Virtual Reality

Why that rumored standalone VR headset from Google probably isn't what you're imagining

Why that rumored standalone VR headset from Google probably isn't what you're imagining
If Google is working on a standalone VR headset, it may not be much different from smartphone-based headsets like the Gear VR (pictured)
If Google is working on a standalone VR headset, it may not be much different from smartphone-based headsets like the Gear VR (pictured)
View 3 Images
If Google is working on a standalone VR headset, it may not be much different from smartphone-based headsets like the Gear VR (pictured)
1/3
If Google is working on a standalone VR headset, it may not be much different from smartphone-based headsets like the Gear VR (pictured)
An example of a PC required to power the Oculus Rift ... there's a reason it's this big
2/3
An example of a PC required to power the Oculus Rift ... there's a reason it's this big
A Google Cardboard developer kit, with a Nexus 6 smartphone inside
3/3
A Google Cardboard developer kit, with a Nexus 6 smartphone inside
View gallery - 3 images

Standalone virtual reality that doesn't rely on another device sounds, in many ways, like the holy grail of VR. But there are still a few reasons not to get too excited (just yet) about a report this week that Google is cooking up said product.

Yesterday the Wall Street Journal ran a report saying that Google is developing a standalone VR headset that doesn't depend on a wired connection to a PC or console, and also doesn't have a smartphone living inside of it. The report frames this in-development product as a possible solution to VR's current limitations:

... virtual reality still appears years away from widespread adoption, in part because high-end headsets arriving this year require expensive PCs, while inexpensive smartphone viewers can give users headaches.

Google's planned stand-alone headset appears to aim for a middle ground: a quality experience not tethered to an expensive PC or game console.

It's the kind of description that can lead many people to conclude "this is the one I've been waiting for!" After walking around rooms with a long cable dragging on the ground behind you in HTC Vive demos or feeling like you have to buy a Samsung flagship just to use an entry-level Gear VR, an all-in-one VR headset sounds like the ultimate destination. And, long term, it may be.

But here's the problem with this rumored one from Google. Assuming the device exists, we have no idea how powerful it is (apart from supposedly having "high-end processors"). No matter what the processors are like, we can guarantee it isn't close to "high-end" compared to today's PC-based and (to a lesser degree) console-based VR. Barring an engineering breakthrough for the ages, that just isn't possible right now.

What's more likely is a headset that's about as powerful as a Galaxy S7-running Gear VR will be (or other high-end smartphone-based VR with a 2016 flagship inside). Or, at most, perhaps a little more powerful than that. Only instead of needing to slide a phone inside, all the phone internals that you need for VR (display, processor, GPU, battery, sensors, wireless and so on) are permanently built into the headset itself. Being Google, it would almost certainly run a VR-centric build of Android ("Android N," likely), much like the Android phones that are powering current Gear VR and Cardboard headsets eventually will be.

Smartphone-based VR will continue to advance: mobile headsets with 2016 flagships inside will be more powerful than today's headsets with 2015 flagships inside. Oculus execs even predict that the Gear VR will only take two or three years to be as powerful as today's Oculus Rift. Any breakthroughs in mobile technology would presumably carry over to smartphones just as they would a standalone wireless headset.

Either way, this product suddenly sounds much less exciting than first impressions would have you believe, no? It's probably smartphone-based VR that you can't use as a smartphone.

A Google Cardboard developer kit, with a Nexus 6 smartphone inside
A Google Cardboard developer kit, with a Nexus 6 smartphone inside

What would it take to make a standalone headset today with horsepower far beyond 2016 flagship smartphone levels? For it to be roughly on par with the Rift and Vive? Well, you'd need a high-end GPU, and that takes up space. Such a graphics card would likely require serious fans to keep things cool. And that's not including the battery, which would: a) need to be enormous to power this beast of a GPU for any extended period, and b) compound the cooling process even further, requiring even bigger fans. Add that up, and you have something that's gigantic, heavy and loud enough that nobody in their right mind would want to wear it on their head.

There's a reason today's high-end gaming PCs aren't the size of a mobile device. If Google could put that kind of horsepower – along with the battery to make it portable – into a HMD, then the Samsungs and Apples of the world could also put that kind of horsepower into a similarly-sized smartphone or tablet. Technological advances for the two are going to run parallel no matter what.

Decades down the road, VR may all be wireless and self-contained. And, whether it's Google or not, some companies will try to make such headsets today. At Gizmag, we've passed on covering pitches from startups trying to make standalone VR headsets, primarily because they redundantly put smartphone parts permanently inside the headset, when we already have those parts in our pockets. It just doesn't make sense to pay a second time for the same components you already have in your phone.

Don't let us be dream-killers here: there will be big breakthroughs in VR through the next few years that make today's gear seem primitive by comparison. Maybe someone like Google will even make us eat our words. Just remember to keep your expectations in check when you dream about such a standalone device today. If it's 100 percent mobile, then its internals are going to have to be very similar to today's mobile devices. There's a balance among power, battery and size that simply has limits.

Those boundaries will keep expanding through the years, but that isn't likely to happen overnight, with that lightning striking only one company working on one form factor. When it comes to mobile innovation (things like processing power, cameras and battery life), it's largely a story of parity across the board.

An example of a PC required to power the Oculus Rift ... there's a reason it's this big
An example of a PC required to power the Oculus Rift ... there's a reason it's this big

There are some other cool insights from the WSJ report, like word of a custom chip from Movidius inside that uses the headset's forward-facing camera to track head movement in relation to space. If that worked accurately as a positional tracker (so it senses not just head rotation but also walking, crouching, leaning, etc), that could be a major perk. One of mobile VR's biggest weaknesses right now is that it lacks body-tracking of any kind. This is a problem Oculus' John Carmack is also focusing on for future versions of the Gear VR: find a way to have positional tracking inside the mobile headset (presumably from an inside-out perspective), without using any external sensors.

Most of the talk around this upcoming "year of VR" has centered around Oculus, Valve/HTC and Sony (rightfully so), but Google is another company worth keeping an eye on. You'd be naïve to take Google's talk of Cardboard being an "experiment" at face value. From the get-go, it was a scheme to get developers making content for the Play Store, planting seeds for future consumer products like this alleged standalone headset, as well as consumer-friendly smartphone-based headsets similar to the Gear VR.

We just don't recommend getting too worked up over the prospect of Mountain View cooking up a dream VR device that strikes the perfect balance of power and mobility. The parts it uses would almost have to be similar to (if not exactly the same as) parts you'll see in 2016 flagship smartphones from companies like Samsung and Apple.

View gallery - 3 images
9 comments
9 comments
phissith
Just develop a game VR rigged that uses Desktop powered but transmitted the image to the VR Headset. Viola the muscle of PC and the portable of wire free. Sheesh!
Mr. Hensley Garlington
I agree with phissith.
The video can be transmitted wirelessly and the inputs can be transmitted wirelessly as well. It is necessary and should result in lighter headsets.
Derek Howe
phissith, I'm sure they have already thought of that. The Rift's head unit has 2 usb 3.0's going to your pc. Which means were talking about some serious bit rate. Which can't be done through wimpy bluetooth, not even wifi's newest 802.11AC comes close enough. Just 1 USB 3.0 port is anywhere from 5 to 10 times faster then a wifi 802.11ac connection. So while everyone would love wireless, were not there yet.
paroway@gmail.com
That's kind of the problem, isn't it?
jati
Will, you should have a look at the benchmarks of the iPad Pro, especially the ones for the graphics performance. And that chip is tiny. I'm not sure if Google could build this kind of VR headset, but Apple most definitely could.

That all these PCs for the Oculus are huge doesn't mean that their processing power is necessary. Maybe they are just overkill for what non-gamers would want to do with a VR headset anyway.

I think you underestimate what such a device by Google or Apple could do for the everyday consumer. They don't count polygons and they don't care for lighting effects.
Jimjam
Yeah I have to agree with phissith that wirelessly connecting a image generated on a gaming PC to a headset looks like the way forward. The problem is that this introduces latency at present, although with improved WiFi or new short range wireless Google and others might be able to eliminate this.
DominicLacroix
I've no idea what you (the author) were expecting, but anyone who knows anything about current SOCs and performance in general would not be expecting anything more than a smartphone SOC powering the device. Also, GearVR is VERY good (apart from the lack of positional tracking) so a bit more of a comparison for those who don't know the difference would've been nice.
Steve Jones
Maybe the new breakthrough is a way to do some of the graphical work - rendering or whatever - in the headset, thereby reducing the required bandwidth and allowing wireless operation.
Mel Tisdale
There are economies of scale to be had here. Virtual reality systems need some technology that enables them to maintain positional accuracy while the wearer moves about. So too do modern smart bombs and the like, if only as a backup for when the GPS signal gets jammed. So far, so good. Now add the need of autonomous and semi-autonomous cars to also cope with GPS jamming and the resulting production volumes should really bring the price down. Collaboration on the design should also bring benefits.