Kinect and Oculus Rift
- Thread starter baten
- Start date
This is why I said that Oculus having an anti-console stance is in the eye of the beholder, because what Palmer Luckey says here is obvious truth to me.
Oculus is trying to drive VR-technology and acceptance. The won't get their first product out until probably some time in 2015, and it will take a few years to hopefully gain support and interest to have a decent momentum a few years after that, in time for updated hardware that can take advantage of both software and hardware developments and extended feedback from developing community and consumers. If the technology starts to come into its own around 2018, they've done well.
How can the Xbox One help them with that? (Sony has their own plans). Disregard the problems of reaching out to developers through the closed platform, having APIs available for it, and the iffy prospects of selling a $400 peripheral for a $400 console to an already limited consumer base. Even from a pure hardware capabilities standpoint - the first consumer version of the Rift aims for (still not locked down) a 2560x1440 display, and a refresh rate of 120 Hz. Even with halved resolution per eye, you are still looking at demands that are five times or so higher than 792p30Hz that Watch Dogs will run at on the XB1. So the visuals would need to be scaled down to roughly on par with Xbox 360, only at higher resolution and frame rate. And that is where it would stay until 2020 or so when a new Xbox may or may not be introduced. How can being locked at approximate 2005 level visuals be a viable path when trying to establish VR in the marketplace?
In terms of graphics prowess the Xbox One is roughly on par with the HD4870 I bought almost six years ago now. My estimate is that VR won't be ready for prime time until we are at the 10nm lithographic node. What's a reasonable performance assumption of a graphics solution at that node, if it is to be coupled with a $300-400 peripheral?
Even if we disregard all other issues, (which is foolish), the Xbox One just doesn't have the graphical grunt to drive VR into mainstream acceptance.
So "dumb things" to you, is "perfectly sensible" to me. Given the other issues involved, I just don't see what other viable path was available to Oculus. Hell, it remains to be shown that even this one is.
D
Deleted member 11852
Guest
He's not anti-console, it's simply that consoles don't fit their business model which is to iterate and sell a new version of Rift on an annual basis - he's said this several times. To do that he needs the host hardware to iterate as well. Console owners aren't going to buy Rift 2 with the higher resolution display and lower latency if the console can't drive it to make it worthwhile.
The Xbox One isn't a PC. It is virtually impossible for a win32 program to extract anything close to the theoretical performance from the graphics system because of all the layers. DirectX12 will help, but I can't see it completely closing the gap relative to a console because even if you can see every discrete hardware function, nobody is ever going to optimise for the hundreds of variations of graphics hardware like they can, and do, on consoles.
However were Microsoft to partner with Occulus, they wouldn't be limited to building an API to work around DirectX/OpenGL, they could imbed the VR APIs right in the graphics system, saving all sorts of overhead.
The dumb thing was what I quoted, which is why I quoted it. Not everything he says is dumb but he's young and has the brashness of youth. It's not consoles lack the power, Sony have demonstrated VR on PlayStation 4 at GDC. Lots of people used it. But doesn't complement their business model.
Of course the business model may now change with Facebook ownership.
Sorry. Is the OR audience actually expected to upgrade their machine every year? What sort of audience are they actually targeting? Sounds like they are marginalising themselves to a tiny niche (no doubt overruled by Facebook who want this tech everywhere and used by everyone).
VR on consoles won't be as good an experience as on a highend PC (middle end as the years progress) but neither are the games. As long as its good enough for the consumer, and with this market you're looking at more mainstream priced to be good enough, it'll have a market. Joe Gamer spending $400 in 2016 to get simple but effective VR on console is far more likely than Joe Gamer spending $800+ and then $100 every year to upgrade to get the latest, greatest VR visuals on PC.
D
Deleted member 11852
Guest
Perhaps not every year but frequently
I mean take a look at the market they are targeting: people who own a PC, game on it, and have a good enough PC to output a decent image in stereo, and are interested in whatever games are released to support OR. That's über niche.
If the expectation is that folks will be upgrading their VR headset as infrequently as they may upgrade their monitor then that's some slim pickins. Before Morpheus maybe they could have sold at a premium for a unique experience but Sony will sell for as low as possible.
Yup. PCs generally require investment (upgrades) to get better, consoles generally require time (familiarity with the architecture, optimisations). I don't know about Microsoft but what the generally positive reaction of journalists who used Morpheus, that'll be more than good enough. Microsoft, if they partner with Occulus, could leverage their benefits of the tech, and combine it with the benefits of a console.
This is incorrect. Console hardware does not get better. Games get more efficient and make progressively better use of the hardware feature set. This applies to PC's as well as consoles. It can apply to consoles to a greater degree but that doesn't preclude the same phenomenon of ever improving graphics on a fixed hardware platform signifcantly effecting the PC as well as consoles.
When I bought my GTX670 2 years ago the best looking games at the time were Battlefield 3 and Crysis 2. Today I can play Battlefield 4 and Crysis 3, in a few months I'll be playing Assasins Creed Unity and Batman Arkham Knight. Will those games have significantly better graphics than what I was enjoying 3 years prior? Absolutely they will. Did I need to upgrade my hardware to experience those better graphics? Nope. Will I be playing even better looking games a year or two after that on the same hardware? I certainly could if I wanted. Odds are though I'll choose to upgrade to get even better graphics than the 670 could deliver in those games which I assume is what Shifty was referring to with the £100 a year comment. PC gamers usually upgrade out of choice rather than necessity.
I would assume it goes like this .
2015 we get Oculus rift . This has all the bells and whistles its priced higher than grannie will want to spend to surf facebook. But early adopters drive the price of this down over the cost of a few years. They pump out hopefully millions of them.
Each year for the hardcore you will get a new screen and what not perhaps wireless power then full body tracking.
But each year that original hardcore model sells more and drops further down in price until its in the $100 oh grannie needs this for Christmas or May the 4th be with you.
D
Deleted member 11852
Guest
Dude, seriously, that wasn't obvious from context? Did you really need to post this because you thought I was suggesting console hardware in some way improved? :-/
D
Deleted member 13524
Guest
Obviously you know console hardware cant improve over time, I just phrased my reply in that manner to highlight the absurdity of the concept that PC graphics can't also improve without hardware upgrades. Considering both platforms run the same games that doesn't make a lot of sense.
Well, I've seen/read a number of intervues of Palmer Luckey, and I can't say that I recognize your assertion that they will churn out VR headsets at a brisk pace.
(Having said that, starting up, there will be opportunities for technical improvements in terms of screens, lenses, interfaces and so on, and later diversification can focus on size/weight/price points and design. In the very long run, headphones are just as relevant an analogy as TV-screens.)
In terms of graphics requirements however, not too much more can happen, unless the underlying paradigm changes. If we use my example from above, and compare the HD4870 with the R9 290x, six years have bought us a factor of three to four in ALU power and bandwidth. At roughly twice the power draw. If that pace holds true going forward, then in 2020, we would have another factor of 3-4 at, yet again, another factor of two in power draw. That's not a blistering pace, and again this assumes that the pace of lithographic advancement doesn't keep slowing down going forward.
So there is every reason to believe that, in terms of graphics processing load, we'll be looking at no higher than split 3960x2160 (or at most 5120x2880) at 120 Hz for a very long time indeed. We simply won't have hardware to drive it.
Now Luckey is an extremely technology optimistic individual, who assumes PC-tech to take great strides going forward, and of course if you have that fundamental outlook, standing still at Xbox One levels of performance for the next 3-5-8-? years is a horrible prospect.
Rather than discuss Luckey, let me quote him directly from this pretty good interview:
Sorry about the long quotes, but I think they demonstrate their strategy clearly. They focus on PC now because they are in a technology/concept development phase. Really long term, they aim for the even more ubiquitous platform, cell phones. Consoles, being low volume, fixed and closed IP platforms, just doesn't make sense to target. The platform holders may choose to add VR support to their platforms at some point, but that is not for Oculus to decide.
He is way more optimistic about future performance levels on PC than I am. He is also much younger, it may have to do with testosterone levels.
Seems just a little anti-console to me. IF he's intending to release on mobile in 'years', in 'years' mobile will get to PS4 level. If that's good enough for mobile, it's good enough for consoles. If that's not good enough for mobile, it'll be another 8 years after that before mobile becomes good enough (assuming we don't hit a wall). Is he just going to ignore mobile for all that time because the quality isn't up to high-end 2015 PC standards? Well, considering how they keep not releasing a consumer product because they keep waiting for it to get better, maybe that's the plan. But then someone else swill steal their thunder, offering a 'good enough' experience.
The consoles shouldn't be ignored, even if the games are visually simple. I think the only problem with Kinect and OR would be latency.
The consoles shouldn't be ignored, even if the games are visually simple. I think the only problem with Kinect and OR would be latency.
From a purely technical point of view, you're correct at the moment, although I suspect that the consumer version of the Rift will use displayport 1.2 because HDMI 1.4 (Xbox One) have issues with both 120Hz (1.4a) and overall bandwidth (1.4a and b). The greater problem is if there will be content on the Xbox One that justifies adding a comparatively expensive peripheral, and if the development environment Microsoft provides would be VR-friendly. If Microsoft wants the Rift on the Xbox, they could just ask Oculus to make a version for the console. However, if they don't do that, can Oculus really target the Xbox without Microsofts expressed collaboration? And seeing as Microsoft has already ditched their own gamble for next-generation interactivity, what are the odds that they would go out of their way for a much more expensive add-on from a third party?
Oculus doesn't own the issue.
I can agree that it's not Occulus's position ot make it happen on consoles, but the reasons don't strike me as fair. If the response was, "consoles are a good target but it needs the console companies to make it happen,"fair enough, but the reply was, "consoles aren't good enough." As you say, the power requirements for the target are ridiculous. If mobile is a target, the current consoles are a great place to start at power far greater than mobile will have for years. OR is already supported in middleware like Unity. Just needs a console driver and interface.
Actually, one issue is probably console outputs. With only one HDMI out, you'd either need a switcher or disconnect it from the TV. The HDMI switcher could easily be included in the breakout box though. In fact it'd work really well with Kinect. "Xbox, show on Rift." "Xbox, show on TV."
Actually, one issue is probably console outputs. With only one HDMI out, you'd either need a switcher or disconnect it from the TV. The HDMI switcher could easily be included in the breakout box though. In fact it'd work really well with Kinect. "Xbox, show on Rift." "Xbox, show on TV."
As Luckey describes it, it is about phases - note that the cell phone scenario is far in the future, and that the main vision there is that you use the phone itself as screen/gyro/MEMS/camera/application platform/headphone driver. In such a VR future Oculus would probably only provide middleware (or apps!) and possibly a holder with lenses and whatever calibration might be necessary for a given phone.
It's clearly an interesting option - you can use your phone as usual, or for specific purposes, mount it and step into the virtual reality. But since that doesn't involve making display hardware, such a vision for the future doesn't really impact what they are trying to do with the RIft, which is provide a platform which allows them to grow acceptance of the concept in the first place. Actually, the first consumer version may be the only one where they can basically market it to enthusiasts exclusively and have relatively steep hardware requirements. It's not at all certain that it makes sense for them to target significantly higher end systems once they try to reach a wider audience. We'll see how it goes. But for the first version they can probably assume that their customers, prepared to pay a decent sum for a product with little or no content produced for it, will simply acquire whatever graphics solution is needed to drive the hardware well. They will be the evangelists that are needed to drum up wider interest. Most CV1 owners will be doing lots of demoing.
I agree that a marriage between the Rift and Kinect would be interesting to explore.
(Regardless of any personal interest in Kinect I think it's a loss to the gaming industry when innovative hardware is abandoned. Being able to provide experiences not available anywhere else can only be an advantage when competing for consumer attention and money.)
Oculus using displayport as an input seems like a given for CV1, but it would be really attractive to somehow get rid of cables completely. I think that is where the cell phone scenario gains a lot of attractiveness, and why you would be prepared to compromise to achieve it. Self contained VR, anywhere.
Similar threads
