Report Highlights Tracking Differences Between Vive and Rift

by Charles Singletary • May 2nd, 2017

Back at the start of March, we reported on a new analytics tool called Ghostline. Developed by Aldin Dynamics, this tool was crafted to be a window into different statistics related to VR usage — how people interact, for how long, and more. Aldin Dynamics shared their program with the developers of SUPERHOT and Gallery: Call of the Starseed, but gave those interested a peek at the stats for Waltz of the Wizard.  Now they’re sharing a bit more info curated by the Ghostline tool, this time focused on the room-scale VR battle between Steam vs Oculus Home.

Ghost1

The report opens with a collection of interesting findings and one of the first things that jumps out is the fact that the NVIDIA GTX 1070 is the most popular graphics card for both Steam and Oculus users. As more begin to invest in VR-ready gear, statistics like this will stand out more and more. Another big thing to note is that Oculus tracking is improved almost two-fold when using 3 sensors instead of 2.

The CEO of Aldin Dynamics, Hrafn Thorisson, says that three sensors improve tracking with Rift so much so that it performs very similarly with Vive. He also cautions that “To be very clear, the report presents data from Waltz of the Wizard specifically, and tracking loss for that specific product — how content is designed can have an impact on tracking loss.” The design is crucial, of course, but this is definitely something to keep in mind. That small amount of players using 3 sensors (14.9%) will likely increase drastically if performance improvements like this are consistently recorded elsewhere.

Ghost2

The audiences across different countries are somewhat similar, except for SteamVR’s representation in China, and the play times aren’t very far off from each other either. An average playtime of around 28 minutes is notable as well.

Ghost3

When it comes to room-scale play, the difference in play space is pretty large. Steam has a sweet spot between 3-6 square meters but has few players using over 8 square meters of space. Only 6% of Oculus users go over 6 square meters and typically stick to 1-3.

These stats show only a small snapshot of the VR gaming community but, as Ghostline is used by more entities, creators will be able to shape their experiences to take advantage of player trends and build better content overall. You can read up more on Ghostline on their website and, if you’re a developer, submit for Early Access updates.

Tagged with: ,

What's your reaction?
Like
Wow
0%
LOL
0%
Dislike
0%
  • Marc

    Sounds like 50% of the value users dont live in a house, apartment or anything residential lol

  • Get Schwifty!

    All those results pretty much make sense – even the double rate for tracking issues.

    • Buddydudeguy

      Naw, most of it is biased and carefully worded to favor the Vive.

      • NooYawker

        How is it carefully worded? It’s laid out in plain site. 14.9% of oculus users have 3 sensor setup, you can’t ignore the majority who only use 2 because you don’t like the results. That means most oculus uses cares about front facing.

        • Buddydudeguy

          Fine, great, they can stick with 2 sensors, but they don’t get to complain about tracking loss, or make flawed claims like ” Oculus suffers 2x more tracking loss”.

          • NooYawker

            Well I agree with you there.

          • Buddydudeguy

            Cool! was all I was saying all along : /

          • JAM

            It’s not a flawed claim. Are you statistically challenged??

            If you take all the Oculus users (which apparently they did) and compare the incidence rate of tracking loss against all the Steam users, you get the result: Oculus players experience 2x more tracking loss than Steam players.

            Yes, over 80% of those Oculus users don’t have the 3 sensor setup that you do. That doesn’t invalidate the result.

            If the Rift requires 3 sensors to compete effectively with the Vive, then Facebook should ship with 3 sensors (an idea you also suggested.) And while you’re at it, if you’re really trying to match the Vive’s experience, then they should include the Touch controllers too. 😉

            And just to wind you up some more, here is another empirical data point you won’t like:
            According to Amazon, as of today, Vive is more highly rated.
            Vive: 630 reviews, 4.4 / 5 stars. (5: 75%, 4: 10%, 3: 5%, 2: 4%, 1: 6%)
            Rift: 497 reviews, 3.8 / 5 stars. (5: 53%, 4: 14%, 3: 10%, 2: 6%, 1: 17%)
            And if you look into those 1-star reviews for Rift, the majority of them are from 2017.

          • Buddydudeguy

            You don’t get it. Saying 2x more tracking loss while leaving out the fact that only applies to two sensor set up is stupid. People with two sensors don’t get to, wait for it, complain

      • No Spam

        Calm down man, I have a Rift and these numbers don’t phase me a bit.

        A lot of people think a 2 sensor setup in 360 roomscale mode is fine. Even two sensors in front of you, raised up a bit and angled down, easily gives you 270 degrees+.

        And “2x more dropouts”? Sure, there can be more data – for example, 2 times more than what? Using how many cameras? In what configurations?

        But that doesn’t mean it’s biased. Just that that data hasn’t been presented in this summary.

        • Buddydudeguy

          lol @ “calm down.

  • NooYawker

    This is interesting information, but how many games are using ghostline?

  • Well, fanboys’ flames are coming!

  • Luca Rini

    I have 4 sensors and at Arizona sunshine averaging 2 hours a time, sometimes more sometimes less. Tracking issues usually one happen when the battery are low or I’m too close to a wall. I’d have them put an arrow at all times in the center of the play area to prevent that from happening if I were them. And my play area is about 3×6 ft

    • SPRUNT

      If you can use OpenVR Advanced Settings on the Oculus, it allows for a center circle to be shown at all times in your play area. Maybe that’ll work for you.

    • Do you mean 3 x 6 meters ? 4 sensors sounds rather like overkill for 3 x 6 ft ;-P

  • Buddydudeguy

    Average Steam play area is nearly 6 meters? I call BS. Fact is most people don’t even have the space to make use of where a Vive shines and that’s ( only) with a larger play area.

    • NooYawker

      My play area is just over 5M. I have a game room/office. Just because you don’t have the space don’t think no one does. And it clearly states that 3 sensor users have twice the tracking of 2 sensors. You don’t lose tracking because you have 3, if you had two, you’d lose tracking.

      • Buddydudeguy

        I know what it “clearly states”. it’s cleverly worded too. Also I didn’t say ” no one does”. Any more straw man?

        • mikowilson

          “Fact is” you are disputing actual data with your BS assumptions.
          “Fact is” data trumps your assumptions.

          • Buddydudeguy

            The “data” is biased and cleverly worded. Oculus gets 2x more sensor loss? I mean come on. Very cleverly worded. Get a third sensor or enjoy your pleb VR.

          • ahvingsing

            Why is it so hard to believe that Oculus users experience 2X more sensor loss? You probably don’t get 2X more sensor loss with the 3 sensors you’re using, but the data may include users with 1 and 2 sensors. You don’t seem to be approaching the data objectively, so you’re coming off as an angry fanboy.

          • Adam Collins

            Angry Fanboy is exactly what he is… It was clear to anyone without a Victim mentality

          • Buddydudeguy

            Don’t be dense. I’m not a fanboy of anything. I fully acknowledge the Vive and have little bad things to say. I’m saying “Oculus has 2x more tracking loss” is a freaking joke. If you only have 2 sensors and set them up front facing, you don’t get to complain about 360/ room scale games.

          • mikowilson

            So if they said: Vive had 1 million instances of tracking loss; and the Rift had 2 million instances of tracking loss; you’d be ok with that?
            Because I think that is what they are saying…

          • Buddydudeguy

            Are we still here? Try and keep up. I’m saying users with three sensors won’t have ” 2x more tracking loss”. They obviously took a two sensor set up, went yup, there’s tracking loss. Gimme a break. Of course there is. That’s why you get a third sensor to basically double the tracking quality. For the last bloody time, it’s a relevant point left out.

          • JAM

            Calm down kid, you do sound like an angry fanboy.

            No one disagreed with you that 3 sensors are better than 2. The data supports you on that by indicating that “Oculus tracking is improved 2x with 3 sensors compared to 2 sensors.”

            (The fact you need 3 to keep up with Vive’s 2 is a different discussion.)

            The data also indicate that 80.9% of Rift users have 2 sensors (and 4.2% have only 1). I would guess that’s due to cost. I’m sure *you* bought the Rift because you believe it is technologically superior, and *you* were willing to spend more money to make that a reality.

            But I would hazard a guess that many people buy the Rift because it’s more affordable than the Vive. And after spending $500 for the base kit, those price-conscious consumers aren’t willing to shell out even more cash for the additional sensors to bring the Rift up to par with the Vive. (and again more cash for the Touch controllers.) The data support this — 80.9% have two sensors.

            Additionally many of those users may not be optimally positioning the two sensors they have (e.g. opposite corners of the room) and that also contributes to the problem.

            With all that taken into account, it seems pretty reasonable that the data is correct — Oculus players (but certainly not *you*) experience 2x more tracking loss than Steam players.

          • Buddydudeguy

            You don’t get it. It’s easy to believe with only two sensors, especially front facing, they get 2x more tracking loss. But…then why would you only have two sensors. With 3 sensors it’s basically flawless. That’s what I’m referring to as misleading and cleverly worded. It leaves out important info just saying ” Oculus suffers 2x more tacking loss”. To take a less than ideal set up then report the experience is less than ideal. It’s downright funny.

    • No Spam

      That’s 6 *square* meters, or about 2.45 meters (8 feet) per side – not 6 meters per side (36 square meters).

      2.5m per side seems like a reasonable upper bound for the majority of roomscale users. I have a Rift with about the same size play space, and I’m just using the empty area in my home office.

      Understandably, there are plenty of Vive users with larger spaces, but the drop-off is large enough that it probably doesn’t make sense to develop mass-market games with those sizes in mind.

      • Buddydudeguy

        Ahh. That’s different. I got that. Most have that.

  • OkinKun

    I’d be more interested to see direct data sets, as well as the dates they were collected, in order to compare them to individual driver updates and tracking improvements in recent versions.
    Based on my own experience, the recent updates/improvements to Oculus’s tracking seem to have put it a notch above the Vive test unit I’ve had access to, when it comes to tracking.

    • mikowilson

      Please define “notch above” and your in-depth testing methodology that allows you to definitively capture said tracking data.
      You know how you test anything? Data.
      You know what this article provides? Data.

      • OkinKun

        ..nice job misunderstanding what I was saying. lol
        I wasn’t saying the article didn’t provide ‘some’ data.. I merely said it’d be more useful to know when the data was collected, or see it in a set sorted by date. As the tracking 6 months ago was no where near as good as it is today.

        • mikowilson

          I don’t think I misunderstood anything.
          You cited your “own experience” over what we can only assume is unbiased data. Your “experience” isn’t valuable.
          I’m sure those variables you point out make a huge difference, but not to the average value of “tracking lost” over whatever period they were tracking.
          Even if the Rift tracking was absolute garbage 6 months ago, and is supremely better than the Vive now; it doesn’t have an impact on the statement they made. That’s the thing about data sets, they don’t care about what we say we “experience.”

          • jlschmugge

            Chill out. This is just a forum, not a dissertation. One person’s personal experience means something in context to the entire conversation.

            Besides, these are consumer products. Perception goes a long way.

          • mikowilson

            That’s the problem. Perception is irrelevant in the face of data. Data is important. It allows people to know, rather than think they know. I’m stoked for real data to be coming out, personally. Speaking from a place of objective truth is always better than guessing.

          • jlschmugge

            You are an outlier then. I wish everyone would find data as important. If data, rather than perception were important to people, the politicians we have today would never get public office and climate change would be a non-issue. The advertising industry would exist only to be informative, not talking geckos and “melty” cheese. It could be argued data is what is used to figure out how to impress people to buy into things, but it’s all about changing that perception to make that happen.

          • mikowilson

            As a developer. Information that is real, solid data, allows us to be informed. The truth is important. It’s been impossible for devs to talk about issues like this with any certainty. Data like this, changes that.
            Yes, people are dumb, and can be tricked by marketing. That’s not a conversation that we need to have, lol.

  • d0x360

    I have an Oculus Rift setup with 3 sensors, not sure the measurements of my play space but it’s plenty big that I can walk around.

    The 3rd sensor really is a must if you ask me and Facebook would be wise to include it in the package. You should get everything and I can’t believe that a company with as much money as Facebook is so easily willing to lose market share especially when their controllers are superior by far.

    I know Oculus is currently working on a tracking system that doesn’t need sensors which will be a big win although I’m not exactly keen on buying another hmd so I’ll wait until they implement that along with higher resolution displays. Their method of tracking is actually quite simple and I’m surprised it didn’t launch this way.

    It works by having 4 tiny depth sensing cameras on the front of the headset so no matter where you look it can obviously follow. I do wonder how they would handle when a controller goes out of view of the front of the hmd though. Of course I’m sure what they showed wasn’t the complete package and it has some tricks up it’s sleeve it doesn’t want made public.

    Once they fix and by they I mean both companies manage to have full tracking without needing to place sensors all that’s really left is a way to get rid of the wire attached to your head. Whether that means needing a small battery pack attached to your waist that can be quickly recharged I don’t know.

    Regardless I’d rather have sensor free tracking and a better display that the removal of the wire so… Once those 2 things are accomplished I’ll buy a new one. The current roadmap has both of those upgrades planned for 2018.

    As for the state of VR today I have a ton of fun with it. Better games come out all the time, I think fallout 4 will be a game changer for VR and I can’t wait for the PC version of re7 to support VR. I wish outlast 2 supported it but who knows…they could always add it in later.

    • For the outside-in tracking they should have four or more cameras on the outside of the headset positioned at the front and back and on both sides. This way they could be used to detect the world position and your hands from all 360 degrees.

      • d0x360

        The prototype was shown with only 4 front facing camera. There isn’t room in the back for a camera but even if there were unless you are looking at your hands depth sensing cameras alone wouldn’t be able to fully track the controllers in 3d space so they have to be using something else.

        It’s possible they will continue to use 2 sensors, 1 in front and 1 in back. That combined with depth sensors on the hmd would be enough to track you anywhere you go for the most part.

        The controllers are really where the difficulty comes from. I just don’t see how they could track them at all times without sensors placed somewhere in the room. They could use an algorithm to infer where they might be for a second but any longer than that and it will break any sense of immersion.
        It’s why the Wii motion plus didn’t work the way we thought it would and to be fair to us the way Nintendo told us it would. Sony has the same issue with the move controllers in psvr but at least on PC we can add sensors I just wish they were cheaper.

        That said after using both the Vive and Oculus I’m glad I went with the Oculus I just think it was a big mistake shipping with only 1 sensor in the main kit then 1 more with the controllers.

        I have to believe that’s the biggest reason for the Vive selling better. You just buy it and you are good to go. Oculus should be selling the hmd, controllers and 3 sensors for $600. Take the loss up front and make up for it over time as VR becomes more popular and cheaper to produce.

  • Paul Karaffa

    I’m at 10.5 square meters with my Oculus 3 sensor setup and it works great.

  • The increased tracking loss on Rift is simply because it’s most set up with two cameras directly in front of you and none behind to account for when you turn fully around. If most people could easily set up even just the two cameras in opposite corners of the room like the Vive then I’m sure that stat would be near identical on both machines.

  • Hacker4748

    Is the report available for download somewhere, by any chance? Or at least to see online in full?

  • jlschmugge

    Oculus should just start selling the system with three sensors included. Forget all this front facing shenanigans confusing game development.