Microsoft HoloLens [Virtual Reality, Augmented Reality, Holograms]

[iroboto]

Sure it'll be better on AR than mobiles, but $n hundred better? Hololens would be the crème-de-la-creme of AR interactive experiences, but it's not essential and that impacts its viability IMO. A lot of the suggested uses I read don't need AR. Kinda like suggesting you need a theatre and real people to convey a story. Well, a flat screen with images of people works to present the same story.

If iPad incorporated a depth cam and Hololens like computer vision, I expect the same industrial AR showcases to be met with far greater sales than the AR headset.

heh uhh I mean those are good points.

I'm going to be objective about Hololens here: the reality of the device is that it works, its designed for function. Sure you don't get the FOV, but everything else it does insanely right.
To begin the AR is fixed onto your own vision, you control what is draw by where you point your head. That alone is useful because you want to see AR where you are looking, you want that perspective. Having to look at a mobile screen that is not at head view means your looking down at a 2D screen drawing into a space that you still don't have actual 1:1 reference to, further more, your FOV is limited to the mobile screen camera. You will be craning your neck trying to see things in AR on a mobile device which is IMO, the reason why AR hasn't taken off on mobile. It just sucks.

The biggest advantage that Hololens has here is that not only is the device comfortable and wearable for long periods of time, but both of your hands are free to do actual work. Something you cannot do with a mobile screen. A company can have a dozen workers around the world in remote locations who maybe do not have the technical know how to do everything, but if they put on a hololens, they can have an engineer from another part of the world guide them through a repair. Something that is much more difficult to achieve without hands free, and something that is made simpler by hololens which is audio assisted, vision assisted, interactive and hands free.

Lets talk about the quality of AR on Hololens. I'm not going to blow smoke, it's not a screen you are looking at, there is no way to describe what or how it is drawing these images into my eye - but I know I did not get sick, and I know the entire experience felt comfortable with the exception of needing to get around the FOV aspect. I did not have to change the way I look at an object like I do with VR, I didn't have to keep my eyes straight to see the AR objects. Its generally impossible for any ipad device to get to this level of AR quality - the technology is just not designed well for it. There is a big noticeable different between 3D VR screens, and this AR screen. It's like its there, I'm not staring at a screen that is close to my eye, I'm just looking at an object that exists right there.

Lets talk about cost. Realistically lets gun for $5000 tomorrow. That's about 7-10 ipads or 7-10 mobile devices. It is still within the realm of reasonable pricing for a company to take on when some companies issue 1 mobile device per employee. Having a few of these around is quite powerful depending on the type of work you are doing. Not everyone needs one AR to be useful.

You can recreate worlds on the hololens, I know it sounds weird but it's possible. If you can scan a room for instance, you could replicate the AR world of it. Then you can do like renovations in AR and see what the room feels like without having to be there, or to do it on your company grounds. Now a larger FOV would be ideal in this case, but that's something that can come in time.

 

[Deleted member 11852]

HoloLens: it's insanely right. I can see Microsoft using that!

 

[eastmen]

I still have my doubts about the insane pricing you guys keep bringing up. WSJ said MS compared it to the xbox one by saying it costs more. Its a long way from $350 to $5,000 . It could be the protypes they have now are using a small run of the HPU which could be factored into people saying its expensive.

 

[Shifty Geezer]

For example, would it be even remotely accurate at placing an object correctly within the real life environment and then maintain that while a user walks around it looking at the computer generated project from multiple angles and multiple distances in real time?...

Hololens is better in some cases. However, Iroboto's examples weren't those...

We've been discussing the usage of hololens for it's usage in mueseums, historical locations where you can see where things where, overlaid where things are today etc. Or for any sort of educational purpose, like wearing Hololens to see what's going on inside a dishwasher while it's oeprating. Learning the biology of an animal without having to cut it up and create all these corpses for study.

All those activities would be doable on a tablet with decent computer vision. Basically passive consumption of information doesn't need an AR headset. Interactive content will want Hololens, but you're getting to smaller and smaller markets there. I'm also not the slightest bit convinced about the fabulous 3D interaction. With zero force-feedback, accuracy of creating things in 3D is going to be hampered. The stage demo of the helicopter thing was a great case in point - the pieces weren't symmetric and the machine would be aerodynamically unstable. Moving pieces around with grid snap and edge snap would work, so it'll be great for architecture say, but a Hollywood style throwing virtual pieces together and creating a real product is IMO nonsense. For anything serious you'll need CAD and a stable interface rather than hands grabbing floating, virtual componentry.

 

[MfA]

Until they are the size of a large set of sunglasses the use cases are limited.

VR because of it's single user, fixed location use can get away with more elaborate hardware ... AR, not so much.

PS. in my opinion worrying about getting it wireless at this point in time is completely counterproductive and will be for a long long time ... I realize wires are kinda bad publicity, but I doubt it's as bad as all the extra weight and volume being wireless adds. Just put all that mass in my pocket or on a necklace.

PPS. if resolution is their main reason for going small FOV, direct laser projectors could be a solution. They are relatively easy to re-engineer for foveated projection (of course the field is probably patented into near uselessness, but as Microsoft showed with depth cameras it's not that expensive to just buy everyone out as long as there is no real market ... because the field is patented into near uselessness).

 

[iroboto]

Hololens is better in some cases. However, Iroboto's examples weren't those...
All those activities would be doable on a tablet with decent computer vision. Basically passive consumption of information doesn't need an AR headset. Interactive content will want Hololens, but you're getting to smaller and smaller markets there. I'm also not the slightest bit convinced about the fabulous 3D interaction. With zero force-feedback, accuracy of creating things in 3D is going to be hampered. The stage demo of the helicopter thing was a great case in point - the pieces weren't symmetric and the machine would be aerodynamically unstable. Moving pieces around with grid snap and edge snap would work, so it'll be great for architecture say, but a Hollywood style throwing virtual pieces together and creating a real product is IMO nonsense. For anything serious you'll need CAD and a stable interface rather than hands grabbing floating, virtual componentry.

Oh. LOL.
Those examples were developer discussed after our hololens demos. They didn't say for us to use it like that. We were trying to come up with easy programs to make to nab ourselves a devkit. Yea they aren't the best examples.
It's not an accurate tool for modelling or creating in 3D. But it's certainly an effective tool to transforming someone's CAD work into real life. You can still use CAD with hololens on mind you. I was using an architectural program and manipulating the building with mouse and keyboard and computer screen with hololens on. But I could also look over at the cardboard model thing and the AR of the model would appear there as well.

Though I will admit fully that this device unless you are exploring it serves as a cooperative experience.

 

[iroboto]

HoloLens: it's insanely right. I can see Microsoft using that!

Lol and that was me being objective[emoji38][emoji38][emoji38][emoji38][emoji38][emoji38][emoji38]

 

[iroboto]

Until they are the size of a large set of sunglasses the use cases are limited.

VR because of it's single user, fixed location use can get away with more elaborate hardware ... AR, not so much.

PS. in my opinion worrying about getting it wireless at this point in time is completely counterproductive and will be for a long long time ... I realize wires are kinda bad publicity, but I doubt it's as bad as all the extra weight and volume being wireless adds. Just put all that mass in my pocket or on a necklace.

PPS. if resolution is their main reason for going small FOV, direct laser projectors could be a solution. They are relatively easy to re-engineer for foveated projection (of course the field is probably patented into near uselessness, but as Microsoft showed with depth cameras it's not that expensive to just buy everyone out as long as there is no real market ... because the field is patented into near uselessness).

It is already wireless. The device has enough clearance for glasses, and it isn't awkward to wear. There is an internal head strap like a halo ring that goes around your skull, the device attaches to the halo ring.

It is quite comfortable to wear and the device is not too weighty at all. I found it to be more comfortable to wear over oculus because it's not front weighted.

 

[MfA]

it isn't awkward to wear. There is an internal head strap

My definition of awkward is different from yours.

 

[Silent_Buddha]

I still have my doubts about the insane pricing you guys keep bringing up. WSJ said MS compared it to the xbox one by saying it costs more. Its a long way from $350 to $5,000 . It could be the protypes they have now are using a small run of the HPU which could be factored into people saying its expensive.

They didn't just say it would cost more than a console, they said it would cost significantly more. Lots more. In other words, nowhere near what an Xbox costs.

One current Microsoft executive said HoloLens would cost significantly more than a game console, which runs more than $400.

From the original source of the quote (http://www.nytimes.com/2015/05/03/technology/microsoft-yes-microsoft-has-a-far-out-vision.html )

Until they are the size of a large set of sunglasses the use cases are limited.

VR because of it's single user, fixed location use can get away with more elaborate hardware ... AR, not so much.

PS. in my opinion worrying about getting it wireless at this point in time is completely counterproductive and will be for a long long time ... I realize wires are kinda bad publicity, but I doubt it's as bad as all the extra weight and volume being wireless adds. Just put all that mass in my pocket or on a necklace.

PPS. if resolution is their main reason for going small FOV, direct laser projectors could be a solution. They are relatively easy to re-engineer for foveated projection (of course the field is probably patented into near uselessness, but as Microsoft showed with depth cameras it's not that expensive to just buy everyone out as long as there is no real market ... because the field is patented into near uselessness).

I've seen user impressions that have mentioned that it is more comfortable and less tiring to use than the Occulus Rift or Sony's Morpheus or pretty much any VR HMD currently in development.

So at least they basically have the ergonomics right.

And for any near future use, it's targeted at industry so size isn't that important as long as it can be potentially used for extended periods of time without generating significant fatigue.

In 2-3 years (or potentially more) it'll be interesting to see where the hardware technology is at for a potential consumer introduction.

Regards,
SB

 

[MfA]

I've seen user impressions that have mentioned that it is more comfortable and less tiring to use than the Occulus Rift or Sony's Morpheus or pretty much any VR HMD currently in development.

So at least they basically have the ergonomics right.

No, they have better ergonomics than current VR ... that doesn't mean it's good enough for AR, consumer or industry.

In 2 to 3 years putting batteries in the glasses will still be a bad idea.

 

[Silent_Buddha]

No, they have better ergonomics than current VR ... that doesn't mean it's good enough for AR, consumer or industry.

In 2 to 3 years putting batteries in the glasses will still be a bad idea.

Your opinion obviously. But there will be use cases, at least in industry, where an untethered device will be far more desirable than a lighter tethered device with wires (which could be a potential health hazard depending on the environment) attached to another device.

But here's the point. It doesn't matter.

If a tethered lighter device is desirable, a company will make it. Microsoft doesn't care who makes the AR device. They just want Windows to be the OS running it.

Regards,
SB

 

[pjbliverpool]

VR is great and can do a lot of stuff but its a closed experience. I wouldn't mind playing games with my friends together and being able to see each other.

A high res camera on the outside of the VR helmet that than can add elements of the real world into the virtual one (people, objects, barriers) would be a better solution IMO. In fact I suspect it's an inevitable development. Imagine being able to see your own hands in the game world - not a rendering of them but an actual high resolution camera feed of them overlaid into the virtual environment with the movement of your hands being tracked by the camera to allow them to interact with the game world.

 

[Deleted member 86764]

I wonder whether the narrow field-of-view is somewhat intentional; i.e., intentional to the point of expanding it might break the immersion. I can imagine it’s there to prevent your eyes from looking to the left or to the right, as the ideal scenario would be for them to remain facing forward in line with your head. If you were to look to the left, would the image remain lined up with the real world? I can’t see how it’s possible unless they included some kind of eye tracking device. Even then the processing would need to increase dramatically to deal with the scenario.

I wonder whether it’s ‘training’ you to look forwards. I can’t see how this would be a problem with VR since you’re not including the real world.

 

[Jwm]

I wonder whether the narrow field-of-view is somewhat intentional; i.e., intentional to the point of expanding it might break the immersion. I can imagine it’s there to prevent your eyes from looking to the left or to the right, as the ideal scenario would be for them to remain facing forward in line with your head. If you were to look to the left, would the image remain lined up with the real world? I can’t see how it’s possible unless they included some kind of eye tracking device. Even then the processing would need to increase dramatically to deal with the scenario.

I wonder whether it’s ‘training’ you to look forwards. I can’t see how this would be a problem with VR since you’re not including the real world.

That is a good question. @iroboto how did you feel the device was for its current use cases? Was the FOV comfortable for say modeling buildings?

It does not seem the FOV is good for watching TV on a virtual 100" screen. Ha!

 

[iroboto]

That is a good question. @iroboto how did you feel the device was for its current use cases? Was the FOV comfortable for say modeling buildings?

It does not seem the FOV is good for watching TV on a virtual 100" screen. Ha!

I actually didn't mind the side viewing angle as much as I had an issue with the height of it. Then again, the example I worked with was vertically biased; it was a building.

I think when it comes to the discussion of whether the FOV is acceptable, I'd say when I'm close to an AR object I'd like to have it fill my FOV entirely. But if I"m far away, I could do with less FOV to ensure I'm not endangering myself.

I see the value of both, it can be easy in certain lighting situations for AR to distract you enough from the standard environment, such that if I were walking in an incomplete building and mapping the environment over blue prints, I would definitely prefer the safety of having less FOV. In a controlled safe environment, I would prefer more FOV.

 

[Deleted member 86764]

I can't help but think AR is much more difficult to achieve compared to VR. You’d need to have a match between the real and the artificial, especially alignment. I can quite easily imagine future iterations including eye tracking and pupil dilation (for focal ranges). Bloody difficult to do properly since the eye moves pretty quickly and blinking might cause problems.

Good on Microsoft for having a really good implementation though. I really hope that they’re able to make the projected image appear to be behind objects that are closer to your face.

 

[Deleted member 86764]

I think when it comes to the discussion of whether the FOV is acceptable, I'd say when I'm close to an AR object I'd like to have it fill my FOV entirely. But if I'm far away, I could do with less FOV to ensure I'm not endangering myself.

Did you always keep perfect alignment between your head and your eyes? I'd be interested to know whether the image 'breaks' if you look away. I imagine when you're testing the device, you'd be looking directly at the projected image anyway! I don't think I'd be looking at anything but it.

 

[Scott_Arm]

Did you always keep perfect alignment between your head and your eyes? I'd be interested to know whether the image 'breaks' if you look away. I imagine when you're testing the device, you'd be looking directly at the projected image anyway! I don't think I'd be looking at anything but it.

Maybe that's part of the reason why they constrained the field of view? Forces you to look straight at it.

 

[iroboto]

Did you always keep perfect alignment between your head and your eyes? I'd be interested to know whether the image 'breaks' if you look away. I imagine when you're testing the device, you'd be looking directly at the projected image anyway! I don't think I'd be looking at anything but it.

It's not like VR where the engine decides where your focal point should be. Ultimately your tilt of your head determines how VR is rendered, but your eyes can freely wander and see the AR. It for the most part was perfect, there were times where it was clear the AR should not have been tilted with respect to my eyesight, but that just might have been poor calibration or the headset was moving.

The pupils of your eyes are measured prior to putting on the device, so that likely helped in keeping this focused and working.