How A Real 3D Display Works

There’s a new display technique that’s making the blog rounds, and like anything that seems like its torn from [George Lucas]’ cutting room floor, it’s getting a lot of attention. It’s a device that can display voxels in midair, forming low-resolution three-dimensional patterns without any screen, any fog machine, or any reflective medium. It’s really the closest thing to the projectors in a holodeck we’ve seen yet, leading a few people to ask how it’s done.

This isn’t the first time we’ve seen something like this. A few years ago. a similar 3D display technology was demonstrated that used a green laser to display tens of thousands of voxels in a display medium. The same company used this technology to draw white voxels in air, without a smoke machine or anything else for the laser beam to reflect off of. We couldn’t grasp how this worked at the time, but with a little bit of research we can find the relevant documentation.

A system like this was first published in 2006, built upon earlier work that only displayed pixels on a 2D plane. The device worked by taking an infrared Nd:YAG laser, and focusing the beam to an extremely small point. At that point, the atmosphere heats up enough to turn into plasma and turns into a bright, if temporary, point of light. With the laser pulsing several hundred times a second, a picture can be built up with these small plasma bursts.


Moving a ball of plasma around in 2D space is rather easy; all you need are a few mirrors. To get a third dimension to projected 3D images, a lens mounted on a linear rail moves back and forth changing the focal length of the optics setup. It’s an extremely impressive optical setup, but simple enough to get the jist of.

Having a device that projects images with balls of plasma leads to another question: how safe is this thing? There’s no mention of how powerful the laser used in this device is, but in every picture of this projector, people are wearing goggles. In the videos – one is available below – there is something that is obviously missing once you notice it: sound. This projector is creating tiny balls of expanding air hundreds of times per second. We don’t know what it sounds like – or if you can hear it at all – but a constant buzz would limit its application as an advertising medium.

As with any state-of-the-art project where we kinda know how it works, there’s a good chance someone with experience in optics could put something like this together. A normal green laser pointer in a water medium would be much safer than an IR YAG laser, but other than that the door is wide open for a replication of this project.

Thanks [Sean] for sending this in.

78 thoughts on “How A Real 3D Display Works

  1. There were a few videos from a couple of years ago where they had sound. It’s quite loud.

    Some fun technical stuff: When air gets ionized by a plasma it turns into a very efficient reflector, sending much energy back into the laser. This can damage the laser, but by using a circular polariser at the lens you can filter out the reflected light. Really interesting physics, there.

      1. None. It’s a pulsed laser, so the physics of absorption is *completely* different that the case of CW laser light.

        Two things: First, the ‘circular polarizer’ is a quarter wave plate, which gives you a quarter wave shift in linear polarization, making it circular. This only works if you also use a polarizing beam splitter cube before the QWP. Light reflected back from the plasma changes handedness (direction of circularity), which is then converted back to linear polarization on passing through the QWP, but with the direction of linear polarization rotated 90 degrees. This polarization is then rejected by the polarizing beam splitter.

        See here (page 3):

        Second, the beauty of pulsed laser light is that you only get absorption when the beam is focused and the intensity within the focal volume is above a certain threshold for absorption and/or optical breakdown. The *average’ power of these lasers is typically only a few watts, but the peak power of each pulse can be in the kW, mW, or even TW (tera-watt) range.

        See here (two-photon absorption is one example):

        At the beginning of December, I’ll be heading to a research facility in France to work with a 100Hz beam line, producing 12 femtosecond pulses at 40uJ / pulse. The peak power of these pulses (for a focused spot size of around 8 micron) is on the order of 16 TW.

        Physics rocks. ;)

        1. I agree with everything you said above. We build ultra-fast lasers-

          With peak powers in the 7GW range for our Ti:Sapphire lasers, or 40MW for our Ytterbium fiber lasers (MHz rep-rates), I have seen first hand the effects of these tremendous peak powers. Photo-ionization (as done in this article) and super-continuum generation in water, diamond, sapphire, and gas capillaries are my favorite effects. You can write your name on a diamond with one of these lasers. The Q-switched Nd:YAG used for this article has pulse widths in the 400ps range, so the peak powers are much lower, but the system cost is also much lower than using an ultra-fast laser.

          1. SWEEET… Clark-MXR was one of the first in the field! I used a system from Coherent for my PhD (RegA 9000 system). Good to way to learn how utrafast laser systems work, but a nightmare to use for experiments. ;)

            I’m currently working at TU/e in the same lab that started the Femtoprint project. Right now, we’re studying ultrafast effects in glass under a project called GALATEA – we’re attempting to introduce different polymorphs in silica using focused ultrafast laser radiation.

            We get our laser systems from Amplitude – so mostly fiber, but we have one ‘hybrid’ system that uses a fiber oscillator and a secondary thin-disk regen configuration. Three systems in total: 500fs (T-pulse), ~375fs (Femtoprint prototype), and 150fs (S-pulse, experimental model).

            You can find a little more information about the projects here:

   (commercial spinoff)

            GREAT STUFF!

    1. Marty reaction to it coming down and biting him would probably be more appropriate with this technology considering he’d literally be bombarded with a laser and tiny explosions all around him…

      1. from what i read on what hand phasers are supposed to be they are wavelength specific to proteins. i don’t think the whole trek universe is consistant because i don’t expect klingons and romulans to run around in ships with meat for hulls

  2. Interesting – back when I was an undergrad our laser lab had a fairly large Nd:YAG laser that had some optics which would do something similar. Not intentional, and the particular optics that did this were on the frequency tripled beam (the laser itself came equipped with the original frequency, doubled, and tripled beams that it could output at the same time – really nice setup). But it would leave a tiny floating ‘dot’ in midair every pulse. Not extremely bright, but bright enough to see.

    So, I’ll admit, I’m a bit reluctant to attempt to try this myself – considering the one I saw was with a kW pulsed laser, but it could be that this is refined such that an even more scary laser isn’t needed.

    I also don’t think this technique will work well in water – water’s a better dielectric than air (unless I’m completely mistaken), so the breakdown voltage is a bit higher, so a more powerful laser will be needed. And I’m not sure how well you could display a point (and not a line) with a laser pointer in water… maybe with some type of compound in the water that has a much lower breakdown?

    1. Saw a similar Nd:YAG green pulse laser setup demonstrated almost 30 years ago… yes, the focus lens made each pulse sound like a firecracker. I’m trying to imagine if that would just become a white noise hiss at the millions of pulses per second needed for a display with useful resolution. 30mm, 3000-round per minute 6000fps anti-aircraft guns don’t seem as loud as a .50 cal rifle, oddly enough. Perhaps same would apply here?

      1. Not true: I’ll bet this a hybrid system: fiber stage feeds a seed pulse into a regenerative amplifier (basically, the ‘seed’ pulse bounces around inside a secondary amplifier cavity and gets amplified 20 – 30 times, and is then ‘dumped’ to the output).

    1. Well that was interesting but I almost had to buy a new BS meter, thankfully I unplugged mine just before it smoked. I certainly can’t fault you for not believing it because nothing short of them actually showing a working prototype would sway me either :P

    2. funny how the only actual shot is an egg with a light coming out of it. I like the , lets all wear the lab coat for a the picture part too. gifted and artistic? this one has all the buzzword bingo in one CF.

      I’m building actual lightfield displays, and they don’t look anything like this

        1. “Guys, how do we make this sound convincing?”
          “Use some buzz words and whatever else comes to mind.”
          “How about.. space-time light modulator?”

          At least reading the technical details of this device gave me a good chuckle, but not much else.

    3. I would have given it the slightest chance of being real until I got to “space-time light modulator”… uhh, yeah… I’ve got one of those… it’s called an LED flashlight!

    4. It’s interesting, that the Wikipedia page for their “lead scientist” Vladimir Titar was created at the end of April and has been updated a lot, but pretty much only by a person named “Terry Benedict”. Probably, this is not the Terry Benedict from Ocean’s Eleven ;)
      The only other page this person has edited (and by edited, I mean created) is the page of Oleg Kokhan, who is also supposed to work at Bleen.
      Both pages only exist in English….

      Another contributer that edited both of these pages is LaoTzuRS, who only edited those two pages and the page of Yuri Denisyuk to add things like “3D holography” to the resume. To most of us, this is probably known as regular holography, but the 3D part seemed to be important to (presumably) Bleen.

      The patent they are referring to seems to be about producing “high definition” holograms. A hologram (from my point of view) does not have a lot in common with the device they are claiming to have built.

      Since I was already in the process of trying to find out more about the names used in their campaign, I also searched for Bogdan Shevchuk. The only interesting thing I could find about this name was related to a domain theft. Mila Supinska seems to be an Ukrainian model. Couldn’t find much about her. But their CTO Max Pavlov is even more elusive.

      All of them are apparently Ukrainians who moved to San Francisco for this project.

      Last but not least, it seems a bit contradictory, that “from 2011, Oleg Kokhan holds a position of content producer in Bleen Inc.” if you consider that the name was (according to their website) chosen in 2013.

      Overall, it seems like Scribble 2.0: More CGI, less in touch with reality.
      I’d be happy to be proven wrong though ;)

      1. The “designer” of course speaks French. always more in the cliché. Or at least should I say that she tries to read a text in French. Her accent is horrible and half of the sentences are wrong…

          1. So many money spend on shooting the videos and designing the website. I think they quickly get there money back and disapear next year. Just show me working prototype or be gone. Hate this shit. The had posted device looks way more legit. Can we forbid ppl from the internet making these fake product and give the funding to true inventors and researchers?

      1. It’s a load of bollocks. Essentially the whole document is just repeating itself over and over again, only using the words ‘ laser’ , ‘lenses’ , ‘ fiber’ etc etc. “The outputs of the synchronization unit 38 are respectively connected to the inputs of the laser 1, the control unit 4 high-frequency generator, a pump unit 8, a quantum amplifier 12, the line key elements 231 and 241”
        It is just bullshit.

  3. This was shown at SIGGRAPH in 2006, and yes, it is loud. The sound is similar to what you hear from a good-sized arc welder. In addition to the noise, the crew running the demonstration was decked out in dark safety goggles. They also took a 20-30 minute break between demonstrations. The rumor was the break was to allow the ozone created by the plasma balls time to disperse.

      1. Depends on the laser system – peak pulse power is inversely proportional to repetition rate, and there are often limits to how fast you can push a system to get useful power out of it.

        You may remove the noise, but you’ll also make the voxels very dim (absorption is highly nonlinear in this case), if not impossible to see.

    1. Sounds like dial-up and transformer buzz mixed together. Is the flickering of the image due to the usual camera filming a screen frame rate thing or does it appear like that to the naked eye as well?

  4. “Having a device that projects images with balls of plasma leads to another question: how safe is this thing?”

    Not safe at all. That’s probably why the ‘display’ is fairly high above the viewers. To get plasma that bright at a 1 kHz repetition rate requires some significant pulse energy, probably in the milli-joule range at several watts of average power. I’m curious to know the time duration of the pulses coming out of the laser. My guess is that they’re only picosecond pulses, as these are much ‘easier’ to produce.. ;)

    .. and PLEASE do at least a *little* research on how this stuff works before posting stuff like the following:
    “At that point, the atmosphere heats up enough to turn into plasma and turns into a bright, if temporary, point of light.”

    The physics of what goes on here is WAY cooler than just ‘heating up some air’.. ;)

      1. You got it. You need enough field strength to strip electrons from the gas molecules. In solid materials, you can actually remove so many electrons that you no longer have a ‘proper’ charge balance, causing the molecules/atoms to repel each other (too much positive charge) and literally rip the material apart (on a micro-scale).

        Theoretically, if you could get the efficiencies of fiber lasers (well, lasers in general) high enough and give them enough portable power, you probably *could* make something akin to a light saber using the same (or similar) focusing technology used for this display device. You’d have to scan / refocus the beam extremely quickly and dump tons of power into it as well. I’m not sure you’d have enough plasma to do any real damage though. At that point, you’d be better off to just use the focused laser itself to cut things. ;)

  5. I never trust science “news” that I see on the internet until hackaday features it, because they are the only outlet that includes comprehensive technical details. I’ve been waiting for days for this to show up on hackaday so i knew it was real and how it worked.

  6. I don’t think water would work, it would require an even more powerful, but I could see a way of making this safer and possibly by having a lower-power laser create voxels in sealed projection area filled with low-pressure neon or the neon-argon mix used in plasma balls.

    1. Argggh, that’s kindof what I was thinking. Some gas/mixture with a low energy quantum transition in the shell. Mixture could use different gases excitable by different wavelengths of lasers.

  7. My prediction: the first practical, *ultra high* quality “hologram” will be a device that tracks the user(s)’ eyes, and projects the image directly into the eyes. I can’t say with what technology, but somehow directly stimulating the cones and rods in the eyes. With multiple “projectors”, you could have a limitless viewing angle. Augmented reality would be trivial with a working device.

    There are a lot of hurdles with this approach (eyes constantly changing focus, head/eye movement, etc), but it doesn’t really on wearable hardware, there’s no requirement for a display medium, works over a large area, easily can be scaled up for multiple users, and is potentially safer — so long as you remain within the energy levels the eyes are used to dealing with.

    The only real downside would be advertising — some non-existant assholl trying to sell you crap as you walk down the sidewalk.

      1. Wow, there is so much wrong with this comment… where to begin?
        1. There is nothing “retinal” about the RTI displays used in the VB. They consist of a single column of LEDs (using LED arrays built for use in some printers in place of a laser, actually; yes, the VB is based on a hack!), a mirror galvanometer, and a focusing lens. With clever timing, the column becomes a 2D plane due to POV.
        2. Where did you get your “most people” figure? The VB has exactly the same capacity to induce headaches/eyestrain as any other artificial stereoscopic display. Last time I checked, millions of people go to movie theaters every day and use very similar 3D display technology with no ill effects. The VB’s problem was negative publicity spread by unthinking fear-mongers. I count myself among the many, MANY people who have never had a problem with stereoscopic displays, and who enjoy the VB’s unique experience and game library.

  8. Any way this could do a color display? Good proof of concept. Now it needs 2x the resolution in all directions and a faster scanning rate. I expect it looks better in person due to POV. POV things never look as good on video as they do to the naked eye.

  9. Ozone source? OK for use outside I guess but in an enclosed area perhaps not such a good idea.

    You could never directly interact with the image either, without getting burnt fingers.

  10. They claim no input gas is needed since it’s just re-purposing the air but isn’t that the input gas? I can’t imagine you can get a large display out of it without cycling a lot of air through a room. Would make for nice drive-in tech though, just might revive the industry.

    1. Technically you could scan a really powerful laser and produce electron-positron pairs in a vacuum, no gas needed. The subsequent anihilation should produce some visible light (and plenty of xrays too). But laser technology is not there yet :-)

  11. Really cool technology, but I feel the goal is silly. In the video they state that they want to use this to roam around in a car and display emergency text. Why would you spend all this money and energy to display emergency text, when something like a led panel would be cheaper and more reliable?

    Still really cool technology though.

  12. Re. colour.
    1) Use argon/xenon mix in a chamber, by tuning pulse energy yes you could do multicolour and energy would be about 5% of that used now
    2) for open air displays you would need a hood that recycles the released gas back into the system using powerful vacuum pumps.

    The cost shouldn’t be that high, and with dynamic mode switching the resolution would be about the same as one of those “Jumbotron” screens but spread out over a 3D volume of tens of square metres.


Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.