ZackieChan
Member
That's sort of their job, isn't it?
-
Hey Guest. Check out your NeoGAF Wrapped 2025 results here!
New Microsoft hololens demo. Goddamn future
- Thread starter Kinitari
- Start date
- Status
That's sort of their job, isn't it?
killer rin
Member
Yes and no. Microsoft is only one company, you cant reasonably expect them to show off every use case of a product, especially not one which hasn't released yet; They are doing what every hardware developer does, they show off its capabilities, give you a few handpicked demos which are easy to build and then give out the APIs to developers to do stuff with.
Wait until developers get their hands on it, that's when you see the novel uses coming out for it.
ZackieChan
Member
Yes, but "creator of product has to show that there is some value to it" is their job, I would think. And for some of us, at least, they haven't met that. Some of us obviously believe they have!
LookAtMeGo
Member
This is amazing.
Alx
Member
Makes sense. I think we're still missing one piece of the puzzle though : tablets (or phones) with integrated depth sensing. I think Dell intended to produce one, but I don't know how far they are. They could be a better entry level product for that AR environment.
Well even if the screen is mostly translucent, it does make it very hard to see the user's eyes, which is a big drawback for social interaction.
We communicate a lot with eyes ("mirrors of the soul") and hiding them is not very socially acceptable, that's why people look like douches when they keep their raybans on when talking to other people (especially indoor).
When Fortaleza leaked I hoped you could see enough of the face through the visor for that to be a minor annoyance, but we're not there yet. That's too bad because otherwise it could be a cool device for social interaction : you could 3D-skype people by having them scanned by a kinect on their side, and display them in your living room sitting on your couch (just like in that Kingsman scene). It's still possible with the current system, but you would see a bunch of people wearing gigantic sunglasses.
jizzlobber
Member
Seen some pretty stupid replies, to many screens and sensory overload... Really?
You can pin you're screens behind your lounge if you want, un-pin when you need it... It's pretty much like having apps on a phone.
I can see this being useful, I thought the first demo was impressive but gimmicky, now I see the potential.
Imagine if apple showed this, best thing since sliced bread!!
You can pin you're screens behind your lounge if you want, un-pin when you need it... It's pretty much like having apps on a phone.
I can see this being useful, I thought the first demo was impressive but gimmicky, now I see the potential.
Imagine if apple showed this, best thing since sliced bread!!
Alx
Member
Thanks, that was interesting.
Shame that the raw data from the sensors won't be accessible from the API, although that makes sense for their goal of having a software environment that doesn't depend on the device.
Alx
Member
I just watched the videos again, and apparently the current version does always overlay the holographic object over the real image. They were very cautious for it not to happen on stage actually (they had a skilled cameraman there), but you can see it for a frame or two here :
https://youtu.be/mSCrviBGTeQ?t=166
Kipman's hand brushes over the robot head for a few frames, and it seems to be hidden by it although he's closer to the camera. You can also see the dots of the planned trajectory appearing over his feet.
I think in theory the hololens has enough information to know there are objects in front of the hologram, using its depth sensors. But it may be difficult to do a perfect rendering of it, since the depth information would be less accurate near the edges of objects (there's also a slight offset between the depth sensor position and the wearer's eyes to consider).
Alx
Member
The base problem is not that hard, it was actually one of the very first hacks with the original kinect.
https://www.youtube.com/watch?v=P3gfMXwQOGI
Handling occlusions when you're dealing with 3D information is quite straightforward, but the main issue is making it look good in all the places where you have less accurate (or missing) measurements. The new kinect data is already less noisy and more accurate than what we see in the above video, but it's not perfect (and will never be).
https://www.youtube.com/watch?v=P3gfMXwQOGI
Handling occlusions when you're dealing with 3D information is quite straightforward, but the main issue is making it look good in all the places where you have less accurate (or missing) measurements. The new kinect data is already less noisy and more accurate than what we see in the above video, but it's not perfect (and will never be).
Pie and Beans
Look for me on the local news, I'll be the guy arrested for trying to burn down a Nintendo exec's house.
Easier to solve when the whole thing is a set video quality being fed back, a lot harder when you're overlaying real life where it would have to then be overlaying crap video feed of the occlusions.
Its going to be something Microsoft hides all the way up until launch and probably even beyond that. Along with the box FOV, its gonna be something AR struggles to solve for a long time.
Benzychenz
Member
Eh, give me actual screens floating around like in Iron Man and I'll care. Don't want to be wearing glasses.
Alx
Member
You don't need to overlay a video feed in that case, just consider in the CG layer a mask of the pixels where it should not appear, so that you would just see "real life" instead. In short, detect where an obstacle is in front of the object, and disable object rendering there.
Only when you're not perfectly accurate, the transition between the real obstacle and the half hidden object will appear weird and/or fuzzy. But it should be better than always having the CG layer at the top, since that would break the depth illusion.
WAR recon zero
Banned
What's the tentative price?
Alx
Member
"significantly more than a game console" is all we know, but that's not a surprise. No way it costs less than $1K imo, probably more.
You'll probably be dead when the kind of stuff you see in Iron Man is realistically possible.
ichtyander
Member
Good find, that's exactly what I was saying earlier in the thread, this needs to be solved and working first otherwise the illusion is completely broken. Imagine pets jumping in front of you, people walking past etc
There was a very noticeable situation like that during the HoloLens reveal when they switched to a "first person" view for a moment to show the glance and hand gestures.
StudioTan
Hold on, friend! I'd love to share with you some swell news about the Windows 8 Metro UI! Wait, where are you going?
Video of 3 devs talking about what they did during the HoloLens dev session. They go through some of the coding and talk about their experience with the device itself.
The comments are very positive. One thing I thought was interesting was when they talked about the audio capabilities which I haven't seen talked about much. They were extremely impressed with it.
Battery life seems like another big issue right now.
The comments are very positive. One thing I thought was interesting was when they talked about the audio capabilities which I haven't seen talked about much. They were extremely impressed with it.
Battery life seems like another big issue right now.
VlaudTheImpaler
tl;dr
Good thing I checked before I posted, was just about to post this.
They seemed pretty excited and apparently the over all impression from everyone there is an extremely positive one.
They spoke about the FOV a bit and that even though it was smaller than they liked they still felt immersed which is good news. However, since the product still seems a couple months from release and the FOV is everyone's biggest gripe hopefully it can be addressed by then.
Anton Sugar
Member
Good find, that's exactly what I was saying earlier in the thread, this needs to be solved and working first otherwise the illusion is completely broken. Imagine pets jumping in front of you, people walking past etc
I'm curious how they'll handle that. They clearly have some way to designate objects that are exceptions to the overlay--when they debuted this in January, one of the demos had you walking around the surface of Mars, but there was an actual computer (which was in the room) on the surface of Mars with you.
Every stage demo you've ever seen has been well-rehearsed, and several impressions have mentioned little to no lag:
ha, are you joking? It's just old stuff with Kinect over there...
it's the reason i'm over here..lol.
so the whole FOV thing...is it because of the (current) limited rendering power of the machine or is it because of the size of the lense? or both, for the time being.
also is there any other devices that use photons to display an images? just trying to get an idea of the image quality here.
Poodlestrike
Banned
Really? The gamingside threads on Hololens were straight-up dumpster fires the last time I checked.
Brian Griffin
Member
A PC already provides me with all the applications a smartphone does- without having to keep a bulky device in my pocket.
LOL @ the thought of a smartphone being "bulky."
Congrats.
P.S. The crucial difference between a PC and a smartphone is that a smartphone is readily accessible anywhere you go. Hololens can't compete either, because you'll look ridiculous in public and you'll quickly feel discomfort with something on your face all the time.
I don't really understand this argument. Phones do what computers did before, just more conveniently and portably. Hololens does what TVs/PCs/Phones did before, just more conveniently and portably.
Brian Griffin
Member
When I actually did upgrade to a smartphone from a tiny flip phone, my pocket did get a lot more cramped. That's not really the point though.
What you posted has been said about every piece of technology that attempts to encapsulate and improve on the features of an existing product. Most people don't need it now but it's likely that it will become a near-necessity in the future like smartphones are now.
I for one welcome our new holographic overlords.
ichtyander
Member
I'm assuming they might've modeled a near exact replica of the computer as an AR object that was then used to hide the Mars terrain. In a controled environment, I'd say it's doable but doing that in realtime situations with all kinds of rooms, it gets much more difficult.
Although, I can imagine having the Kinect hardware create these occlusion meshes as well as having something like the Leap Motion create such a mesh for the user's hands. I've no idea how feasible all of this is at the moment, probably won't be quite there yet for a few years.
StudioTan
Hold on, friend! I'd love to share with you some swell news about the Windows 8 Metro UI! Wait, where are you going?
Yeah, I was being sarcastic.
The interesting thing about the FOV discussion is that it seems the prototypes they had the press try in January had a much better FOV. So that likely means there is another reason the FOV has gotten smaller that's not related to inherent design limitations of the visor. Maybe it does have to do with processing power and battery life?
Yeah I'd like to know if there is a headphone jack or how the audio works.
There's also this rumor: "Paul Thurrott had heard but was unable to confirm that the hologram field-of-view was purposefully kept smaller to reduce motion sickness (something common with VR headsets like Oculus)." (source)
Edit: damn you, Lazaro
People get motion sick from augmented reality? That seems interesting.
Mister Pink
Neo Member
GuitarAtomik
Member
cartoon_soldier
Member
I am guessing that based on type of app, motion sickness is an issue with AR or VR.
Probably a bigger issue in VR though.
If you're already wearing the visor, perhaps that's true. But 1) I'm not going to wear a visor outside my home or workplace, and 2) I'm not going to wear it for any extended period of time.
Theonik
Member
- Status