* Processor/GPU power (HoloLens is x86-based with a dedicated HPU unit for hand/environment tracking)
* Custom OS based on Windows Universal
* Multiple apps can run
* FOV is small
* Display doesn't work with nearby objects
* Brief update delay for real objects - there's a short delay for direct object mesh tracking and a longer delay for spatial anchor updates, IIRC
* Can't draw blackness
From the last few limitations, it seems they may just be using the same waveguide displays as the HoloLens. Variable DoF sounds interesting - I suppose they're using eye tracking to provide blur-DoF effects, but I wonder how quickly they can adapt the image. Would also be interesting if variable DoF means something else.
A tracked controller is nice but perhaps nothing too new. I wonder how they track the controller if it's out of the FoV of the head-mounted camera setup? HoloLens has a clicker controller but that's inertially tracked so it's much more similar to a gyro-mouse than e.g. a Rift controller.
I was really happy to see eye tracking in this list as I consider that a "gateway technology" that opens up a variety of additional really cool features. The big one is probably foveated rendering [1] which both allows for a more natural look as well as (hopefully) more efficient visuals.
We still have to wait for the exact FoV, and the eye tracking seems interesting but it doesn't seems to be better than the Hololens, and won't be from the 2018-2019 Hololens.
Wait, the Hololens has eye tracking? That's interesting, I'd never heard that before.
Anyway, I think the main innovation Magic Leap seems to be pushing over its competitors is its multiple depth of field display. It'll be interesting to see how big of a difference that actually makes in practice.
No HoloLens doesn't have eye tracking. I think they're saying the device as a whole doesn't seem to be better than HoloLens. Which I might agree with. The form factor may be stranger though, which would be an accomplishment.
Yeah the variable depth-of-field not working well at < 1 ft is a just a property of the hardware sensors. I guess I don't know exactly what they are but if it's time of flight or stereo then they could make it work better at closer ranges but then it wouldn't work as well farther away.
That's referring to the user being < 1 ft from a virtual object, not a physical one.
The docs say "Currently we do not recommend to having interactions with objects within 30 cm of your view for prolonged periods of time. When users approach too close to content for us to display, fade content out gracefully rather than sharply clipping the view."
I think it's very unlikely you'll get motion sick in AR, _especially_ AR with a small FOV. The vast majority of your vision will still be filled with reality, so it won't be all that different from wearing a pair of regular glasses, at least from your brain's perspective.
Not to mention that Magic Leap's multiple depth-of-field display will likely simulate normal vision much better than current-generation VR displays, and that there's no artificial locomotion in AR.
I get dizzy about half an hour into using the HoloLens. I am not sure why, I think it may be because the tracking/rendering doesn't quite keep up with fast head motions making the holograms "inconsistent" with the real world. It might also be that it's because of the accomodation/convergence mismatch ...
Do you think it is related to the weight of the HoloLens? I find that many people feel uncomfortable after 20 minutes of use. And at least one person got a headache using it.
What does everyone make of perpetual license to content agreement? Is this just so they can operate the service, or are they able to sell and use your content without cutting you in?
> Specifically, you grant Magic Leap a worldwide, irrevocable (except as set forth below), perpetual, non-exclusive, transferable, royalty-free and fully sublicensable right and license to use, copy, display, store, adapt, publicly perform and distribute Your Content so that Magic Leap can operate, deliver, and improve our Services.
When lawyers set these things up, they tend to keep things as broad as possible. In practice this means that they can show how cool their technology is in a commercial, while showcasing your app without necessarily your approval. Obviously it would be absurd to make money of your content and not cut you in, since those types of actions ripple very quickly in dev communities.
In practice it also means they can bundle your killer game with their headset and not pay you for it. Would they? Probably not. Could they, according to this? Yes. (Not a lawyer blah blah.)
I read the T&C's and decided against signing up. Way too broad and restrictive on just about every aspect from content ownership to legal recourse should they decide to screw you over.
Call it hyperbole, but in light of the last few days Facebook revelations, these terms of service contracts are suddenly quite interesting. On one hand, this one at least, is written in marketing speech and legalese, switching narrative voices frequently, salesperson telling us how thrilled they are about entering a legally binding agreement with us, smoothly segueing to lawyer explaining to us, without a glimmer of salesperson's self-awareness, the lopsided agreement that has both salesperson and lawyer so frickin thrilled.
I hope we are rapidly approaching the day of reckoning where these types of agreements are actively struck down as categorically irrelevant. Claiming up front, complete lack of obligation or blame for anything negative that may ensue from the use of, and the claim to exclusive rights to anything positive that may ensue from the use of, your software, is bogus, at least in the manner these terms of service do it. If you really claim all of these things, and want anyone to believe your marketing spiel about enthusiasm for legal arrangements, print it on a T-shirt, and go tell the world about it. Don't bombard me with marketing, and then take me off in a corner and get me to sign away any rights just as I've decided to use your product.
I don't advocate for companies being defenseless, but it seems to me that this imbalance of culpability, and ownership rights, if they truly exist as the terms of service claims, might inspire the type of hubris that leads to what we see being revealed today.
If the United States judicial system is so terrible that you aren't willing to subject yourself to it in matters concerning your behavior, fix the judicial system. Don't run to some parallel system you've captured simply by sophistication lent by your resources.
What do you think about FOSS then? Doesn't the license holder of FOSS software deny any liability for using the software, while claiming ownership of any improvements you make and distribute?
My understanding as a non-lawyer is that unless there's an assignment of copyright by contributors they retain copyright to their contributions. I think in practice they also license that contribution under the project license, but some quick duck-duck-go research indicates that this is at least a complex topic. The most reliable source I was able to find in the time I care to spend on it is from the FSF[0].
Even in the case where copyright ownership is assigned to the project owner, it is a very different situation than the case where your created content is just hosted on a service that claims rights to it.
My point is not the veracity of any of the underlying legal principles. My point is the underhandedness of how we are entered into these agreements.
FOSS has in fact, printed their stance it on a T shirt, and subsequently told the world about it. It's actually one of their flagship contributions to the world.
What is different with the standard flow of the presentation of Terms of Service is more like being presented with the documentation and destination fees, and any other fees on the tail end of a car purchase, after you've talked at great length about THE price. For you see, the sales person conveniently considers those as completely different, and very odd to bundle together with the actual cost of the car.
It comes back to my original point. FOSS isn't switching narrative voices throughout their legal document. They actually want you to truly understand what you are agreeing too, because assuming good intentions, the goals of the users of FOSS align with the goals of the creators of FOSS. Contrast that with Magic Leaps TOS. Why else would you layer in persuasive, folksy, nonthreatening ad copy with your legalese, except to shape any negative opinion I might arrive at through only reading the legalese. Thats because their goals are profit at the expense of their customers, which necessarily means adversarial relationships with those customers, on some level.
It's nice that we can start building with the SDK, but I'm more interesting in experiencing what I've made. I'd rather have the device and build a simple hello world demo that inspires me to build more vs spend a bunch of time diving deep into an sdk and while having no idea how the end user will experience the stuff I create. With iOS for example, you could develop a majority of apps by just using the simulator, since the experience is (mostly) 2d. But I remember when starting out, the more fun projects to work on were using the accelerometer. The only way to experience the magic in that case, was with an actual device. With Magic Leap, the entire experience is spatial. So it's probably the other way around where the device is mostly the only way to really understand what it is that you're building.
Yea very strange that they released the SDK and tools without the hardware. How r u supposed to know the ins and outs, quirks, user experience to expect on the glasses without playing around a bit with them first, maybe with a few pre supplied apps by magic leap. I can think of a few reasons, none of them good, to release tools before the platform
I think releasing the developer hardware too early can ruin the whole product launch. People will form their opinion based on the reviews of the devkit hardware and sample apps.
Launching the consumer product without apps is not good either. By launching the SDK early, developers can play around with it. Maybe they can even pick serious and promising apps to some "insider" program and provide access to real hardware.
This is a modern thing which I don't think applies to technology that is (ostensibly) category setting. In the early days of consumer computing, the hardware was released with a handful of supported software that the hardware maker teamed up on, then MAYBE a devkit would come out.
I love seeing open source projects with multiple active developers. Georgeswinger has 47 commits on the project now, and is the third most prolific developer when ranked by commits.
Can someone comment on what are some compelling use cases for Magic Leap that can be developed using these APIs ? Still seems to me that they are constrained by a device that you need to carry
This is my biggest concern with Magic Leap. I'm starting to believe they are going to ship on time, and they will deliver a fairly polished and high quality product.
However, I haven't seen a single compelling use case. I don't want to browse my email with a giant pair of googles. It's less convenient and less efficient than using a phone or computer. I also don't want to spend a couple of thousand dollars for dedicated hardware to play mobile quality, gimmicky AR games.
What problem are they solving? I want to be excited to buy this product, but their demos are just showing random 3D objects floating in space. Billions of dollars and years of development, and they're trying to sell me on some 3D jellyfish? I have money, I like technology, I've been following Magic Leap for years, and yet I feel no buying interest or excitement towards the product.
For that reason, I think it's destined to fail with consumers. The product will have niche use cases with certain businesses, but I think it'll end there.
Maybe a better way to implement AR would be by combining real-world images (through cameras) with CGI into a VR goggle. One problem though: focusing your eyes doesn't work as expected. But at least you can get better black pixels.
The Oculus Rift supposedly made that possible in one of the latest versions. But I'm curious why you strain your neck looking at a monitor and wouldn't looking at a virtual monitor. Don't you just need to adjust the monitor physical location to whatever the virtual location you would use?
That's easier said than done. I don't know the requirements of the person you're replying to, but it can be things like needing to lie in bed with screen fairly directly above you, or it can be that the physically constants of your space limit the positions that a monitor can be placed in.
I know that historically a lot of their photos that look like renders are real photographs, but I thought the general consensus was that they had almost completely shifted to renders.
> "they clearly don't have any problems creating a physical representation"
I watched a couple of minutes of that video and couldn't see any physical representations of the product. Can you link to the point in the video where the device is shown?
Okay, well at least they've made physical representations, that's something positive. However, my general point still stands, why so much secrecy over a device that's meant to launch this year?
Compare this with the launch of the Oculus Rift, where you had multiple public iterations before the release of the CV1. To me, that's a sign that you had a company that was interested in pushing forward the medium. At the moment there's no clear indication that Magic Leap have developed anything beyond the current market leaders in AR (HoloLens, etc...) in terms of user experience (aside from a smaller form factor).
I'm a bit fed up of the childish negativity that always seems to be sent in Magic Leap's general direction.
Instead of putting energy into this I'd suggest pointing the energy towards creating something yourself.
To the Magic Leap Team - I am a believer :-) especially if you nail the stuff that others don't focus on like smooth interaction models (i.e. voice, gesture, 6DOF mobile controller etc.)
It's less negativity and more caution. People have been burned by early adoption of breathlessly-advertised tech before.
People claim that their products can solve all sorts of interesting problems every day, and it's easy to get excited. But if they cannot demonstrate that their product actually can do what they say it does, then people are going to get suspicious.
Combine the lack of transparency with the long lead times that consumer hardware devices tend to have, and people will start wondering why nobody has been allowed to interact with a live demo of the product when it's supposed to be available to anyone within 9 months.
Everyone wants a product like this to work, but we're also worried about getting taken in by little more than slick marketing; that's how you get companies like Theranos.
I disagree that it's childish. Magic Leap hasn't been clear on what they are actually demonstrating in their demonstration videos. They started out with some really bold claims and have been gradually scaling them back into something more reasonable (and less unique). In my opinion, the skepticism is well earned.
Thanks for the upvotes. Gratifying to see there are others who choose to think positively too. I do understand some of the skepticism, but I think saying:
"The product doesn't exist. The only image of the device on that Creator page is a 3D render"
..about a company that has received over $2billion in funding is pretty childish.
If Magic Leap are watching and want to send the team here at R01 a Creators Edition we'd be really happy to start creating on the LuminSDK with UE4 and represent here in London.
Anyone from @magicleap.com - feel free to use the chat widget at the bottom left of our site to get in touch (http://realityzero.one).
Edit: took Trolls comment off - because I realised, ironically , I was now the one being overly negative. Sorry about that.
Epic Games Founder and CEO Tim Sweeney said: “Magic Leap has truly created something magical in this combination of hardware breakthroughs and software innovations, which represents a giant leap towards the coming spatial computing revolution.”
Tim Sweeney, CEO of Epic Games and head geek for Unreal Engine 4, must have tried Magic Leap One and he doesn't strike me as someone who just goes ahead and drinks the Kool-Aid. I guess we'll all have to wait and see.
Apologies for annoying so many people here. Now I'm getting down-voted into oblivion :-(
The triggering point is just the belief that just because entity X received Y dollars in funding, that X has any legitimacy. There are more than enough counterexamples in the valley (juicero, theranos, bodega, etc) that demonstrate investors really lack the capabilities or wherewithal to conduct proper due diligence. The downvotes you are getting is a direct reaction to this.
And I'm a bit fed up with the fanboyism over vaporware and the lack of perspective while simultaneously accusing those who /have/ perspective as being lazy.
I, for one, /did/ try creating something at that company. Instead all I got was a cut in pay, overwork, and a lack of career prospects while still being required to play roles with rediculous levels of responsibility under working conditions even OSHA would find intolerable.
Mind you, this was all happening while being told that my many years of experience actually shipping consumer products at other companies was worthless and that the failed leadership forklifted from Nokia would save the day as they proceeded to continue making short-sighted decisions and ignoring the experience they hired.
No, I will /happily/ sit back and watch and heckle. And why am I not starting my own thing? Because when they fail to deliver the ill-defined "magic" they so heartily promise, they will destroy an entire field I used to have faith in actually taking off. Funding will dry up and VCs will turn away after this fiasco.
Believe what you want, but keep your accusations of lack of trying to yourself.
Interesting. I'd certainly love to hear more about the inside scoop (but obviously don't say more than you legally can).
What's your take on the current push with the SDK coming out? Do you think this signals that they're really getting ready to ship something, or do you feel that they're still unready based on what you saw there?
I don't think the negativity is childish at all but completely warranted. This product seems to have all the hallmarks of vaporware designed to pump the value of the company through pure hype and zero delivery.
Ah man, the irony of complaining about childishness, insinuating everyone against you is a loser, and then making a new comment just to thank people for upvotes... 20 minutes after posting the first one.
It smacks of when Theranos supporters were shrieking down suggestions of malfeasance.
>>Instead of putting energy into this I'd suggest pointing the energy towards creating something yourself.
attitude. Just because I don't play football doesn't mean I cannot criticize the coach. Just because I'm not in politics doesn't mean I cannot criticize politicians.
Magic Leap appears to be this generations Phantom Console.
Lots of promises, lots of hype, a crap ton of money poured into it... and then basically nothing of substance for years.
Eventually, when it is finally released it'll likely be very underwhelming (so far they really are just showing normal hardware for the space, so they're already there).
It's not surprising people are negative - they've given us nothing to be positive about and lots to criticize. I'm more suspicious of those who are positive about them at this point.
The thing is that if they are real they'll be doing exactly this developer play right about now - a compelling product without any software at release would result in a lot of disappointed early adopters - you can't use this as a proof for (or against) vaporware
It's worth noting that Weta just announced yesterday that they are jumping in with both feet
Well, if it keeps investors aboard. That's what a company desperately scrambling to reach a goal that seems attainable but keeps slipoing out of reach might look like. Theranos comes to mind...
Magic Leap promised a completely goggles-free AR experience. You could load up hundreds of people in a room and show them AR demos. You could be on a beach and watch a communal AR event with thousands of people. You could teach in a classroom without any kinds of headgear or goggles or anything.
Magic Leap is 100% vapourware, even if they release some headset. Magic Leap are liars and cheaters and full of BS vapourwear claims until they release their Magic Leap thing with no headset that gives perfect AR to everyone. Until that day they are liars and cheats.
Their early promotional material sure implied that strongly.
They may have a cool product but they over-hyped the hell out of themselves three years ago and they don't even have physical units to demo yet. That's a massive marketing failure.
I think most of the hype was coming from uninformed bystanders, not Magic Leap themselves.
Magic Leap revealed almost nothing about what they were working on until just a couple months ago; before then all we had to go on were rumors, statements from people who demoed some early prototypes, and a couple incredibly vague teasers from Magic Leap hinting at what they were doing.
If it weren't for the fact that they made headlines after receiving massive amounts of funding from investors (including Google, which makes for great headlines), most of us probably would have never heard of them.
The early marketing showed _nothing_. No images or descriptions of actual hardware, not even a straightforward explanation of what they're actually building.
Again, if you went and interpreted that to mean "we invented a holographic projector" or "we invented glasses-free AR" or "we discovered a way to project forcefields strong enough to suspend whales in midair" that's on you.
Not yet, but the rumored 2k+ price tag should temper most peoples rush to buy one so it's unlikely to be sold out immediately like the Switch for example.
Craziest part of this is it just might be worth it.
I feel like a site shouldn't have to assume that you're familiar with their product to understand what they're trying to push. This single page is useless to actually explain what this is in context, you need to click back to the main site (which is at the bottom of the page) to find out, and even then it's very sparse.
Some years ago I saw (here on HN?) a video of an iPad app that showed kids drawing lines on blank paper, then tooking a picture of the drawings with the app. After that, they would be able to run nice games with those same lines serving as the scenario.
Is it unrealistic to hope for a linux SDK? They have Unity and Unreal SDKs (both of which engines support Linux in my understanding), and their custom OS is based on Android.
The front page should tell us exactly what this is and who/what it's for. Those with knowledge of this industry might be able to jump on board quickly, while everybody else is left scratching their heads.
I disagree. How is the OS supposed to know the difference?
If you want to make a nice animation, how about not including an audio track? Or using the attributes to specify it should be played muted so it won’t cut my audio by default (what YouTube and others do).
What is it that's only allowing you to play audio from a single source at a time? On my desktop that sound just plays over other sounds, nothing pauses. I hate auto-play media as much as the next guy but your real problem is an awkward configuration or oddly limited device.
Can concur. I'm running a pair of D435 under Linux. Still some teething issues, see the SDK and issues at [0], but the quality of the data out of these things is amazing compared to previous generation sensors.
Looking for something comparable to the raw resolution of the iPhone X camera (which is a descendant of the old Kinect camera). Mainly older cameras on the market: https://stimulant.com/depth-sensor-shootout-2/
On iOS, Apple won't let apps access the raw depth sensor data, only a reduced-resolution sample.
I had to go half way down the page to get a hint as to what this software is. I did not find anywhere on that page that describes it for anyone who doesn't already know what it is.
This is an odd way of looking at in my opinion. Apple started making the newton tablet in 1987 and the iPad came to fruition in 2010. The underlying technology for a compelling and consumer-ready VR/AR/<buzzacronym of choice> is much harder problem to solve, it will naturally take longer. I think the software industry sometimes has too short a timescale in its understanding of "bleeding edge".
I'd say it also helped that Apple had already shipped the Apple II and derivatives before 1987 so that they could at least burn some percent of revenue on R&D (if you want to also ignore how they were also struggling from the mid 80's to the 90's even after already shipping an actual product to the market).
>…it will naturally take longer.
Whatever they're working on at this point, maybe it would have been better to stay in academia for a while longer before starting up the hype machine…
But oh well, we shall see eventually, one way or another.
* 6-DOF tracked controller
* Eye tracking
* Multiple depth of field display
* Persistent location tracking across multiple rooms
* Processor/GPU power comparable to "high end mobile hardware"
And software:
* Custom operating system based on Android (but heavily modified)
* Multiple apps can run at once and be displayed in the same physical space
And a few limitations:
* FOV is going to be rather small
* Variable depth-of-field display doesn't work well with objects less than a foot (30 cm) from your face
* Takes a few seconds for the system to adjust when you move real objects around (so if you move your couch, it won't notice instantly)
* The display can't draw darkness or black (virtual objects can't cast shadows on real ones)
[1]: https://creator.magicleap.com/ (Requires login)