.

So I can understand those feelings. But still it isn't all that different from controller latency in a standard controller, assuming they actually do get it that low.

Regards,
SB

 

[Crossbar]

It's not too different from a platformer where you press a button to jump and X ms later your character jumps. Or when shooting. Or turning in a FPS (controller latency here is particularly infuriating). Heck with Wii, I've noticed considerable controller latency for motion controls. Haven't tried the new 1 to 1 controllers, so not sure if those have been improved.

So I can understand those feelings. But still it isn't all that different from controller latency in a standard controller, assuming they actually do get it that low.

The difference is that when you press a button you move your finger just 1 mm and that information is intantaniously sent to the console. To detect a movement using a camera you need a larger movement which will be sampled by the camera by a certain interval and then you have the image processing time which all adds up and make the time from that you start your movement until it is detected by the console certainly longer than if you press a button.
If you also have some filtering that removes unintended movements, that will add even more time.

It is quite a big difference in my opinion, but I don´t think this is an issue for the games that Natal is inteded for. People seem to be enjoying the Wii boxing games even though it´s far from perfect so no need to make a big deal of it.

 

[warb]

Jason Bradbury's impressions from the Natal preview event: http://fwd.five.tv/gadget-show/videos/other/project-natal-preview

 

[obonicus]

http://multiplayerblog.mtv.com/2010/02/22/project-natal-timing-the-delay/

If the average lag is 100ms, that's excellent as that's the same as the theoretical minimum controller latency at 30FPS.

If theyre getting 80ms at times, it should mean the software is running faster than 30fps,

From Digitalfoundry's twitter:

MTV Natal lag timings done with a stop watch?! Latency from normal joypad on 30fps game is 100ms minimum. Natal closer to 200ms at gamescom.

 

[obonicus]

Seems to be an undercurrent of thinking that MS isn't capable of doing anything good, and looking for ways Natal will fail...

Where is this undercurrent, exactly? It must extend to more than Natal, which just hasn't demo'd very well to people who actually look at response times. And it doesn't seem like anyone is particularly kind (nor should they be) towards Arc/Wand.

 

[liolio]

Lag to an extend is not only a matter of ms, the brain seems to have higher stander with full body motion sensing. The same is true with musical instrument, you can notice tiny lag while playing guitar though various numeric devices when you are more than unlikely to notice the same lag while playing (hopefully).

 

[Prophecy2k]

I have a question regarding Natal's ability to track more than one person at a time...

IIRC, MS hasn't demoed Natal doing this in any meaningful capacity, yet its been stated (I think) by them and many others who follow the tech that Natal can indeed do this capably.

Given the reports of lag (also bearing in mind it is a WIP) and the recent suggestion that the number of tracking points have been reduced... theoretically, how well would Natal in it's currently demoed incarnation be able to track more than one person? How would the percieved latency be affected too, bearing in mind the heavy number crunching will now be done on the 360 CPU rather than a custom chip?

If Natal was tracking say, 11 points per person and giving 200ms latency...
How would the processor load be affected by tracking 44 points on four people? Would that imply decreased accuracy in multiplayer titles?

I've been wondering this for a while...

 

[Shifty Geezer]

I wonder why all the press are still complaining about the Natal lag when they've shown it is as good as Halo 3 on average (sometimes even better) and much better than the input lag in games like KZ2 and GTA4.

Perhaps because there are different experiences and measurements, and no-one has a clear picture yet. At the end of the, no-one experiences lag as a ms figure, but as a responsiveness, which various with both actually input-to-display latency and the application. 10ms of input-to-display latency is undetectable in a shooter or platformer, but utterly crippling in trying to play a musical instrument. Likewise if your mouse were to lag a tenth of a second, it'd be very annoying and make drawing things hard. If the applications being shown require extremely low latency (and following your own actions on screen 1:1 with your own body movements, this is something that's going to be far more noticeable), the same degree of latency as another game could be more prevalent and disturbing.

We really are just going to have to wait and see on this one. Measuring isn't going to be much use (unless the measured latencies really are on the tiny side!) as that doesn't provide a worthwhile context. Until we get ahnds-on ourselves, we can only go by user reports what they personally experienced. Comparing numbers from different devices and in different contexts isn't going to be any use in determining the speed of Natal as an input device.

 

[patsu]

Need to experience it myself.

I was thinking in full body tracking, the lag for multiple body parts may add up in order to interpret a complex action. It is not tracking a simple, relatively constant thumb twitching action. So overall when a third party tries to observe the player, he/she may notice a more severe lag although locally, if we break things up to measure, the lag may be constant and small.

Personally I still prefer the original IR + camera tracking method. e.g., tracking finger gestures, identify player for green screen effect, and ignore the rest, just because it is more efficient. Only some games need to track the entire body (e.g., exercise, dancing, acting). Even so, the exact match may not matter. It may be more important for the users to have fun rather than match the exact 3D posture Natal is expecting.

 

[Silent_Buddha]

IIRC, MS hasn't demoed Natal doing this in any meaningful capacity, yet its been stated (I think) by them and many others who follow the tech that Natal can indeed do this capably.

Aye only demo with actively tracking multiple people was the paint demo as far as I can recall (2 people). Otherwise it's mostly just shown the ability to pick a person out of a crowd and successfully track them while disregarding the rest.

Doubt we're going to get any other significant information prior to E3. Although GDC might have a shot at revealing more info...

Regards,
SB

 

[Silent_Buddha]

Where is this undercurrent, exactly? It must extend to more than Natal, which just hasn't demo'd very well to people who actually look at response times. And it doesn't seem like anyone is particularly kind (nor should they be) towards Arc/Wand.

Just look at the comments in the link that Warb posted above.

PS3 is better END OF... xbox's natal is probably gonna be a fluke like the consle itself they'll hype it up and it will end being total bull.

That's the undercurrent I'm talking about among many (obvious not everyone) in the general populace. And there's plenty more like that in various game blogs/newsite comment sections. By volume, more than people defending X360. Perception among many being that MS still hasn't proven they can do a console right.

I have a few friends and relatives that after having used both consoles mirror those feelings. PS3 is better, X360 has no chance. When asked why. Response is quite often either,

It's a Playstation, duh.
Xbox 360 is made by Microsoft, duh.
I don't know, why, it just is.

Again this obviously isn't everyone, X360 wouldn't be doing as well as it is if it wasn't. But I run across this far more than I run into people saying X360 rules and PS3 has no chance.

And that feeling filters down to Natal as well. Interestingly enough, I've also run into far more people that know about Natal than know about PS3 Arc.

Regards,
SB

 

[Squilliam]

Doesn't ARC have even more useability flaws than the Wiimote?

People complain about waggle, and how tiring it is to repetitively perform motion with the Wiimote, but in general use it can be rested on the knee for pointing style games. This is because the camera is in the controller and a small change in angle nets a large change in perspective.

The Arc is different, as it has the camera outside of the controller you'd have to make large movements of your hand for typical motion/pointing controls which are comparable to the Wiimote. You can't just rest your hand and you can't leave the pointer centred to the middle of the screen for aiming etc because that would be tiring.

So how can you play for an hour of say Resident Evil 5 (Arc) if you cannot keep your arm at rest like you can with the Wiimote or a traditional controller? If its even more tiring to use than the Wiimote then its even less suited for core games and even more suited to casual short pick up and play style games.

 

[obonicus]

Just look at the comments in the link that Warb posted above.

SNIP

What you're describing are console warriors. Some people still hate Sony for 'killing Sega'. Some people hate Sony for usurping Nintendo's rightful place in the videogame hierarchy. Some people dislike Sony just for having been successful with the PS2. Some people carry grudges against Nintendo from the 16-bit wars. Some people hate Microsoft just for not being Japanese. I was talking about the press, though -- they're generally pretty positive towards Natal, even though it hasn't demoed that well.