D
Deleted member 13524
Guest
There's two console targets for all titles. Why would VR be different?
Sony VR Headset/Project Morpheus/PlayStation VR
There's two console targets for all titles. Why would VR be different?
Because it was supposed to be one target. Now PSVR isn't a single target but two. The console space has been fragmented into XB1, PS4, Scorpio, PS4Pro, PSVR and PSVR Pro. The sequel consoles weren't a problem when they were just running the same game code only faster, but targeting them is requiring more work from the devs for no net gain.
If I were creating a VR game, I'd completely ignore PS4Pro's specialist functionality as a waste of money. Only if PSVR on PS4Pro is the larger market would it be worth targeting specifically.
If I were creating a VR game, I'd completely ignore PS4Pro's specialist functionality as a waste of money. Only if PSVR on PS4Pro is the larger market would it be worth targeting specifically.
upnorthsox
Veteran
If I were creating a VR game, I'd completely ignore the XB1 and Scorpio as a waste of money.
Of course you would because they don't even have a VR option. But what would you do about PS4 and PS4Pro versions? How much would you be willing to spend to design a PS4 Pro specific engine (or modifications thereof) to target a subset of a subset? Historically optional power upgrades have been well underutilised on consoles for exactly that reason. The hope before we knew what was really going on was compatible hardware clocked higher just running the same code fast, which could be targeted with a few parameter tweaks to use different assets. But now we have something more akin a new console, needing specific code. But it's not even a new console where you can severe ties with the old code base because everyone's moving on. It's just a decidedly awkward position for devs to be in. And it means PSVR could end up fractured too, with PS4Pro specific VR titles, in the same box as the PS4 game because PS4Pro is just a PS4, only it has PS4 Pro exclusives...it's getting messier and messier.
D
Deleted member 13524
Guest
Are you aware that most games/experiences/demos with PSVR are actually PSVR exclusive?
Look at the list of PSVR games/demos:
https://blog.eu.playstation.com/2016/10/05/heres-the-playstation-vr-software-launch-line-up-in-full/
Number of PSVR "compatible" titles: 14
Number of PSVR "required" titles: 40
And if you look at the "compatible" titles, some of them are actually small DLCs of current existing games (e.g. Tomb Raider's and Battlefront's VR level) that were made in VR to test the waters but are otherwise mostly uninteresting to play in a TV.
I think what we'll have in the end is:
1 - TV-only games
2 - VR-only games
3 - VR-or-TV games.
4 - VR-and-TV games.
5 - TV-only games with spontaneous (non-required) VR mini-games
- TV-only will be the ones we've had so far, and those will have two render targets: Pro and regular. These will be most FPS, 3rd-person action, side-scrollers, etc.
- VR-only will have again only two render targets: Pro and regular. These will be the kind of games we see in that list above.
- VR-or-TV will in fact have 4 render targets, though the amount of games in this situation are probably very few. We're probably only looking at the car and spaceship simulators that choose to support VR.
- VR-and-TV will be mostly party games that use the social screen. Here the VR is still mandatory, so there's only 2 targets again.
- TV-only with VR mini-games (like e.g. Tomb Raider) will probably use VR in very limited quantities, perhaps even for exclusive content, so the amount of effort for having 4 targets is greatly reduced.
In the end, the only ones having an actual problem with VR and Pro are the devs making simulators, but those are probably also the ones who can gain the most with the added features.
If I'm split between two similar spaceship simulators, I know I'll choose the one with VR support and better graphics on the Pro in VR.
As for XBone and Scorpio.. I don't know about Scorpio but I don't think the XBone will ever get VR in any way.
And by the time Scorpio gets out, both Oculus and Steam may have released 2nd-gen cheaper headsets (most probably seated-only), so Microsoft may be wise enough to just officially support one of those.
Yes. What difference does that make? Currently devs have been targeting PSVR as a single platform on a single hardware with a potentially sizeable install base. Now they're being asked to split development across two machines, so you have to make two flavours of PSVR games to sell the same number of headsets. That is, is PS4Pro didn't exist and Sony sold 5 million PSVRs, you'd write PS4 code for a potential audience of 5 million. The introduction of 4Pro means that 5 million is now split across two consoles which need specific code. What's in it for the devs?
I think you misread/misunderstood what I was saying? What we may have now is PS4Pro only PSVR titles. You can get a PSVR this Christmas in the legitimate belief it'll play all PSVR titles, because PS4 Pro is just a PS4 and isn't getting exclusives, only to find next year a few PS4 Pro only PSVR titles. Or, more likely due to economics, the extra VR features of PS4Pro will go largely unused because there's no reason to spend time optimising for it.
D
Deleted member 13524
Guest
So what?
Do you want devs to not take advantage of the Pro in VR titles?
Are PSVR devs forced to make a Pro target? You may have a point if they are, but I haven't seen such statement anywhere.
Sales. Specifically, more sales in console form than PC form, because the former are more profitable. That's ultimately what the Pro is about, according to Sony (either people are willing to believe Sony's statements or not is a whole other story though).
I'm pretty sure Sony will forbid that, just like they've forbidden Pro-only titles for TV.
If they're not, why would they choose to add work?
How does PS4Pro PSVR equate to more sales? If PS4 Pro didn't exist, would those customers not buy PSVR? I'm suggesting we've gone from 5 million PS4 PSVR buyers to 3 million PS4 owners and PS4Pro owners. What's the argument that the total PSVR install base will now be greater, so maybe 3+3 or 4 million thanks to the existence of the Pro and its specific VR enhancements?
I think PSVR enhancements going from a PS4 -> PS4pro aren't likely to be that time intensive. An increased buffer size by itself should be a relatively trivial change, while being quite a noticeable visual improvement. Other optimizations like multi-res buffers would require a bit more engine-level tinkering, but even then I'm not sure they'll be all that big of a deal for the bigger developers to integrate. Whether it's Epic, Unity, Crytek or EA/Frostbite, there's a pretty big incentive for these companies to be spending R&D time on exploring and implementing VR optimization techniques as they're going to be equally applicable across all future forms of AR/VR rendering for the coming years.
D
Deleted member 13524
Guest
Because better experience = more sales.
Considering PS4 convinces X number of customers, PS4 Pro convinces Y number of customers, Y and X being mutually exclusive and greater than zero.
The fairest point there is the middleware targeting it. I expect everyone targeting VR at the moment is using middleware, so that moves the problem to them. But still, given the hardware optimisations enabling more stuff, how do you create, test and implement those cost effectively? It's still the case that any PSVR game now needs a second development thread. It might not be a problem if the games are targeting PCVR too, but I still ask what the ROI benefits are for developers. Saying, "it's not going to add too much to their costs," isn't really a great motivating factor if the returns don't increase!
Why would Y not be a portion of X? That makes no sense to me. Before PS4 Pro existed, there were X million people interested in PSVR on PS4. PSVR has been specifically designed and managed to provide a good experience, and all accounts are PSVR is a good experience. 4Pro now adds a bit more to that experience, but how many people are going to be swayed by that option who wouldn't otherwise have been content with PS4 and PSVR? If 4Pro is ten times better, sure, but that seems implausible. If the difference is just the loss of time-warping artefacts in the periphery, are sales really going to increase?
D
Deleted member 13524
Guest
Because Y is an amount of people who aren't convinced with PS4 + PSVR quality graphics, but are with PS4 Pro + PSVR.
I will get back to you on this thing next Thursday when I get my PSVR set up at home, but it seems to me that since the render resolution on the PSVR is rather low, aliasing might be a bigger problem than time-warping artifacts.
And if the Pro will be working on the exact same assets as the regular PS4, it will have a ton of freely available CUs for compute-based antialiasing.
D
Deleted member 13524
Guest
Supersampling may not be a great idea because there's not that much more (theoretical at least) fillrate and bandwidth in the Pro.
If I already had a PS4 and could spend 400€ on it this Christmas, I'd definitely go for PSVR instead of the Pro.
But if I had only 400€ to spend and no console, I'd go for the Pro.
Me.. I'm probably going to spend ~200€ on upgrading my PS4 to the PS4 Pro (400€ Pro minus 200€ PS4 2nd-hand) and I don't have a 4K TV, I have a PSVR pre-ordered.
By supersampling he meant using a higher resolution on the Pro (like 2x more pixels with similar performance), which is exactly what the Pro is good at based on the Pro games we have seen so far, compared to OG PS4 versions.
So this will dramatically reduce the aliasing seen with the standard PS4 games.
D
Deleted member 13524
Guest
I know, but rendering at a higher resolution is preferable for a display panel that supports a higher resolution (e.g. 4K TV). Compute-based AA methods are generally less intensive and provide better performance for lower-resolution displays.
The Pro still has that same 60 FPS minimum target
According to rifters, SS for VR is muuuuuch better tho. It also probably easier to do without complications
I think you guys should read through this
http://media.steampowered.com/apps/valve/2015/Alex_Vlachos_Advanced_VR_Rendering_GDC2015.pdf
Valve Recommends 4X MSAA minimum with 8x if they can. They also recommend raicing the render target scaler up on faster GPUS .
I am sure Sony will use both to make a better image for the user
http://media.steampowered.com/apps/valve/2015/Alex_Vlachos_Advanced_VR_Rendering_GDC2015.pdf
Valve Recommends 4X MSAA minimum with 8x if they can. They also recommend raicing the render target scaler up on faster GPUS .
I am sure Sony will use both to make a better image for the user
PSVR teardown by SIE
https://www.wired.com/2016/10/psvr-teardown-video/
https://www.wired.com/2016/10/psvr-teardown-video/
Similar threads
