Mel Slater’s Theory of Presence includes two major components including the Place Illusion (the degree to which you’re transported to another world), and the Plausibility Illusion (the degree to which you believe what’s happening in that world is believable). Presence researcher Richard Skarbez thinks of these two components as “immersion” and “coherence,” and we previously discussed his presence research where he found that both illusions are vital for a deep sense of presence. The place illusion is largely enabled by the objective details of the VR technology with features such as 1:1 head tracking, low latency, and large field of view. But all of the different dimensions of what makes an experience plausible are still widely unknown, not very well researched, and also difficult to isolate and determine.

Skarbez was back at the IEEE VR conference in March of 2017 showing some of his efforts to break down presence into different factors that included interactions with virtual humans, high-end VR tracking to get fully immersive body tracking, interaction abilities within the environment, as well as whether or not the scenario was coherent and believable. He also deployed a new research method where the subject would experience the full fidelity of the screen, and he’d slowly dial up the fidelity of these different factors to determine which ones were the most helpful in cultivating presence. Presence surveys are not that great at aggregating multiple individual subjective rating the degree of presence on any type of scale, but you can judge relative degrees of presence based upon your own previous experiences. Skarbez was able to determine that full body tracking was one of the biggest indicators of the depth of presence that someone felt within their experiment.

SEE ALSO
Meta Overhauls 'Horizon Worlds' Avatar System for More Realistic Representation

I had a chance to catch up with Skarbez in Los Angeles at the IEEE VR conference where we talked about his presence research as well as how his different components of presence plausibility mapped over to my Elemental Theory of Presence.

LISTEN TO THE VOICES OF VR PODCAST

Support Voices of VR

Music: Fatality & Summer Trip

Newsletter graphic

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. More information.


Kent Bye
  • Blaspie

    Virtual body tracking is not that hard, it was done many years ago by kinect, and the tech only improved since then. I can easily imagine next generation of VR to include it.

    • yexi

      Full body tracking is relatively hard in VR, because you need to do invert kinematic (IK). That mean that you need to guess the location and rotation of some bones with a very few bones which are tracked.

      Kinect track all the bones and joints, but it’s impossible with lighthouse or constallation, because camera tracking have to much latency (kinect have 60-100ms of latency), and you cannot ask home player to buy and wear a full body tracking combination every time they play.

      In reality, you can approximate full body tracking with 5points (head, hands, feet, back), but it’s very hard and not perfect.
      One the HTC Vive, you can do it with three Tracker (~300$, in addition of the ~850$ of the headset+controllers), and that work pretty well but not perfectly, for example you cannot get the elbow exact location). It will be pretty hard, if not impossible, to do the same on the Oculus Rift.

      And here I do not speak about physical collisions… else it’s even harder.

      • Blaspie

        You need low latency tracking for the headset and hand controllers, sure. But for lower body tracking (feet, back, legs)? 60ms may very well be enough. And there certainly is a room for further improvement.

        • yexi

          I do not think so, unless you want to be sick :)

          • Caven

            Why would mild to moderate latency on lower limbs make anyone sick? I don’t get sick when a Vive controller starts having tracking issues. The headset is the only component where low latency is absolutely critical.

      • rabs

        Oculus could also provide Constellation compatible trackers. It would need to be bigger than HTC ones to have a tracking as robust, but it’s technically possible anyway.

        Adding some depth camera to capture the whole body would let adjust some IK calculated coordinates. Latency can be higher, that’s just another source of data.

        With Tango and Apple next devices, it seems depth data will be more and more comoditized. It’s also used in self driving cars for example.

        So yeah, I also guess will see that added to headsets to capture facial expressions, in base stations to capture the whole body and objects entering play area, and stuff like that.

  • Get Schwifty!

    Even games like Star Trek: Bridge Crew with it’s sort of fixed sitting body add more to the experience… not too surprised at this results really, we are naturally creatures that have evolved to expect to see part of our selves interacting with the environment for reference.

  • Muzufuzo

    For me, having a full virtual body that correctly corresponds to what I am doing in reality is crucial for a convincing VR experience.

  • I know… that’s why I founded the Immotionar startup in 2014… but providing full body is hard and people do not want to pay much extra money to experience it (we used Kinects for the body tracking… and they’re expensive). And that’s why we shut-down!
    So we have to find a cheap way to implement it…

    • WyrdestGeek

      What about an in-between?

      Instead of trying to draw a complete body, then getting it wrong, breaking presence, what if it were common practice to show, like a vague representation of a torso tapering to a point?

      Over a short period of time, you’d get used to it knowing that’s how you look in VR.

      It wouldn’t look like ‘you’, but at least you wouldn’t feel like a disembodied ghost.