At CES this week, Pico unveiled a pair of its latest standalone VR headsets, Neo 2 and Neo 2 Eye. While the company is targeting enterprise customers in the West, the Neo 2 line is eyeing up the same hardware category as Facebook’s Oculus Quest, and doing a solid job at that.

The only appreciable difference between the the Neo 2 and Neo 2 Eye is that the latter includes integrated eye-tracking from Swedish firm Tobii. It also comes with slightly more RAM than the standard version to account for eye-tracking as well as a different color scheme; Neo 2 is white and Neo 2 Eye is dark grey. I’ll get to eye-tracking below, but for now let’s talk about the Neo 2 line in general.

Pico Neo 2

The first thing I noticed about my short time with Neo 2 was how comfortable and well-balanced it was—more so than Quest’s front-heavy design—thanks to Pico’s inclusion of a rear-mounted battery, which is incorporated with the headset’s ratchet system with a single knob. An overhead strap is a rubbery affair that has a few fixed length settings, something I wish was instead a bog standard velcro strap for easier and more accurate fitting, but it did the job.

Photo captured by Road to VR

As a company, Pico has been pretty cognizant about weight distribution in the past with its 2016-era Neo headset, which cleverly housed the battery in a Nintendo-style tethered gamepad, so it’s nice to see the company is still focused on keeping one of the heaviest pieces of a standalone VR headset in good balance.

The second thing I noticed was its slightly mushy, albeit entirely serviceable optical positional tracking. To its credit, Neo 2 wasn’t jittery or too lurchy like the company’s previous Pico Neo from 2017, putting Neo 2 in the ‘acceptable’ range for tracking.

My demo didn’t quite live up to Oculus Quest, which by all accounts is the bar to reach when it comes to inside-out tracking for head and hands, but with a useful guardian system and pass-through capability to boot I was mostly satisfied Neo 2’s 6DOF tracking. Granted, the CES show floor’s bright lights and bustling crowd make for a challenging tracking environment, so I’ll have to reserve my ultimate judgement until I see it in a wider range of environments.

Qualcomm Chips Now in More Than 30 Headsets, First XR2 Devices Expected in 2H 2020

By this point, you’re probably wondering whether this really a hands-on piece or a straight-up comparison to Quest, which sells to consumers starting at $400. And you’d be right in saying its a bit of both. The only other real device in the product category currently is HTC Vive Focus Plus, which hasn’t presented much competition for Quest in the West. That said, Pico Neo 2 and Neo 2 Eye are set to launch at some point in the West to businesses for $700 and $900 respectively, both with 128GB internal storage. Meanwhile, Facebook charges $1,000 for the 128GB Quest through it’s ‘Oculus for Business’ program.

While there’s important platform considerations not to be ignored, here I’m looking mostly at how Neo 2 compares to Quest from a hardware and capabilities standpoint.

Tracking aside, what did fare remarkably well was Neo 2’s 6DOF controllers. Although it doesn’t have the precision or finessed ergonomics of Oculus Touch, its range of motion and overall reliability seems more than acceptable for the headset’s enterprise-focused use-cases.

Unlike Quest, which uses inside-out tracking to track both the headset and Touch controllers, Neo 2’s controller tracking is based on a NDI’s Atraxa controller platform which fuses data from an on-board electromagnet (EM) and inertial measurement unit (IMU), resulting in a positionally tracked controller that doesn’t suffer from occlusion (but may suffer from sources of EM interference). I didn’t get a chance to put it through the ringer with something like Beat Saber (an easy benchmark for latency and accuracy), but it felt more than acceptable throughout.

Photo captured by Road to VR

The controller’s plastic feels a tad on the cheap side, and button placement isn’t the greatest, although I was glad to see thumbsticks instead of touchpads here. Ergonomically it isn’t anything to write home about; it’s more wand-like and uses a ‘click-to-grip’ button, whereas most headsets have moved toward a ‘hold-to-grip’ trigger.

Integrated stereo speakers are very similar to Quest’s (hidden in the head-band), though I couldn’t get a good feel for them considering the noisy show-floor environment.

Hands-on: Pico 'VR Glasses' Prototype is the Most Impressive VR Viewer Yet

On both Neo 2 and Neo 2 Eye I played Angry Birds VR: Isle of Pigs (2019), which I know quite well from my review of the game. Although frame rate wasn’t at all where it needs to be right now on that game in particular, I suspect the Neo 2’s Snapdragon 845 would be more than up to the task after a little bit of headset-specific optimization; in this case it seemed that Pico just grabbed for the nearest game they could to have something to show at CES.

The headset’s Snapdragon 845, which should be capable of running VR apps at an acceptable frame rate for the 3,840 × 2,160 (1,920 × 2,160 per lens) LCD panel, which is clocked at 75Hz. For comparison’s sake, Oculus Quest offers dual panels with a per-lens resolution of 1,440 × 1,600 at 72Hz, driven by the more demure Snapdragon 835.

Image courtesy Pico Interactive

Neo 2 uses a single panel with software-based interpupillary distance (IPD) adjustment, something that Pico says should serve users with an IPD of between 55mm–71mm. My IPD is a smack down the middle at ~63mm, so I didn’t have any issue, but a physical IPD adjustment is greatly preferred for its ability to help a wider range of people get their eyes into the sweet spot of the lens.

Pico Neo 2 Eye

Then there’s the Neo 2 Eye, which integrates Tobii’s eye-tracking tech. I had the opportunity to go through the Neo 2 Eye demo, which was nearly identical to the one I tried at Vive Pro Eye (which also uses Tobii eye-tracking) at CES 2019.

One of the big benefits Pico is trumpeting for Neo 2 Eye is its foveated rendering, which is supposed to improve performance by only rendering the scene at full resolution where your eye is looking, while reducing resolution in the periphery. Unfortunately the lower-resolution in the periphery was more noticeable than I would have hoped for an eye-tracked solution. Ideally you aren’t supposed to notice the edge of the high-resolution center, which should be locked onto the user’s fovea (the center view of the eye which sees in high detail). It was simply too inaccurate a demo for the illusion to work, which is a shame because we’ve seen remarkably solid Tobii eye-tracking in other headsets.

Though the foveated rendering was easily noticeable, it did what foveated rendering is designed to do: allow for higher frame rates and more complex scenes.

To that effect, I was told the headset was using a generalized eye-tracking profile, which was a possible reason why it wasn’t offering the sort of accuracy I’d seen in Tobii’s tech on  the PC-tethered Vive Pro Eye. Again, it’s something I’ll have to test in a less hectic environment where we can go through a proper eye-tracking calibration process and see if Pico was truly able to use Tobii’s tech to its fullest on the mobile VR platform.

And while we’re looking forward to testing the Pico Neo 2 in the conditions of our choice, the hardware seems an admirable entry in the 6DOF standalone category. In may not match Oculus Quest in a few categories but businesses looking to get a solid, lower cost 6DOF standalone with slightly more horsepower, resolution, and the option of eye-tracking wouldn’t be remiss by casting a curious gaze at Pico’s latest and greatest.

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.

Well before the first modern XR products hit the market, Scott recognized the potential of the technology and set out to understand and document its growth. He has been professionally reporting on the space for nearly a decade as Editor at Road to VR, authoring more than 3,500 articles on the topic. Scott brings that seasoned insight to his reporting from major industry events across the globe.
  • impurekind

    I find it so ironic that many of these system’s hand controllers are simply getting closer and closer to my design from around 20 years ago at this point, yet still aren’t actually as good (I have a better solution for the grip button/trigger imo for example): (the design lower down)

    • asdf

      ive seen you post this a lot. I remember back in 2005 my friends and I talking about how we though they should do vr. We all agreed theyd either need to stick outward facing cameras to a headset display and use depth sensing/IR to locate the players movement in realtime, or just put up two inward facing cameras aka basestations. now basestations are not cameras, theyre IR lights but damn we were close.

      Others and I have had so many ideas that others did before us that it kinda shows some good ideas are really just obvious ideas. For example the controllers look the way they do because you’re both considering ergonomics and the form of the hand first, and the rest fills out from there. Its pretty cool and confirming when you see a product develop in the same way youd make it or have said it should be made.

      Also side note, any idea you ever had, you have to assume someone else just had that same idea. So watch as vr object verification via blockchain will be huge someday, even ready player 1 hints at it.

      • impurekind

        Yeah, my idea was actually just a natural evolution of a bunch of other things around back then too, like Microsoft’s haptic flight stick and controllers or the rumblings for the Wiimote for example (then it was called the Revolution and nothing about the controllers had been shown yet). It’s just funny how long it takes some of these companies to get there.

  • Darshan

    I believe it’s little off the mark by including SD845 it need to be SD855+ or SD XR2 soon as they can. Software refinement of Oculus is long way to reach, no doubt shouldn’t matter in business environment. I still believe guarding system is must.

    • Ad

      I’m sure they’re scrambling to do that. It would require retooling a lot to make cooling work, but if this just killed the quest in terms of compute power then I would buy one at 700 or 800 hundred assuming it was going to have an adequate software library.

    • Andrew Jakobs

      Well, every upgrade of the soc will increase the price even further.. I even see it here, ‘oh if they add this or that I would buy it for $500 and even with that added for $600’ not realizing that the extra’s they want would make the headset way more expensive then they want to spend… Yeah I want 4K per eye with 180 degrees FOV clear lenses and eyetracking, and I want it for $400………. Ohyeah, I also want the GPU to drive those for maximum $200…

  • Rupert Jung

    I really, really hope that eye tracking will be a part of all future headsets. It’s IHMO the only missing part for a perfect VR experience where the rendered world ‘knows’ what you are looking at and is able to react to it.

  • Ad

    Good lord if they, on their own or with Valve’s help, can make a 32GB version without eye tracking and an OS that could easily have games ported from the quest and sell it for 500 I would buy one in a heartbeat. I can’t buy anything from Facebook so I’d love a quest alternative. If they can get the XR2 chip into this for 600 it would do pretty well too because it would have a lot more capability on top of the comfort and resolution. I wish eye tracking was cheaper too but that’s not their fault. Software is obviously worse on this but that’s fixable if they worked with someone else (could HTC license this out to make a consumer version?) and the tracking sounds worse than the quest but if it doesn’t have occlusion issues then it seems competitive.

    Someone needs to break the quests monopoly fast and this looks like a good contender, the price just isn’t there. Although if this could run all quest software I would personally pick one up.

    • asdf

      this. I do love my quest but the FB data hoarding mentality is starting to kick in full gear with having to link a fb account to get features. I love vr but hate social media and am now back at the whim of fb with my data because of it. this was thier plan the entire time by moving to hardware and it sucks tbh.

      • Ad

        Did people think they were selling hardware at cost just for fun? Especially now that the quest can use steamVR they’ll work hard on Facebook integration.

  • blue5peed

    We desperately need competition for the quest if to do nothing but keep Facebook honest. Every Intel needs an AMD.

    • Paul Schuyler

      Absolutely right on. Facebook is not the company we want owning the space based on the experience of many developers and the Quest. The ultimate would be an open platform without Facebook’s irrational (and sporadic) developer relations.

    • Sofian

      We ll see more competition in the coming months when the HMD built around the Qualcomm XR2 are announced, I don’t see the Quest being relevant after the end of this year.

      • Blaexe

        “Ecosystem” is the key word. And the XR2 won’t magically enable high quality tracking and a low price point. Quest will sadly be as relevant as ever.

        • Sofian

          I don’t mean that people will stop using the Quest, just that it will outdated much faster than Oculus expects.
          The “not any time soon” HMD Michael Abrash talked about at OC6 better comes soon.

          • Blaexe

            And I 100% disagree. With no real competition, Quest won’t be outdated.

    • Adrian Meredith

      true, but were a long way from that if the quest had sold 100 million units then oculus would be getting complacent.

    • NooYawker

      The problem is FB can sell their devices at zero profit or even at a loss since their profits come from data collection. Plus they have their own storefront. Other companies don’t have that luxury and need to make a profit from their hardware.

      • G-man

        theres no reason other standalone headset makers couldnt make their own stores. that sell the exact same apk files the quest does. devs pretty much wouldnt have to do anything.

  • mellott124

    Pico definitely stepped up their game this year. I’ve never been impressed with these guys in the past as everything seemed 80% done but this year they looked good. The Neo 2 looked great visually and tracking was equally good. Their pancake optic HMD was also very impressive and better than some others I’ve seen. I’ll be keeping an eye on this company moving forward.

  • Jim P

    These systems will not survive they have no Ecosystem. They have to rely on Steam. Simple as that. Not shitty on the hardware it’s nice just telling the facts.

    • Sofian

      This is not for consumers, enterprises can build their own apps.

      • Vernimator

        I create training apps for Enterprise. My clients are looking at the Quest for their training solutions. Just to give you an of Oculus in Enterprise check this out.

        “Walmart announced its plan to expand its pilot VR training program by delivering 17,000 Oculus Go headsets to stores across the country, making it one of the largest deployments of its kind in history.”

    • G-man

      rely on steam? this is a android based device. it has nothing to do with steam

  • Mike


    • Adrian Meredith

      I get what you’re saying but there are a ton of benefits of high quality lcd vs oled, sde, black smearing, mura etc. The blacks aren’t even that good on vr oleds because of pixel persistence means they aren’t pure black like normal oleds

      • Bamux

        I also do not understand why some think OLED is the better solution, in all headsets with OLED (Rift, Quest, PSVR) that I have tried so far you could see a clear mura in dark scenes. Either I had extreme bad luck with the panel lottery or I am very sensitive to mura.

        • Sofian

          Because even with mura it looks much better than black turned into grey.
          There are already LCD panels with miniLED tech being developed, it could be a good option while waiting for microLED.

        • Mike

          What Sofian said. Grey-blacks are extremely immersion-breaking and non-aesthetic to me. Also, see my reply to Adrian above.

      • Mike

        It’s not that black-and-white (no pun intended). The Vive limits black levels so it doesn’t achieve 100% black levels, which prevents black smearing but causes Mura – but this looks much, much, much more realistic than grey-blacks, and the Mura is easy to ignore.

        The Odyssey+ actually DOES achieve 100% black levels and no mura, at the expense of occasional black smear in occasional scenarios. BUT, this black smear can easily be eliminated using the brightness adjustment in the OpenVR Steam plugin. You can adjust it for each game, as needed, or just leave it somewhere halfway between Vive-levels and full-blacks. It works great.

        SDE is not intrinsically better on LCD. You’re talking about pixel structure being Pentile vs RGB. OLED tends to be made Pentile, but that’s due to cost – the PSVR is RGB OLED. Additionally, there are pros-and-cons to each, at least at typical VR resolutions. At 1440p on the Pimax 5K+, RGB actually has more-apparent SDE than the 1600p Vive Pro – this is because the human eye is evolved to easily detect straight lines, so the neatly-lined-up RGB pixel edges are more visible than the broken diamond pattern of Pentile. Additionally, the Odyssey+ is 1600p with a slight blur filter that gives it the SDE of a 2200p screen (which I’ve verified by comparing it to a Reverb – the SDE looks the same).

  • The article’s title is a bit click-bait-ish. This thing is, in no way, capable of “giving the Quest a run for it’s money”. It’s inferior in all ways, and apparently they lack both the hardware ability and software ability to keep up. You can’t fake genius like John Carmack with a pile full of less engineers. The Quest was his baby for at least 3 solid years.

    Reminds me of high-school, ages ago, when they were trying to tell us that Think-Tanks and committees were going to put an end to the great genius inventors. There is no number of monkeys on typewriters that will eventually write Shakespeare. Zero multiplied by any number is still zero. And no amount of lesser minds can beat one dedicated genius.

    • talk talk talk

      There is no number of monkeys on typewriters that will eventually write Shakespeare

      this is pure racist, it’s because people like you that we have a cartoon character for president.

      • G-man

        did you hit your head today, and also the day before, and many times before that?

      • brubble

        Sloooooow clap. Slower….sloooower.

    • G-man

      inferior? except for resolution, frame rate, power, controller tracking, potential for eye tracking, has the exact same software availablility, being android it can run all the same things quest can…. so wtf are you talking about

  • USA 2020 Keep America Great

    VR Headset makers need to ” OFF LOAD ” components like the CPU – GPU – Battery – Gyros, Sensors, Cams from the already heavy front visor/display area and integrate those components into upgradable modules in clip in/clip on modules into the metal HALO part of the headset in the temple area of the face/head/headset.

    Leave the display area only for optics, lens, display.

    Counterbalance the weight by moving those components either integrated into the metal HALO band of the headset or create ” POWER PACKS GPU – CPU POWER PACKS ” that can be worn in a FALLOUT type wrist band ? or Shoulder Pack ? Or Light Military tactical vest with a backpack, and multi pockets for batteries ?

    Use the metal band of the headset’s HALO for heat dissipation, a heat sink.

    PIPCO is headed in the right direction with locating the battery in the back of the PIPCO NEO 2 Headset.

    Have ? Duel Processor Module units with multicores on both sides of the head/headset to run each eye, or display screen.