Breaking, Business, GPU Technology Conference, Hardware, News, startups, Virtual Reality (VR)

Camera Maker Lucid VR Receives $2 million Seed Round

Raising capital is never easy, even if the idea you have might be a revolutionary one. If your idea is a piece of hardware, raising capital in a software-driven industry is extremely hard, but not impossible. Case in point is Lucid VR, a tiny startup launched by a Berkeley and Stanford graduates.

During the GPU Technology Conference, we met with Han Jin and Joe Hill from Lucid VR. Han is a co-founder and CEO of Lucid VR while Joe is a chief hustler, i.e. company evangelist. Described as “world’s only 3D 180-degree Virtual Reality camera”, LucidCam is a combination of innovative hardware design and software that enables instant watching of VR Content using Apple iOS and Google Android.

Their quest for funding was more difficult than software-only developers, as software startups will typically ask for $.5-1.0 million in initial funding, while Lucid VR had to complete a $2 million seed run in order to fulfill the tooling, start manufacturing etc.

Lucid VR LucidCam overview

Just a week after the finish of the GPU Technology Conference, Lucid VR announced that they received $2.1 million in Seed round lead by Wistron, S2 Capital, Lab360, TEEC Angel Fund, 17 Miles Technology and a number of private angel investors. The key investor for the company is a manufacturing agreement with Wistron, one of world’s largest ODM (Original Design Manufacturer) that will be in charge of bringing the camera concept to life.

“LucidCam brings people together across the world by capturing incredible immersive experiences in VR, the same way as your eyes see them, and which can be easily shared with friends and family within seconds,” said Han Jin, co-founder and CEO, Lucid VR. “With our relationship with Wistron we will be developing a beautifully designed and engineered camera that is going to surpass our original plans. We are very excited to announce that we will be upgrading our backers with a first production version of the LucidCam with double the resolution and frame rate, in addition to extended battery capacity. The viewing experience in VR headsets will be so much smoother with 2K and 60 frames per second.”

This is second round of investments, after the company successfully completed its Kickstarter campaign, which slowly but certainly, is becoming a good path to bringing hardware products to life, and ultimately receiving conventional seed / Series A / B / C funding. Time will tell will for Series A, Lucid VR turn to crowdsourcing through platforms such as SeedInvest. After all, in a crowdfunded Series A, VR platform startup Virtuix attracted over $4 million.

  • gc9

    Is the evangelist using ‘virtual-reality’ here just as a buzzword to attract attention?
    The only connection to virtual-reality is that a stereoscopic head-mounted display could be used to view the recorded video. It records actual reality, not virtual reality, and does not create a virtual reality model.

    This looks like a camera for capturing wide-angle 3-D stereoscopic video, using a single interpupillary distance. They could make the peripheral view even wider if the cameras were mounted on a curved face rather than a flat plane, so the cameras weren’t in each other’s view. This looks useful for capturing first-person sports video, as long as the player’s head motion isn’t too violent (else viewers may get hand-held video nausea). But VR?

    In 3-D stereoscopic video, the camera operator controls the view, and you the viewer watch passively. You can move your eyes to look around, but if you your head then the inter-pupillary distance will be wrong, producing unexpected eye perspectives for nearby items, and you might feel sick. This camera has a non-adjustable inter-pupillary distance, so the result only looks real if your own inter-pupillary distance is similar to the one in the camera, otherwise things can appear too small or too big, like toys or models. (Maybe you, the viewer, can make your effective inter-pupillary distance narrower by looking off to the side rather than straight ahead, but you can’t make it wider.)

    In Virtual-Reality, you, the viewer, can move around, turn your head, look down into a box, sideways under the desk, or up into a hole in the ceiling, etc., and the virtual reality sensors track your movement and adjust the view. Virtual reality requires a computer model to generate the appropriate view for each eye depending on the position of each eye in space, which depends on your inter-pupillary distance, and changes as you turn your head and move around. To generate different views, it needs to have the distance to each surface so it can calculate which parts are occluded in one eye and exposed in the other.

    A two lens camera doesn’t record enough information to create this virtual reality model in one shot. Can a 3-D stereoscopic camera be used to create a virtual reality model from multiple shots? Maybe if the scene has only static objects, like buildings or bridges, and the viewer keeps their eyes level. If the camera lenses are as far apart as
    the largest human inter-pupillary distance, then it may record all the surfaces visible from just one eye at any inter-pupillary distance around an edge or corner (but not into a grate or keyhole). But when looking up or down (say up at the bottom of a bridge or chandelier or stairwell, or down into manhole or box containing more items), the camera-operator’s eyes are aligned along one compass angle, and the viewer’s eyes could be aligned in a different angle, so the two camera lenses will not capture all the surfaces visible to the viewer’s eyes at a different angle. For static scenes, the two-lens camera could be rotated on multiple axes to capture views from multiple angles, but this won’t work for dynamic scenes. Microsoft Research has software for inferring a model from camera views, but don’t see anything that says this start-up is doing this.

    So it looks like VR is being used as a buzzword to attract attention, but there is no VR model involved. The consumer-priced camera records wide-angle stereoscopic 3-D video, which happens to be viewable on a stereoscopic head-mounted display.

    • As far as the company showed me, the goal of the product is to sell the parts which were successfully launched on a crowdfunding website, and then to increase the capability of their cameras with a professional mode. The goal is to create a recording in which you could move around. They’re a startup and they have a long way ahead, but I’ll let Lucid guys to give their opinion of your comment. Thanks for the insight.