OP
OP
Eye-tracked foveated rendering is needed to get past 2K x 2K. Otherwise hardly anyone, if anyone, can render these displays natively with today's or future content.
-
Ever wanted an RSS feed of all your favorite gaming news sites? Go check out our new Gaming Headlines feed! Read more about it here.
-
We have made minor adjustments to how the search bar works on ResetEra. You can read about the changes here.
Oculus/Facebook Exec: ‘Quest Is The End Of Our First Chapter Of VR. What’s Next Is Where Things Really Get Interesting’
- Thread starter DarthBuzzard
- Start date
So would you like it better if it were Google, Apple or Microsoft leading the way instead? Or do you want it to remain niche? Big tech is big tech. VR's capabilities inevitably go well beyond gaming and big companies are going to take advantage. Where would PCs be if they remained niche? The tech won't be locked down. Yes, Facebook has a head start, but if it really starts being successful, you better believe other companies will invest heavily, and then it will be Facebook that has to find their "social VR" niche which is where their expertise/following lies.
If you're talking gaming only I wouldn't worry about it either. Valve has their own platform and I don't see that changing anytime soon as long as Steam stays big. It may be higher end right now, but that could change, plus other gaming/tech companies like Microsoft might increase investments once they see more profit potential. Long term I really don't see it being an issue. And right now Oculus is doing the best job of getting the software and cost efficient hardware out to an extremely untapped market. Plus you have the console industry (Sony) getting into VR as well. It's coming from all directions, whoever is in charge now probably won't be in 10 years; and if they are, they'll still have some hefty competition.
Last edited:
The problem with that most Quest titles, several PSVR (especially base), Pimax, and everyone with lower end GPU's already render well below their native panel resolutions. I'm not convinced there wouldn't be appreciable gains upscaling to those much higher resolution panels, and it doesn't explain the lack of HDR, which definitely should have trickled down by now.
I'd begrudgingly rather see Apple or Microsoft take the lead than Facebook, yes. They're bottom of the barrel in my giant company trust index. And I'm not sure what you're saying about computers. When they began to enter the consumer market in earnest in the 70's and 80's they didn't require a giant company that had dominance in tons of other fields to pump money into it. The very concept made businesses. You may think a massive company trying to consolidate the image of a new technology to be only their brand is a necessary part of it leaving the niche market, but I don't think that's true. After all, every company in this post had fairly humble beginnings in their respective fields. This isn't a half century ago and tech companies are larger and a small number have a lot of control over what hardware, software, and content the average person sees than they ever did in the previous millennium. I think more than consumer computers ever were, the VR market is vulnerable to be effectively taken over, or at least heavily segmented by a large actor to no benefit to VR as a whole. As much as Facebook wants you to believe it, these aren't the good old days.
Humm...ok I guess ?
I was being serious and was not trying to spin anything. As a PSVR owner I can´t wait to see how next gen VR will evolve.
But you (and that video) made my point exactly. Eye tracked foveated rendering will enable much higher quality visuals once we have super high resolution screens and high powered devices. It won't help low end devices like the Quest much more than what we are seeing today in games that use Fixed Foveated rendering. Unless you think the Quest's CPU is capable of smart enough AI to recreate an image from 1/20 of the pixel data in realtime faster than it would take to just render that image in the first place.
OP
OP
The Quest is irrelevant here. What we should be considering is the Quest 2 or even 3, as those are the models that will have eye-tracking. Not to mention that Oculus have said they've made progress on getting their photorealistic codec avatars working on mobile compute. They're very good at optimizing, so I'm sure a Quest 2 or 3 will handle it without issue.
I agree, I was responding to a post that said
"Just adding eye tracking and foveated rendering would make the 835 in the current Quest perform nearly as well as an Xbox One X and PS4 Pro put together.
It would make the PS5 perform better than what we can expect from a PS6 Pro."
That's just unrealistic expectations on the abilities of Foveated Rendering. It's not a miracle optimization, definitely not by itself. And the video that was used as reference was based on having a dedicated AI chip (or powerful enough CPU) that could recreate a scene using 1/20 of the pixels in less than a frame with just a small detail loss.
Gen 2 doesn't start until you don't have to strap something to your face or if you do it's as light and simple as a pair of glasses. Socialization and telepresence just won't be there until this happens.
Foveated rendering alone is a game changer for VR and traditional gaming as well. What is incorrect about your Vader Immortal example is that since developers don't know where you are looking at in front of you, they still have to render most of the pixels. Adding fast eye tracking will eventually allow developers to know exactly where are you looking, and disregards the rest of the image. See the images below so you can know how much can be removed without you noticing any changes.
Full Res Image
95% of the pixels removed. See how few pixels are necessary to fill what you can see in detail.
Missing Pixels filled by Machine learning (You wont notice as the pixels that you are focusing on will be rendered in full detail).
We already have examples of headsets like the Vive Pro Eye with real Eye tracking and Foveated rendering, but they still don't track the eye fast enough (100hz) and because of that they have to leave a large margin of error to render in full detail because of the slow eye tracking. Below is the margin of error for the Vive Pro Eye. The green circle is full resolution and then it goes down from there. The goal is to only render the same amount of pixel on the example above.
Last edited:
And again, that requires a very powerful CPU or dedicated AI to be able to fill in the detail of that image, as well as a really high resolution display to even have that much detail to not render while still rendering a good amount of detail in front of your eye. That's not possible without really high end systems, which means this is for making very high end devices get much better visuals, it's not a miracle optimization for low end devices. When Oculus has talked about Foveated Rendering, they always talk about it for making truly believable visuals, not for running VR on mobile chipsets. You aren't going to see much better performance out of the current Oculus Quest than what Fixed Foveated Rendering gives even if you add a good high refresh eye tracker, definitely not "better than PS4/Xbox One".Foveated rendering alone is a game changer for VR and traditional gaming as well. What is incorrect about your Vader Immortal example is that since they don't know where you are looking at in front of you, they still have to render most of the pixels. Adding fast eye tracking will eventually allow developers to know exactly where are you looking, and disregards the rest of the image. See the images below so you can know how much can be removed without you noticing any changes.
Full Res Image
95% of the pixels removed. See how few pixels are necessary to fill what you can see in detail.
Missing Pixels filled by Machine learning (You wont notice as the pixels that you are focusing on will be rendered in full detail).
OP
OP
That's not how generations work, and it won't need to be a literal part of glasses before socialization and telepresence takes off. Once it's at a lightweight visor form factor, people will start to buy in to a much higher degree because the comfort should at that point enable comfortable use for hours even for average people. The value is what will make people want to wear it, hence why telepresence and other killer apps are important.
Is it possible the oculus name is holding the product back in terms of sales ? Wouldn't incorporating Facebook in there somewhere increase public visibility.
Anything to improve rendering is great, but it's not really what is going to make VR sell. There are two key things to overcome. And they're BIG things to over come.
First, you need virtual worlds and characters that can respond appropriately at whatever behaviours and actions the player performs in them. And that's a massive undertaking and not happening any time soon.
And then there's the problem with movement - your virtual movement is frequently limited by physical boundaries, resulting in either virtual movement (teleporting, analogue stick), giant slippy walking peripherals (tm) or alternative movement mechanics that will never be as intuitive as walking or running naturally. I honestly don't know how this one will be solved, but I hope there are some tricks we can play on the mind which means I can feel as if I'm exploring worlds of any size without having to leave my house.
First, you need virtual worlds and characters that can respond appropriately at whatever behaviours and actions the player performs in them. And that's a massive undertaking and not happening any time soon.
And then there's the problem with movement - your virtual movement is frequently limited by physical boundaries, resulting in either virtual movement (teleporting, analogue stick), giant slippy walking peripherals (tm) or alternative movement mechanics that will never be as intuitive as walking or running naturally. I honestly don't know how this one will be solved, but I hope there are some tricks we can play on the mind which means I can feel as if I'm exploring worlds of any size without having to leave my house.
OP
OP
I've thought about this, but I feel like that if they ever do that, they'll wait until they can produce a truly mainstream capable device like a Quest 3 and then advertise it everywhere on Facebook.
OP
OP
With a 2nd gen headset that has eye, facial, hand, and body tracking, you'll have almost a mind-reading device. That will enable AI to react to you on very personal levels. So it'll be possible, it just depends on how much work devs are able to put in. At the very least, expect weeping angels.
Boneworks is the best example of a physical world. Everything responds to the player physically. Even the enemies run entirely off physics.
Movement doesn't need to be perfected for VR to sell, but it will certainly help. Future solutions include galvanic vestibular stimulation, drowning the vestibular system in white noise, rollerskates, walking on the spot, redirected walking, and treadmills.
Some of these are more promising than others. White noise won't actually make you feel the movement, but will eliminate motion sickness in nearly all cases if the current studies are to be believed.
The closest term solution would be run-on-the-spot perhaps combined with the white noise mentioned earlier combined with full body tracking, and additionally have a consistent locomotion system for when you're not on the ground. Like when you're swimming or floating in space since those are solved areas now.
The overhype. Every single time there's never metered and reasonably low expectations that are met. Its always "This is going to be MASSIVE the first time people use it, GAME CHANGER" And then we get the previous 4 years.
Is it progress? Yes. Is it what the hype jobs were predicting it would be? No...
I wish the links to GAF weren't banned so I could link those years old threads as examples of some of the over the top predictions.
You said the following:
"You want to know how foveated rendering would make the current Quest perform? Play Vader Immortal, it uses fixed foveated rendering, which just means it's foveated rendering that always assumes you are looking straight. Adding eye tracking wouldn't increase performance much from that"
This is how fixed foveated rendering works on the Quest.
As you can see there is still a large percentage of the screen that still needs to be rendered at full resolution. If the Quest could do eye tracking developers would be able to know exactly where you are looking and render a lot more of the screen at 1/4, 1/8 or 1/16 of the full resolution. There are also a lot of games that don't render at native resolution on the Quest, that problem would be solved with eye tracking. The Michael Abrash example is basically the end goal where you render the minimum amount of pixels needed, but that doesn't mean there will not be massive performance savings until we get to that point with other foveated rendering solutions. Your statement about how adding eye tracking to the Quest wouldn't increase performance much is false. You could have a 1080P screen and not having to render 60% or 70% of the pixels at full resolution would be a massive performance increase and battery savings for a device like the Quest.
Last edited:
The fact that a few insiders might have been a bit over-enthusiastic a few years ago doesn't negate the incredible achievement that is the Quest.
Just 3 years ago we had a $600 Rift which required separate $200 controllers (released 7-8 months later) and outside-in trackers, and an extremely powerful PC for 2016, with an obnoxiously complicated setup and a cord always in your way.
A month ago I spent $400 for an all-inclusive headset which included touch controllers and had it up and running playing Beat Saber and SUPERHOT within 10 minutes of taking it out of the box.
This progress is nothing short of mind blowing. I can't wait to see what comes in 3 more years.
Just 3 years ago we had a $600 Rift which required separate $200 controllers (released 7-8 months later) and outside-in trackers, and an extremely powerful PC for 2016, with an obnoxiously complicated setup and a cord always in your way.
A month ago I spent $400 for an all-inclusive headset which included touch controllers and had it up and running playing Beat Saber and SUPERHOT within 10 minutes of taking it out of the box.
This progress is nothing short of mind blowing. I can't wait to see what comes in 3 more years.
That's not going to make the existing Oculus Quest faster than a PS4/Xbox One, which is what I was countering.You said the following:
"You want to know how foveated rendering would make the current Quest perform? Play Vader Immortal, it uses fixed foveated rendering, which just means it's foveated rendering that always assumes you are looking straight. Adding eye tracking wouldn't increase performance much from that"
This is how fixed foveated rendering works on the Quest.
As you can see there is still a large percentage of the screen that still needs to be rendered at full resolution. If the Quest could do eye tracking developers would be able to know exactly where you are looking and render a lot more of the screen at 1/4, 1/8 or 1/16 of the full resolution. There are also a lot of games that don't render at native resolution on the Quest, that problem would be solved with eye tracking. The Michael Abrash example is basically the end goal where you render the minimum amount of pixels needed, but that doesn't mean there will not be massive performance savings until we get to that point with other foveated rendering solutions. Your statement about how adding eye tracking to the Quest wouldn't increase performance much is false. You could have a 1080P screen and not having to render 60% or 70% of the pixels at full resolution would be a massive performance increase and battery savings for a device like the Quest.
I agree with that. What I was refuting was the "Adding eye tracking wouldn't increase performance much from that", part of your comment.
Software advancements are not hardware advancements. Slapping some cellphone guts and cameras into a modified Rift was the easy part (Qualcomm did most of that work), making it map/track space and run meaningful experiences was the hard part. Oculus' software has always been industry leading and generations ahead of everyone else. ATW/ASW, Home, and now their inside out tracking shit all over anything Valve, Sony, or MS have done from a great height. There is no doubt their future inventions will be similarly uncontested. The problem, to use your example, is that the brand new Quest hardware is not very different from that now ancient (in CE term) more than 3 year old CV1. The Quest is much less comfortable (due to the increased weight and front heavy imbalance), has significantly downgraded audio, no FOV increase at all, SDR pentile displays, and only a modest uptick in resolution. Functionally, very little has changed from that $800 2016 Rift kit, except you don't have to set up webcams or own a "gaming PC"--but to be fair that also comes with significant trade offs like noticeably downgraded visuals and many experiences that won't run at all (i.e. most of the "AAA" VR games).
Last edited:
Gen 1 VR already gave me the best gaming experiences i had for many years, I can only salivate at the prospect when we reach Quest levels of untethered with better visuals, Eye tracking, FoV, Finger tracking and full body motion.
And yet here i am still amazed by Farpoint, Lone Echo and soon to be No Mans Sky.
Star Citizen/Cyberpunk 2077 in VR is the ultimate future. Maybe Gen 4 of VR.
People still comparing it to 3DTV................I dont remember it getting a gen2, Considering it was still panel gaming and nothing like VR.
And yet here i am still amazed by Farpoint, Lone Echo and soon to be No Mans Sky.
Star Citizen/Cyberpunk 2077 in VR is the ultimate future. Maybe Gen 4 of VR.
People still comparing it to 3DTV................I dont remember it getting a gen2, Considering it was still panel gaming and nothing like VR.
What matters to me is the experience I have in the end, as a user and customer of VR.
The experience has improved so drastically in just a few short years, despite the fact that hardware advancement takes a long time, that it makes me even more excited for the road ahead once hardware has a chance to catch up too.
In the meantime, I'm so utterly astounded by the progress made in VR that I cannot wait to see what comes next.
OP
OP
Star Citizen already has planned VR support. It's a matter of whether the game releases anytime soon.
Well by the time it does i would assume the VR headset will be untethered and in glasses form, Nay maybe even a contact lense.
Which is great, but as someone who had a gaming PC and desk to place a couple cameras, the experience for me hasn't changed at all and is in fact actually a regression in many respects.
I have no doubt the tech could/will get "there" reasonably soon, but at what cost? Until they can make a stand alone headset that will blow everyone's minds for sub-$200, I just can't see it catching on with the mainstream. I even think $200 is a lot of money for people to spend on something that they can't really see without putting a headset on, and it would be incredibly expensive to set up kiosks EVERYWHERE with new headsets for each person (because trying on a headset that someone else just had on their head is something that a lot (most?) of people won't be willing to do). I just don't get how you make enough headsets where everyone can try it before they buy it without trying one that someone else already had on and do it all for a price point that the general public is going to accept. I want to see it happen. I want to be convinced so badly, but again, I'm not going to try on someone else's headset, so I don't get how you convince me how good it is. A sub-$100 price point and I'd probably buy one without needing to try it on first...provided the reviews were absolutely glowing.
OP
OP
You wouldn't pay $200 for a decent phone or a PC. VR isn't just for gaming. The type of VR we'll have in the next 5-10 years will be offering thousands of dollars worth of value just on the basis of the physical goods it replaces. To say nothing of the travel expenses/times it will save, or it's many other uses.
People will accept $600+ price points without much opposition once they see the value is there. Those who want less premium versions can opt for the $150-200 versions.
You make a good point though that marketing VR is really hard. What if you were offered a disposable face mask to try on a headset at X convention? Would you do it then?
A lot of people are still willing to try the device after they see people have fun with it in person. The bigger issue in my mind is getting enough headsets around, which is getting a bit better with places like Best Buy.
Another thing to consider is that games like Boneworks and Blade and Sorcery show that you can market 1st person VR quite well through videos if done a certain way.
I think I should introduce you to a very interesting concept called showing a device to friends and family.
Turns out my Mom has never tried a quest at a kiosk, no, she tried mine and then bought one.
You don't need demos everywhere, you just need more people to buy the tech and share it with your friends.
As for blowing people's minds, Beat Saber and Super Hot are already doing that on the Quest, and in 2-3 years that should be available for $200.
To your first point, the difference is that I need a phone and PC for work. Maybe eventually owning a painting company will require me to also own a VR headset, but I don't see that being the case in the next 5-10 years. Maybe a person can convince themselves that they are "traveling" instead of just wearing a VR headset, but I can't imagine the sad, lonely existence of taking a week off work to spend a week in Tokyo in VR. VR isn't and will never be replacement for going out and exploring the world.
I want the premium version, and I want it for less than $200.
As to your other point, there are no "conventions" around me where I would be able to try it on, and even still it just seems like it's pretty gross even with a disposable face mask. Maybe I could be convinced if I were like the second person to use it out of the box, but my fiancee wouldn't. Best Buy, at this point, is not a store that most people ever go into. You would need kiosks everywhere. Like public sidewalks, parks, and other non-retail locations since let's be honest retail is going to be changing substantially in the next 5-10 years.
Like I say, I want to be convinced, but the only thing that's going to convince me is trying it, (and we've discussed the restrictions around that), and the price point needing to be substantially lower.
OP
OP
You're going to need a VR/AR headset for work in the future too. Phones and PCs didn't start off immediately being required. As for travel, the idea is that you don't waste time driving to work and just visit a virtual office instead, one that people would find more enticing than a real office, with coworkers represented as avatars indistinguishable from their real selves. Granted, this is for office based work. On-site, you'll be using VR for training.
You also underestimate how powerful travelling via VR can be. People who like travelling in real life will likely appreciate what it offers. There's no end to the amount of people that cry touring their old house in Google Earth, despite it being far from real in terms of specs and what-not. You may also want to visit concerts, sporting events, friends. The list goes on. Most people just can't do travelling that much, and VR will offer that ability for them.
It isn't a replacement, but it's very much going to be a valid alternative, or a reality-bender since you can't travel everywhere at once and see a different concert in a different country every night.
I also think you're overshooting the amount of kiosks needed. It doesn't need to be that drastic.
Don't understand the logic behind the "BLOWS EVERYONE'S MIND", why does it have to be sub-$200 in order to work? Smartphones, consoles, gaming PC's cost a lot more than that and sell hundreds of millions of units, so don't know why it can't be the same for a VR headset, especially if it eventually becomes a "Blows everyone's minds" type of product. All of the headsets that are out now become a way to promote the technology and that will only increase over time. VR is a tecnology that you instantly want to share as soon as you try it and the Oculus Quest will help a lot with that. May I ask why you won't try someone else headset?
Last edited:
If it becomes something that everyone needs to own like 200 years from now, they will also be affordable enough that you'll be able to pick one up at 7-11 for less than $100 much like phones now.
Anyway, you guys can quit with the sales pitches. I've explained what it will take for me to jump into VR. If it's that important to you guys that VR does well than just go buy multiple VR headsets. Problem solved.
Are you coming back to even try to defend your embarrassing "take"?
I guess not...
I too want a flagship product for 50-80%+ off >.>
If that is a deal breaker for you I highly suggest you stop looking at anything VR period right now.
It's going to be a long time before anything remotely close to what you want happens because it sounds like even if the Quest gets down to $200 in 3 years you still won't want it because it won't have the new features the new headset has at $400-$500
and that will rinse and repeat for decades as this is newer tech with lots of room for advancement right now.
As to this. Congrats on having money to travel.
I also don't understand why you would take a week off to spend a week in Tokyo in VR. The beauty of VR is that you can get an experience instantly. So you can just check out some places after work and head into work the next day.
Not everyone can afford to travel the world and experience it first hand. Vr tours aren't meant so much to be replacements as much as affordable alternatives.
You are more realistically looking at that possibility in 40-50 years if you go by how cell phones went :P
OP
OP
Woah now. 200 years is a bit steep. I was talking in the context of like 10 years.
Also, all I'm doing is trying to explain to you why a price higher than $200 is perfectly valid for it to take off. Products sell based on the value they give. You may wait until sub $200 prices if that's what you wish for, but that's because you've been undervaluing what the tech can do.
I think you are being just about as unrealistic with the time frame as they are.
There is no chance at all that there will be widespread every day use of VR in 10 years. absolute min would be like 20 years but even that is likely pushing it.
Hell tons of companies still don't let employees work at home even though all the coding and tech support could be handled at home because they are worried about data leaks, and VR would do nothing to help solve that, let alone setting up stuff like virtual work places.
Black and white to colour perhaps? Analogue to digital?
That sort of thing will also be attractive for people who went to the actual places.
Like with museums, VR could let you hold and inspect the artifacts like supermarket produce.
Oh absolutely. I went to Scotland a few years back and I would love it if I could get a nice VR version of some of the countryside because getting that sense of scale and size in traditional pictures was impossible.
OP
OP
10-15. Given the vast progress going on, especially with photorealistic avatars, I feel this is not unreasonable.
Question is, should a company be worried about data leaks or about ongoing office finances? They can get rid of most of their physical space in order to save money, employ a wider range of people, and enable better convenience for their employees and thus enable greater satisfaction.
Different companies will prioritize different things.
They could do all of your thought right now.
There isn't a need to drag in people to an office to make them work 9-5 for the majority of office jobs anymore but they still do.
VR isn't going to change that any time soon. The ability to see other people at work isn't the thing holding back companies moving to mostly people working from home.
You are also forgetting not everyone will want to do that.
I know I sure as hell wouldn't want to work anywhere that is having me use a photorealistic avatar and expecting me to be hooked up to VR for 8 hours a day.
I don't care if VR advances to mere glasses that I can wear I wouldn't want that.
What you call better convenience for employees I call a way that your work will make sure that you can be put to work no matter where you are.
Hell a lot of companies expect you to respond to emails in your personal life as it is.
and yes, some companies might do what you are saying, but there is zero chance at all it will be widespread and common in 10-15 years even if we wake up tomorrow and VR tech that is advanced 50 years is suddenly available. Way more factors than just the quality of VR in play for something like that.
Yeah it's insane the leap we will see. Legit better visuals in VR than not in VR which could make developers sit up and take notice. Also they "could" let 2D games be played on a virtual VR screen with all that extra horsepower used for the 2D visuals as well.
VR (and AR) are gonna get crazy within a decade, or more like 5 years.
You seem to be reading talk about foveated rendering being useful for taking advantage of higher resolution panels and assuming that it's only useful for that.
You also don't seem to know just how tiny the sweet spots of our eyes are, or how terrible our vision is outside of them.
Even the portion of our FoV where we're completely blind is actually much larger than the high resolution portion.
Last edited:
OP
OP
There's less reason to be doing it the more advanced VR becomes.
Photorealistic avatars of real people is just an example. They could keep it anonymous if wanted.
There's less reason to do it the more advanced computers become in general. Still hasn't been a huge trend of companies changing their ways.
Companies are typically very slow to adapt.
You're underestimating the collaborative functionality of presence, which as a technology VR affords and other technology does not. It's still in an early phase, but it's rapidly advancing. Sure, traditional companies will remain that way, but just like significantly more jobs are remote than there were 10 years ago, 10 years from now VR collaboration will continue to catalyze that trend.