Apple ‘Personas’ on Vision Pro are already the most likelife real-time avatars you can find on any headset today, but in the next version of visionOS, they’re taking another step forward.

Apple today announced that its Persona avatars for Vision Pro will get a major visual upgrade with the launch of visionOS 26, due out later this year.

Personas on Vision Pro are generated on-device after users take a short scan of their face using the headset. Once generated, the avatar is used for social experiences like FaceTime.

Currently, they’re the most lifelike real-time avatars available on any headset today. Although they impressively capture subtle motion from the user, they have always felt somewhat blurry or ghostly.

VisionOS 26 promises a big visual update that will greatly reduce that ghostly look, and present a more complete view of the user’s head, including a “full side profile view.” Apple is also promising more realistic hair and lashes, and more than 1,000 variations of glasses, so glasses-wearers can find something that looks just right.

View post on imgur.com

Although visionOS 26 will be available as a developer beta starting today, it isn’t yet clear if the Personas upgrade will be available in the first version, or roll out in later versions of the beta.

Beyond the visual upgrade to Personas, visionOS 26 will also make improvements to how social experiences work on the headset. New developer tools will allow for the creation of co-located virtual experiences; meaning two headset users in the same physical space will be able to see a shared virtual experience that’s visually anchored in the same space for both. That same system will allow for remote participants to join as Persona avatars, making for a mixture of in-person headset users and remote participants in the same virtual experience.

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.

Ben Lang
Ben is the world's most senior professional analyst solely dedicated to the XR industry, having founded Road to VR in 2011—a year before the Oculus Kickstarter sparked a resurgence that led to the modern XR landscape. He has authored more than 3,000 articles chronicling the evolution of the XR industry over more than a decade. With that unique perspective, Ben has been consistently recognized as one of the most influential voices in XR, giving keynotes and joining panel and podcast discussions at key industry events. He is a self-described "journalist and analyst, not evangelist."
  • Christian Schildwaechter

    Short reminder: Facebook announced they were working on photorealistic avatars with a blog post in 2019-03 ("Facebook is building the future of connection with lifelike avatars"), and showed them off in an infamous interview with Zuckerberg by Lex Fridman inside Quest Pro (youtu_be/EohIA7QPmmE 2023-09), which required both to be 3D scanned in a studio in a rather elaborate and long process.

    Admittedly creating Meta's Codec avatars now only requires a single iPhone instead of the previous setup with 171 high-resolution cameras. But the biggest progress that made it to the users of Meta VR avatars so far was getting legs for their floating torsos around the time of the Fridman interview.

    The roles seem somewhat reversed: Meta, usually following a "move fast, break things" philosophy, is somewhat held back by perfectionism, sticking to cartoony avatars despite having way better options in the lab. While Apple started with a "just do it yourself with the AVP cameras" approach that created slightly creepy, but workable avatars, and now gradually improves upon them, instead of waiting until they have something that feels finished, as they usually do.

    • Jonathan Winters III

      So true, Meta failed on this feature and are now being left behind. However now that Quest 4 won't be here for around 2 1/2 years, they have lots of time to get it right.

      • Christian Schildwaechter

        In their defense, the codec avatars rely heavily on eye and face tracking to deliver a life-like performance, which is why all the interviews and demonstrations used the Quest Pro. Meta sort of booted themselves out of this (and a lot of advanced UI options) when they didn't integrate at least eye tracking in the 2022 Quest 3.

        They probably expected to introduce the codec avatars much sooner as a Quest Pro-only feature, a plan that was foiled the moment Quest Pro crashed straight into the ground. But it at least means that a 2027 Quest 4 will most certainly incorporate eye and face tracking.

        • Thud

          "2027 Quest 4 will most certainly incorporate eye and face tracking."

          By then toasters will employ eye tracking ;)