Wintel VR: Intel, Microsoft, And Their Two-Pronged Plan To Democratize VR, AR And MR

Status
Not open for further replies.

d_kuhn

Distinguished
Mar 26, 2002
704
0
18,990
2
First of all... Hololens is textbook AR. In fact - Hololens is more AR than current "AR" things like Pokémon Go. I think if people just reject a Microsoft branded AR term they'll eventually give in and call it what it is. There's VR and there's AR... there currently are no other fundamental technologies in play, just variants of those two things.

As far as the Intel approach of Merging camera input instead of using a passthrough... I see it as having some advantages and disadvantages. One advantage is it won't suffer from fading in very bright conditions, and the computer based content (Holograms on Hololens) won't have to be delivered through a more complicated viewing technology (which currently adds artifacts). On the downside... human vision is MUCH better than any camera (especially the super-compact ones used in headsets), and also the display tech used can't cover the full range of a real scene (in a number of ways... dynamic range, color reproduction, temporal reproduction, etc...). I also suspect that the Hololens pass-through approach will ultimately be more energy efficient (battery life).

I think short term it will be worth exploring both technologies and both will work better in some situations... long term I'd put my money on Hololens style pass-through for AR.
 

wifiburger

Reputable
Feb 21, 2016
415
1
4,810
9
waste of time, right now VR is not worth the effort with putting up with accessories,
it's bad cause instead of having game studios create amazing games they dedicate their resources into this gimmick because it's the new thing !

it happened with wii mote, everybody jump unboard for some truelly garbage games that sold on the fake promise of that wii mote accesory, it's just another reapeat and i'll pass no thank you
 

scolaner

Reputable
Jul 30, 2014
1,282
0
5,290
2


Have you tried any VR/AR experiences?
 

alidan

Splendid
Aug 5, 2009
5,303
0
25,780
0


the only people who don't believe in vr have never tried vr, id willingly go back to what i used 15~ years ago if i could have it today, but what we have is currently to expensive for me, and i would need massive platform upgrades to deal with it.
 

scolaner

Reputable
Jul 30, 2014
1,282
0
5,290
2


That's what Wintel is trying to solve--they're trying to bring the cost down without sacrificing (much) performance. I hope that happens. I don't care if it's Intel, MSFT, Google, Oculus, HTC, or whoever, just as long as it happens.
 
Feb 18, 2016
463
0
4,960
57
VR & AR, weren't enough? There has to be a "MR" too? Can't we just shorten the first two down to MR? Am interested to know what long term effects MR may have, or potentially have, on the human brain. Or even how many hardcore Pokemon Go players have started seein' those virtual critters in their dreams, and other places they aren't supposed to be, so far.
 

scolaner

Reputable
Jul 30, 2014
1,282
0
5,290
2


Agreed, the additional "R"s are a little much. I addressed that very nomenclature issue in the sidebar.
 

alidan

Splendid
Aug 5, 2009
5,303
0
25,780
0


I honestly don't trust intel/microsoft to being the hardware cost, as in the computer needed, down in price. they may be able to make a headset cheaper, but osvr also does that. and for my money, i see no use in room scale because i dont have the space for it, however a racing sim or even an arcade racer, that, just having that extra axis is enough to change how the game is played, even if its not in 3d.

Hell, there is a 4k vr out of china I have been looking at, not so much for gaming, but for watching movies.
 

scolaner

Reputable
Jul 30, 2014
1,282
0
5,290
2


You don't "trust" them? Or you don't think they can do it?

I mean, they're not doing it to be altruistic. They want to dominate the market and make buckets of money, and the best way to do that is to exploit the mainstream portion. So it's one of those situations where the Venn diagram of what benefits those companies and what benefits end users aligns.

Also...as I wrote in the article, they're really trying to develop a set of specs that work and then will get OEMs to make the actual stuff.

Also also, I feel you on the roomscale issue...but that's where HoloLens and Alloy are ideal. They give you whatever environment you're in. The XR experiences don't have to be "room scale" the way Vive is. Besides, I think a lot of the value of AR headsets like the HoloLens will be around productivity and content consumption.
 

bit_user

Splendid
Ambassador
I totally agree on this point. That MS will try to use their trademarked term is almost a given, but I really hope everyone rejects it, in favor of "AR".

I disagree. What's truly novel about Intel's Alloy is its ability to introduce non-translucent imagery. They can also mess with your environment in other ways, such as replacing your room with a scenic vista, while still letting you see all the furniture. It can also selectively remove or alter items in your environment, to a large extent.

I think it's hard to really get your head around the differences between MS' approach vs. Intel's, without actually trying both. In the long term, I think Intel is on a better track. The downside is that it's currently more obtrusive, invasive, and fatiguing than MS' Hololens.

So, I accept the term MR as a seemless mixing of VR and AR that can achieve things not possible when limited strictly to either one.

The big question for me is how extended & frequent use of VR affects vision and depth perception. Lightfield displays would largely allay those concerns, I think.
 

scolaner

Reputable
Jul 30, 2014
1,282
0
5,290
2


This is why I'm moving to "XR" as a catch-all. What we're seeing with "MR" is that VR and AR are blending. So...XR.
 

bit_user

Splendid
Ambassador
You have to re-engineer how the pipeline works to a degree.
Yeah, with moar ray-tracing, plox. They could totally buy Imagination, if they wanted. But I know Intel looked at realtime raytracing, back in the years prior to Larrabee's cancellation. I recall they even had a demo of realtime-raytraced Quake on an 8-core machine (dual Core2-era CPUs).

Intel is looking to solve the problem by leveraging technology it knows best: processors.
I don't see this amounting to much. GPUs have evolved for over 20 years to solve exactly this sort of problem. Where they differ from CPUs, it's with good reason.

there may be advantages to doing things like barrel distortion, and time warp, and chromatic error correction on the HMD itself
True. This is fertile territory for fixed-function hardware blocks.

and you can do things like decouple that from the resolution and framerate of the application
This isn't worth a whole lot, IMO. You really can't drop the front-buffer resolution or framerate much, without hurting the experience. Maybe not in a vomit-inducing way, but it'll still be distracting. I doubt they can get more than a 2x savings, before it becomes unacceptable.

Just as you can’t buy an Intel PC
Ahem, NUCs?

Also, let's not forget Intel's SSD business. I believe they don't want to make HMDs, but with them building motherboards, NICs, and now SSDs and NUCs, the lines between chip-making and end-user products have certainly gotten more blurred.
 

scolaner

Reputable
Jul 30, 2014
1,282
0
5,290
2
Ahem, NUCs?

Also, let's not forget Intel's SSD business. I believe they don't want to make HMDs, but with them building motherboards, NICs, and now SSDs and NUCs, the lines between chip-making and end-user products have certainly gotten more blurred.
Ha! You got me there. Good point.

The sense I get is that they really don't want to make HMDs. I think they think RealSense cameras and deep learning are going to be big, and that may be satisfactory for them. I mean think about it--they have this already-expensive IP around those things. Why go through the agony of building an HMD? Design, manufacturing, marketing...ugh. What a pain.

That's not to say they won't...it's clear to me that there's much they haven't decided on or figured out yet. I think they've said the right things and tagged the smart bases so far and have left things sufficiently open-ended.
 
Status
Not open for further replies.

ASK THE COMMUNITY

TRENDING THREADS