A new video of Magic Leap’s augmented reality tech in action was shown at a conference in China today.

On stage at the event, Magic Leap CMO Brian Wallace had a one-on-one discussion with Alibaba CMO Chris Tung (Alibaba is one of Magic Leap’s major investors). Brian said that the company is working on an AR headset that’s portable, lightweight, and can be worn all day, and further said that the aim is to someday replace smartphones completely.

brian-wallace-agic-leap

Part way through the discussion Wallace introduced a new video showing off the experience of Magic Leap’s mysterious augmented reality tech. A video of the full discussion was published by Tech.qq.

Photo courtesy Mark Zuckerberg
See Also: Facebook CEO – ‘In Five to Ten Years AR Will Be Where VR is Today’

While prior Magic Leap videos have focused on entertainment and productivity use-cases, the new video’s concept involved a pregnant mother shopping for new items in her baby room. This is the first video from the company to show voice control as a method of input.

Using voice commands, the user in the video measures the available space and is then presented with several options which will fit. She’s able to preview the products such that they’re convincingly inserted into the real world. After trying out a cloud lamp and giraffe lamp, she opts to buy the latter, and we can see the lamp animate to confirm the purchase.

SEE ALSO
Magic Leap One Controller Appears in FCC Filing, Release on Track for 2018

While Magic Leap’s video demonstrations are impressive, they’re also shrouded in mystery. The company hasn’t publicly shown the device that’s enabling these videos.

Prior Magic Leap videos, which the company says were filmed through their AR technology with no special effects, show a robot and a solar system in an office, an augmented reality workstation interface, and life-sized Star Wars characters.

This article may contain affiliate links. If you click an affiliate link and buy a product we may receive a small commission which helps support the publication. See here for more information.


  • Steve Biegun

    Well this is… cryptic.

  • Opto

    Any guesses whether the user’s eye or the system is zooming-in on the lamp?

  • Kevin Williams

    the rooms are getting darker in each video seen now!

    • darkflame

      Ideally, we need a way to selectively remove photons from getting to the eye, rather then just adding.
      MagicLeap is tackling the focus issue though, trying to reproduce a full light-field, so I think it will be probably too much of a challenge to that on that as well.

      Best we can hope for, I think, is a pair of AR specs with a single variable opacity across the whole back.

      • I believe a big selling point on Microsoft’s HoloLens is the black masking they do to the background before painting the new image on top, so they don’t need to overblow the light to make the AR image appear.

      • Jack H

        There have been several attempts to selectively filter the ambient scene which I have come across for use in optical see-through head-mounted displays.
        The first method is to use an inverse filter of the image captured by the camera to cancel out the relative strengths of different colours. This is performed solely in graphics and again I state it is only relative not absolute cancellation and even then only of colour balance not of intensity (and results in an overall increase in intensity).
        The second method is an imaging method, whereby the ambient light is focussed onto a “spatial light modulator” such as an LCD, LCOS or DMD which selectively filters out some of the light before it is brought through a conjugate lens system and delivered to the eye. This has been demonstrated in one or two universities and patented by Microsoft (using LCDs) before Magic Leap decided that replacing the discrete lenses and mirrors with a waveguide counts as a new invention.
        The third method relies on the filtering of the rays while they maintain their vectors. This needs at least two (and preferrably at least three) spatial light modulator layers to be useful. In one instance nVidia Research and UNC Chapel Hill (I think) developed the “Pinlight Display” where several small LCD layers were driven by time-varying patterns of vectored occlusion. This is a similar idea to the MIT (or was it Stanford?) “Tensor Display”.

        I am developing a method using the ray vectored occlusion of ambient light, which should it have satisfactory performance parameters, result in both the additive colour and the occlusion being driven by a single off-axis display and waveguide, unlike the nVidia/ UNC method requiring the the use of three displays

        Kind regards,
        Jack H.
        Halo AR

        • darkflame

          Wow, thanks for that “state of the art” update. A lot I was unaware of, despite following AR developments passionately.
          Looks like I have a fair bit of reading to do.

          Lets me know if theres any place to follow your developments on this.

          • Jack H

            No probs little homie :*
            I’ll try to be more active on Twitter (@Halo_AR_ltd), particularly some good old fashioned smacktalk on rival headsets.
            PM there and I can take your email for future mailing list.

            Stay legit xoxo

          • Jack H

            PM me on twitter (Halo_AR_ltd) and I’ll keep your email for a future mailing list

  • ImperialDynamics

    HoloLens can already do that. By the time Magic Leap’s product is officially announced HoloLens will have reached second gen and become more affordable. Magic Leap is too late to the mixed reality party.