What should we expect from the next gen of VR
The Vive Pro is out and the Pimax is on the horizon, whilst it’s too early to tell what the Pimax will be like, the reviews are in on the Vive Pro and they all seem to be “Meh, it’s a nice little bump but hardly gen 2”. So… What DOES Gen 2 look like?
For me, this is the #1 thing that Gen 2 needs, we’re starting to see Wireless add-ons for headsets now and that’s great, but they still require a PC to power them.
I think the Gen 2 headsets should not only be tether-free but truly stand-alone. Look at the headsets in Ready Player One for an example of this, to access the Oasis, they didn’t need a PC, they just put the headset on.
This may sound implausible but it’s not, or at least, it won’t be soon.
Last year, AMD started releasing their mobile graphics chipsets known as the ‘bifrost’ series. The latest one the ‘Mali-G71’ has some impressive benchmarking scores:
The blue bar is the Mali-G71 and the orange is the Nvidia GTX 1080, now although the Mali is quite far behind, just take a look at that graph again, it’s just below half the speed of the most powerful graphics card currently available on the market.
That means that a smartphone with this chip in it is already half the power of a top of the line gaming rig. A few more years and we’ll easily have mobile devices that can power 4K VR.
Even now, we have fairly powerful gaming laptops, the size of those laptops can be reduced quite considerably if you take the display away from the build and put it into a connected headset, this makes standalone backpack VR rigs possible, right now!
2) Fully immersive displays.
One of the major problems with current gen VR is the Screen Door Effect (SDE), in a lot of games it’s easy to forget about but when you want to watch a movie, play a game with a detailed UI (Elite Dangerous for example) or just sit down and do some work in VR, the SDE is a major block to that.
Again, we’ve seen display bumps with the Vive Pro and the Pimax 8k promises 4k per eye, which is a fantastic step in the right direction but reviews of the Vive Pro and early reports of the Pimax 8K are reporting that although reduced, it’s not the end of the SDE.
One of the big problems here is the amount of processing power it takes to power a 3D 4k display, however, a new technique has recently been developed called ‘Foveated rendering’ which could mitigate those problems until our processing power catches up with our display technology. Rather than waste time in this article explaining what it is, take a look at the linked Wikipedia article above for more information.
Another major issue is Field of View (FoV), early prototypes of the current-gen were terrible for this but thankfully the consumer versions made it a lot better, however no headset from the current gen that I’ve personally tried has managed to eliminate the feeling of looking at the world through welding goggles.
Early reports are suggesting that the Pimax has actually solved this to a degree creating an ultra wide FOV that goes right to the edge of our peripheral vision, although it’s still not perfect, reviews are suggesting that the vertical FOV is still notably shallow and that lens distortion in the peripheral vision is quite noticeable. Of course, the Pimax isn’t out yet so any demo versions that have been seen still have to be considered as ‘beta’ devices, so we still have to wait for the first consumer versions to be shipped.
For a headset to be truly next-gen, the SDE has to have been reduced so dramatically that users can read a word document without any squinting or moving their head, they should be able to feel completely immersed in the world around them and see no black bars or lens distortions.
3) Inside-out tracking
Ok so this one already exists in the current-gen, Microsoft’s ‘Mixed Reality’ headsets use this and although they work, they don’t work well. It feels very much like the first instance of this technology and needs to be a lot better to be able to call it ‘gen 2’.
Inside-out tracking is basically the ability to track the position and angle of a headset without the need for base-stations or external cameras to tell the headset where it is. Current gen does this with cameras in the front of the headset which use software image processing to determine what it can see.
This is actually a great way to do this technology but it needs to be improved, also two cameras in the front is not enough, there needs to be multiple cameras dotted around the entire headset in order for it to be able to detect when you have moved your controller behind you.
Inside out tracking done this way has positives and negatives
- People don’t like cameras in their house, so strapping cameras to their heads may be offputting for some
- Current gen uses black and white camera footage to process images, however, future ones should also implement infra-red as the trackers work poorly in low light.
- Image processing eliminates the need for powered external trackers, hands and feet (and other objects) could easily be tracked simply by putting a sticker on them. As processing technology improves, even this will become unnecessary.
- It would be a lot easier to use these devices to implement augmented reality headsets
- Inside out tracking means you can wear your headset anywhere, not just in your designated ‘VR’ space.
- Once it is advanced enough, it can probably be used to track individual fingers and possible even facial expressions.
- I’m pretty sure that this tech is very close to being reliable, for gen 1 it’s already incredibly well-done and with so many companies working on this, it’s not going to take long for someone to crack it.
4) Haptic feedback
I mentioned Haptic Feedback before in my previous VR article, but it is worth repeating as this is a major part of what will make VR really feel ‘next gen’. Right now, we don’t have hands or feet in VR. Sure, we can simulate our hands fairly well with controllers but to truly immerse us in the virtual world, we need to have complete control over each, individual finger.
We also need our feet back, even though we don’t have infinite treadmills or gyroscopic rigs yet, just being able to see and track our feet in VR adds to the immersion to an incredible degree, I highly recommend trying out VRChat to see what I mean, they have implemented a system which approximates your foot position based on where your head is and if you walk around, your VR feet shuffle around with you.
The company ‘HaptX’ have recently developed some haptic feedback gloves which not only give you control over your fingers but also give you the feeling of touching something, they even have the ability to apply resistance to your hands so you can feel yourself squeezing something.
It should be noted though that these gloves a) have not been released yet and b) are not targeted at consumers at this point. It’s not really surprising though, just looking at them you can tell that these things would be incredibly expensive to produce so putting them out to the professional market first makes a lot of sense whilst they figure out how to get their costs down. So I’m not expecting technology quite this advanced in Gen 2. However just high-fidelity finger tracking and feet tracking would be enough for me, for now.
With all this in mind, I imagine a real Gen 2 of VR is still probably a couple of years away. If I were to guess, I’d say 2020 at the earliest before we see a released headset with all of the above features out of the box. However for those of us who can afford it, a lot of the tech will probably become available as ‘add ons’ before then.
Who knows? Give it a year or so and perhaps I’ll see you in the Oasis.