Looking through their most recently filed patents, it looks like their main innovation is a HMD device that uses a combination of a high speed digital zone plate, a masking device, and a traditional imaging device.
For each eye, there is an imaging device (720p/1080p resolution for example) that displays an image. In front of the imaging devices there are high speed mask device (LCD?) that blocks off portions of the image that are not at the current focal depth. The focal depth cycles quickly between 12 different values and easily driven by the z-buffer values from the renders. The light not blocked by the mask is partially collimated and guided towards the eye via optics and then a high resolution high speed digital zone plate is used to refocus the light at different depth levels using diffraction. The mask and zone plate update together at 12X (360Hz-720Hz) the rate of the image device (30-60Hz). The result is that pixels that should be at different depths are actually focused near that depth and the eye blends the 12 partial frames together to form one image that has a range of depth in it.
I expect they are using off the shelf parts for the imaging device (and maybe mask LCD) but manufacturing their own digital zone plates since they need super high resolution and high speed for those (at least 360hz) but only on-off values not color or grayscale. This is probably where a bit of their money is going.
If it works, you could get high resolution images per eye, and also see 12 buckets of depth so your eye would refocus to see things close up versus far away. This could potentially create a pretty good impression of something existing close to you, compared to existing AR where all objects are focused at a single fixed depth and there is a conflict between accommodation and stereopis cues.
Of course this is one of the patents filed, it could be just an idea they decided to file on, not what they are actually building.
In an interview with the South Florida Business Journal in February, Abovitz
said “When you see this, you will see that this is computing for the next 30 or 40
years. To go farther and deeper than we’re going, you would be changing what
it means to be human.”
As exciting as all this is, this kind of language is giving me flashbacks to the buildup to the launch of the Segway.
This definitely seems to have the makings of Segway 2.0. One can only hope this yields better results, but I'm extremely skeptical. I can't imagine this in a form factor that would be comfortable for the average user.
Segway still has a transformative future potential, though. Once they're completely autonav, along with all the rest of transportation, they become an important part of the overall automated system.
> In front of the imaging devices there are high speed mask device (LCD?) that blocks off portions of the image that are not at the current focal depth.
It would be interesting to see how this works in practice. The eye can't focus on something that close. For an example, imagine a speck of dirt on your glasses - it shows up as a blob in your vision, not as a dot.
The eye is not focusing on that depth. The light coming out of the lens or LCD would be collimated / diffracted to form an image on the retina that appears to originate from a specific distance (or in this case, many different distances, multiplexed together).
Just like with Google Glass, where the image appears to be at a distance and your eyes do not focus on the actual screen.
Fascinating. I've been pondering this problem for quite a while and this is an entirely unexpected solution. Such a system can potentially be very small because there's no refractive optics.
For me, the biggest headline out of this not being mentioned is that half a billion in VC was just raised in Hollywood, FL.
I'm building a startup community in Myrtle Beach, SC and Paul Reynolds is one of the founders of Startup.SC with me. Paul is involved in Magic Leap as a director, and after working remotely, just moved there last month. It's very exciting to see unfold!
It drives home the fact that a startup can make such headlines not only outside Silicon Valley, but even outside other well known northeast tech metros. It's the sort of news I feed on to explain scalable startup growth potential: Magic Leap has done a spectacular job attracting talent from all over the country. There are worse places to be than the Fort Lauderdale, FL area.
I watched the revival of the Boston startup scene while involved with TechStars, DogPatch labs, etc. from 2009-2012 and although we certainly don't have the same resources, I'm doing my best to apply those principles to the Myrtle Beach market.
I hope to one day have a startup come through Startup.SC which can also demonstrate the ability to attract talent and investment dollars to a coastal city not known for tech, and demonstrate that a company can thrive!
Yep, Magic Leap is headquartered in Florida, but they didn't get the investment until they opened an office in Palo Alto:
http://recode.net/2014/10/20/look-who-else-is-joining-google...
"People familiar with the company say Magic Leap was having trouble convincing Silicon Valley types to drop everything and move to Hollywood, Fla., where it is based... So it recently opened a Palo Alto, Calif., engineering office..."
The primary bottleneck among most top-tier VCs is partner time and travelling is a huge time-sink. Hence why most VCs tend to focus over a relatively small areas.
If a VC is looking at investing in a startup in a location where they have no activity then they're going to have to do trips out just for that specific startup (for due diligence, board meetings, etc.) and they're also going to have a harder time providing value-add (lack of local contacts, etc.)
Hence for a startup outside of a cluster to be worthwhile it has to be significantly better than the local dealflow and the deal has to be of a significant size. So such deals are always likely to be a rarity.
(Having a founder who's already built and sold a billion dollar company also helps immensely; if you're in that small category of people you can probably raise money regardless of where you're based)
I look forward to either disproving this, or being defeated by it. My stance is that great connectivity and communication (think google hangout) combined with cheap flights between Myrtle Beach and Boston can make it happen. Investors may be enticed by additional reasons to head to Myrtle Beach: golf, vacation, warmth.
VCs (largely) want solid personal relationships with their companies and that's something that's not happening over skype/hangout anytime soon. A few thousand dollars on flights also isn't a big deal when you're investing millions; time is much more important than money.
Your best bet would probably be to:
(1) Get solid local VCs who other (non-local) VCs are happy to co-invest with.
(2) Build thematic clusters, because a partner at a VC who's interested in multiple companies in an area is much more likely to fly out then for an individual company.
I think #2 is the salient point there and displaces #1.
Creating a "groudbreaking, path forging technology and an amazing team" is the outlier, not raising a half a billion in funding. If you can achieve the former, then your odds of the latter go up by many orders of magnitude.
All I remember is that Webvan attracted ridiculous investment as well and look how well that turned out. Google can waste billions on utter speculation and never notice the blip. If you can attract such enormous funds why not, the worst that happens is you fail.
For me, the main story is that there is still money thanks to people like Google and AH to go after the big bold crazy technology dream. Kudos to all Rony, Google and AH for the courage and vision.
Startup.SC is really exciting -- I live in the Charleston area (but work for SV companies remotely) and hadn't heard of them before.
The quality of developers and ideas coming out of this town hasn't been that great, or (as I hope) I am having a terrible time finding the right community.
You should come visit our community and mingle with our members at an event: http://meetup.com/Startup-SC
Or reach out to me on twitter: @MikeSchroll @StartupdotSC
We have some top-notch devs working out of our space: Ex: Two remote workers for Quirky.com.
I too am in a secondary market so I want to believe ;) But it'll take more than this to really prove anything. It seems like every 18 - 36 months there's this sort of "exception" to the rule vs indicating any type of sustainable change in where the majority of dollars and interest flows. This feels very much like that.
I live in the area and we have a lot of difficulty finding talent. I see they are hiring for a lot of very specialized positions and I wonder how difficult it will be to fill them in this area.
Its going to be very difficult for them to hire such specialized skill sets ANYWHERE. They need to find very high end optics and microprocessor experience, and thats tough to find either in palo alto or hollywood, fl
@logicX would love tot alk to your friend Paul. I run RefreshMiami, the largest startup community in South Florida, we'd love to get someone from MagicLeap to give a talk.
> It’s rare that a company can stay relatively secretive while raising a huge amount of funding, but Florida’s Magic Leap has managed that.
Only rare because journalists are extremely lazy by default, and need their stories pre-drafted and handed to them on a silver platter, which companies tend to do. Avoiding press is very easy, just don't do the journalists' job for them. :)
> Only rare because journalists are extremely lazy by default, and need their stories pre-drafted and handed to them on a silver platter
That doesn't follow. If that is true, then it would be very easy for a company to be able to stay relatively secretive while raising a huge amount of funding; with lazy journalists as you describe, a company could raise funding by networking and so long as no one involved issued press releases, they would succeed in being secretive.
"journalists are extremely lazy by default, and need their stories pre-drafted and handed to them"
This seems to be a consensus opinion on HN. I have no idea if it's a majority opinion, but it's fairly typical. The "journalists are lazy/stupid/inept" argument generally misunderstands what journalists do, how they work, and how they pick (or, as is more often the case, how they're assigned) their coverage. If anything, it gives some journalists too much credit, and for the wrong things.
I'm not a professional journalist, though I've written frequently for some major and minor publications. My typical process looks like this:
1. Pitch a bunch of story ideas to my editor.
2. Have 99% of those ideas rejected for various reasons: too narrow; too wonky; not enough meat on the bone [1]; not timely enough [2].
3. Settle on a topic that is somehow deemed acceptably timely, broad, approachable to general audiences, and well substantiated by some degree of data or prior coverage.
The astute reader will start to see how objections [1] and [2] come into direct conflict. By definition, breaking something new means covering ground that hasn't been covered before (or at least not often, or not in depth). But having prior substantiation often means being able to point to pre-existing coverage, as proof that your topic is of some interest to the public. [3] Also, you have to sort this all out on a deadline: anything from the end of the day to the end of the week, per story. (As an exercise: try investigating, substantiating, and beautifully composing an in-depth story in a single working day).
[3] The public, as typically defined, is a very broad swath of readers. Even major tech sites, like TechCrunch, now have a very broad readership -- with highly varying degrees of technical sophistication, subject-matter expertise, and topical interests.
On the flip side, look at any of the blogs or publications that HN readers respect. Chances are, they write in greater depth. They are not afraid to get into the wonky, technical details of their coverage. They don't just give you charts and infographics; they give you their statistical reasoning. Niche publications can do this.
Broad publications theoretically can, but often choose not to, out of the belief that the effort/payoff ratio just isn't there. It's a lot of effort that will appeal to only a sliver of their readerships. Deep coverage, or investigative coverage, or breaking new ground, requires a shitload of hard work. It is extremely tough on a standard deadline. It often requires a dedicated budget, and a broadly extended deadline, just to ensure that the writer(s) can afford to develop it.
Bottom line: I don't disagree with you. Most tech journalism sucks. That said, I don't blame the writers per se. They are the proximate end of a very long chain. Maybe some of them are capable of writing the stories we'd want to read. Maybe not. But they don't get the chance in the first place.
Before we impugn the entire field of journalism, let's be clear which publications, and which branches of journalism, we're impugning. It's silly to paint the whole field with one brush. I'm not saying the ratio of wheat to chaff is a good one, in the aggregate. But there is definitely wheat to be found. For every two dozen tossed-out, deadline-harried stories you find, I can point you to one or two amazing pieces of long-form writing. Sometimes the former and the latter are produced by the same writer. It all depends on the publication, its business objectives, and its beliefs about the demands of its average audience. The stuff you read in TechCrunch, for example, is meant to be consumed and immediately understood by almost everyone: you, your non-technical friend, and even your Baby Boomer parents. There's a certain type of material that fits that broad audience's tastes. Chances are, it's not the type of material that moves readers like us.
Journalist here. This man speaks truth. Pitching is 99% of the job, and I've lost count of the number of major stories my editors turned down, even when I had those stories as much as 3 years before major outlets. This is exactly how it works.
:) Funny thing I noticed, going through the job descriptions - Physical Requirements for HR Generalist[1]:
> While performing the duties of this job, the employee is regularly required to sit;use hands to finger, handle, or feel and talk or hear. The employee is frequently required to reach with hands and arms. The employee is occasionally required to stand;walk;climb or balance;stoop, kneel, crouch, or crawl and taste or smell.
> The employee must regularly lift and/or move up to 10 pounds, frequently lift and/or move up to 25 pounds and occasionally lift and/or move up to 50 pounds.
> Specific vision abilities required by this job include close vision, distance vision, color vision, peripheral vision, depth perception and ability to adjust focus.
I did a web search on those sentences and they appear thousands of times in job descriptions, e.g. 133,000 results for the "Specific vision...focus" sentence. The "taste and smell" seems to appear mostly in food service jobs, where it makes sense.
These seem to be some sort of standard conditions that many jobs cut-and-paste from somewhere. I wonder what the original source is. But they seem very out of place for an HR position.
I guess I just haven't seen one of these (such specific or probably any Physical Requirements for a job) before and imaging all of it together, there was this comic vision, in my mind, of what will that person actually do on such a job. :)
I definitely didn't want to offend anyone and I hope I didn't.
It's a bit strange how the CEO claims it's not AR while it looks completely like it.
> since it goes well beyond that and provides truly integrated, 3D digital objects that looks as though they were physical objects, alongside the real world
The Ingress game is perhaps the most famous example of a company using "Augmented Reality" to mean something substantially more limited than Magic Leap intends.
The consensus seems to be that the "cinematic reality not augmented reality" message probably refers to the realistic accommodation - virtual objects being in focus at the correct focal depth - which Magic Leap is apparently aiming for and which other 3D display systems don't offer. If that message seems confusing, especially in light of the fact that neither 2D nor (standard) 3D cinema offer realistic accommodation, let's just say that Magic Leap's PR operation has not always been the tightest ship: https://www.youtube.com/watch?v=w8J5BWL8oJY .
The language used, including the brand name, make it sound like it's a leap ahead of the competition. It has our attention especially now with this large raise. Hoping it's truly a decent improvement.
As I said when there was speculation into an investment this size by Google, etc into Magic Leap:
An investment this size isn't designed to get "VC-size" (10x) returns. It is designed to block competitors getting access to the technology before Google.
The 10x factor generally only applies to early stage investments, VCs making later stage investments tend to need smaller returns because there's less risk involved.
"Magic Leap and Weta Workshop are collaborating on a truly next-generation Dr. Grordbort’s first person shooter on a world-changing new platform in an effort to defend Earth from robotic overthrow"
I hope they're OK with remote work, and that they're open to paying more than the typical IT wage in Florida, because I can't imagine their talent pool will be substantial otherwise, being in Hollywood, and all.
I grew up in Fort Lauderdale, I worked there for a while after I finished college and own a house there. (Now in Bay Area) It's not just the wages they will have issues with: even if they pay SV rates the talent just is not there. Maybe they go after some Disney people however aside from a limited pool up in the Orlando Area, I'm pretty sure they are going to need to import a big chunk of the talent and that gets expensive.
This struck me too. The secrecy probably excludes remote work, so . . . they're going to have to pay massively to get the experience they apparently need down there. This is a rare case where relocating to SV may actually be a cost reduction measure.
I don't know. They could pay a 20-30% premium over the typical IT wage there and still be far below SV averages, plus reap the benefits of Florida's tax incentives for entertainment industry companies (at least they used to have a massive tax break for such companies a couple of years ago, before I fled to the Bay Area--I doubled my salary, but my real cost of living only went up 30% or so, and the best part is I'm not working on some web/mobile app in a sweatshop/startup using fad-driven-development on a macbook).
Maybe they're hoping to poach from the DoD/simulations and entertainment video game industries that form the backbone of the IT industry in that area, but even in that case I'm skeptical they're going to find much.
The bigger problem, from my point of view, is basically one of culture. Most of florida is extremely conservative, focused on tourism-based entertainment, and (my opinion) neither very family friendly (for families who live there--the big tourist parks' existence doesn't make it a family friendly place) nor a good place for the younger Bay Area startup (stereo)types in terms of cultural and social attractions.
I don't know. I can see it going both ways. Honestly, how many places sound this interesting to work at? Feels like Steve Jobs famous line about you can keep on making sugar water or you can change the world with us. It's hard for SV people to leave the Kool Aid but it wouldn't be too hard to pull anyone from a cold part of the US and/or already on the East Coast.
This company isn't targeting the SV "Kool Aid" crowd for engineering on the main product. Perhaps their web/mobile apps or other secondary products might target that crowd, but their core product is one that requires hardware and software teams working together, with the software teams implementing systems (applications or even lower level) programming, not fad driven JS-client-on-Ruby-backend web/mobile "apps".
There just isn't enough such talent especially in Hollywood for them to attract it locally. By the same token, the tech industry there in general isn't strong enough to be attractive to someone coming from a tech-rich area like the Bay (or even Chicago or New York). I would not move to Florida for work without some significant salary and stability guarantees, because in the event the company goes belly up, relocates to a place I couldn't/wouldn't follow, or terminates my employment the quality of the "fallback" jobs is atrocious (especially in comparison to the work apparently being done at this place).
"SV Kool Aid" crowd are people who will never leave the Bay Area because they believe it's the only place to be. They wouldn't move anywhere else - even a major secondary hub on the West or East Coast.
I also understand that you personally wouldn't take the risk and move. I just think that's a small sample size.
Their technology trademark is 'Dynamic Digitized Lightfield Signal'. A lightfield camera has multiple lenses to record multiple focal points simultaneously. Have they found a way to reverse this concept, 'project' to a wearable an eye refocusable image? Their list of job openings leads to a lot of possibilities.
There's no theoretical reason this wouldn't be possible - there's been a lot of very cool work on lightfield technology. The best I remember was a display where by adding multiple LCD layers (3 I think) and calculating a high frequency distortion signal, you could achieve wide-field of view 3D TV, glasses-free, with pre-rendered images.
There's other work to do the same thing to the retina - throw a lightfield at it which simulates the proper wavefronts to focus a 3D image at a short distance.
Of course this is all kind of begging the question of what this company is actually doing: for example, the big problem not answered is where they get the portable processing power to make this work, and what the actual resolution is? The Oculus Rift already tells us 1080p is not enough and even 1440p will be imperfect.
Of course you could try to end run around the problem: I've been curious what you could do with 3-color lasers shone through an interferometer setup mounted on piezoelectrics, so you could add or subtract interference peaks by advancing the mirrors a few nanometers and then scanning the output over the eye (or, as one company is doing - using a DLP mirror chip). Focus that laser beam small enough, and you'd be addressing individual rod and cone receptors in the retina, granting infinite focus to any projected image.
EDIT: Looking at that jobs page, a laser and fiber optic engineer being top position seems very intriguing.
But the Rift has a very high FOV compared to what many AR applications require. If you only need to draw a relatively small AR object instead of drawing the whole FOV, 1080p goes a long way.
Kinda like the 1080p laptop screen in front of me is cleary enough, but stretching that to the whole FOV decreases the pixel density too much.
Very interesting, I looked up the CEO to see if he had a Disney/Imagineering background, and was surprised to see I was familiar with his previous work, which was founding former public company MAKO Surgical (which ended up being acquired by Stryker). MAKO's product used a robot/surgeon hybrid + software for visualizing/planning surgeries for implanting joint replacements. Interesting career twist (edit: but one that makes a lot of sense)
edit: @glxc, completely agree, I just didn't know a lot about his background prior to seeing this news. Interesting guy.
As others have pointed out, it's unlikely that the company needs the money and this is more about denying competitors the technology before Google gets their hands on it.
Right, I get that. That's why the company wants the capital. It's impossible for any young company to need this sum of money at once. But it doesn't justify the market giving it to them.
The ultimate effect is the entire market got denied a substantial amount of capital (not just ML's competitors) in order for a single entity to gamble it.
That's irresponsible of everyone, founders and investors included. Either acquire, go public, or JV to secure an opportunity. Why sap such a huge swathe of capital from the opportunistic ecosystem?
They are making hardware; hardware is expensive, and a lot of money can significantly improve time-to-market. I doubt they will be spending this on Aeron chairs.
I don't think that's a good example. For all the angst around them online, they both have their redeeming qualities if you actually interact with either.
I know people who work on Wall Street and I know people who work in the White House. They are some of the most brilliant people I know.
That doesn't make the systems they collectively run any less corrupt or terrible than they are.
You originally said:
> These are very intelligent folks we're talking about, I'm sure there is a rational explanation.
If the examples I mentioned aren't good, then logically, there may not be a rational explanation.
But in any case, I think all 3 situations are quite similar. VCs are probably just as smart as the top people in finance and politics, yet they often appear stupid from the outside. It's actually quite intriguing. If I were a researcher, I'd definitely want to investigate this further...
"artificial, but extremely realistic images .. projected directly onto a user’s retina" launching without predecessors as a commercial product "fairly soon" would be quite a surprise.
They have an app on the App Store called Hour Blue, which they had run as a demo at Comic Con a few years back, in conjunction with string labs. It doesn't seem to work for me in the iPhone 6 though. You watch a video of the demo if you search on YouTube however. The demo was a basic augmented reality example. I'm assuming, with an investment of this size, they have achieved something far more substantial.
AR is okay, but it will probably be gimmicky for a long time before it's useful. VR is the closest thing we have to the Holodeck, and it's almost here. I thought Google loved Star Trek. It shouldn't be ignoring VR.
Did anybody notice the typo on the company page? "It is biomimetic, meaning it respects how we are function naturally as humans (we are humans after all, not machines)."
From the gif, I gather they are doing AR without fiduciary markers. Since the gif shows a hand that is moving while the elephant remains in place, they might be doing 3d surface construction (possibly in real-time). How off am I?
For each eye, there is an imaging device (720p/1080p resolution for example) that displays an image. In front of the imaging devices there are high speed mask device (LCD?) that blocks off portions of the image that are not at the current focal depth. The focal depth cycles quickly between 12 different values and easily driven by the z-buffer values from the renders. The light not blocked by the mask is partially collimated and guided towards the eye via optics and then a high resolution high speed digital zone plate is used to refocus the light at different depth levels using diffraction. The mask and zone plate update together at 12X (360Hz-720Hz) the rate of the image device (30-60Hz). The result is that pixels that should be at different depths are actually focused near that depth and the eye blends the 12 partial frames together to form one image that has a range of depth in it.
I expect they are using off the shelf parts for the imaging device (and maybe mask LCD) but manufacturing their own digital zone plates since they need super high resolution and high speed for those (at least 360hz) but only on-off values not color or grayscale. This is probably where a bit of their money is going.
If it works, you could get high resolution images per eye, and also see 12 buckets of depth so your eye would refocus to see things close up versus far away. This could potentially create a pretty good impression of something existing close to you, compared to existing AR where all objects are focused at a single fixed depth and there is a conflict between accommodation and stereopis cues.
Of course this is one of the patents filed, it could be just an idea they decided to file on, not what they are actually building.