More resolution is always going to be preferable theoretically, but I take it to mean you're saying why not use that additional performance just for super sampling or whatever, and yeah, that's certainly a fair point.
Sure, and I'm not disputing that, but at the same time, "moar graffx" is always going to be preferable, theoretically. That's why we get all of the lol@ps2gfx posts whenever PSVR comes up. There's lots of stuff you can do to make the picture prettier apart from increasing the resolution; better AA, more polys, better lighting, etc. All of that stuff will improve the look of your world, and by extension, increase believability. They also all
effectively come at the cost of resolution, by which I mean, if you double your pixel count, you now may not have time to compute the improved AA and stuff.
I think the GTA thing hits at the very center of the debate really. GTA might seem like the perfect VR game to some people, it presents a huge world, lovingly crafted. But in reality, it would be the very worst choice for VR.
I wasn't really cheerleading for GTA:VR in particular; it was just recognizable and easy to type. lol It does bring up an interesting point though.
They're right about games needing to be designed from the outset for VR. Now, everyone seems to be assuming that means saying goodbye to all the franchises that we know and love, but that's not necessarily the case. In GTA you assume the role of a street punk who hits it big in the underworld
. While you're likely correct about GTA5 making for a terrible VR game, that doesn't mean there can't be a great GTA VR game. I'd argue that if R* makes a game where you become a punk with dreams of taking over the mob, then that's GTA, whether it's a GTA5-style game, or a VR experience that makes you literally feel like you're the punk. In fact, the series has already undergone a similar re-imagining, when it transitioned from top-down to what we have now.
Speaking of best practices though, I think MMOs may become really huge in VR.
So we're supposed to rethink the way we design our games. Let's try that for a while.
The name of the game is presence. We want the player to literally feel like they've been transported in to the world we've created for them. We've already established some rules for establishing and maintaining that effect. Head motion is law. Latency is king. Sound matters. These are mostly just technical issues though. What about the game design itself? What can we do there to strengthen presence and minimize immersion breaks?
I think one thing most gamers will agree can kill the mood in an instant is shitty NPC interaction. A best they're crappy robots. At worst they're little more than buttons. Since they can't think like a real person, they can't respond like a real person, which is a huge limiter to immersion. In GTA, when you're captured, beaten, and dragged up before the Don who explains how he's going to kill you, as a player, you laugh and say, "Yeah, yeah. Blow me, bitch."
Clearly, that player doesn't feel present at that moment. So how do we avoid it? The obvious solution would be to remove all of the NPCs, but we can't have every game be about the player wandering around alone in the world either. So why not populate the world with other players? Imagine you're playing GTA:VR and you get dragged before the Don, but this Don is a real person who built himself up EVE-style to king-like status in the game. When he gets up in your face, are you still gonna laugh at him, or are
you gonna be the one whose knees are about to buckle?
At his speech on designing good VR experiences, Palmer was saying simply hanging out with other avatars even if you're not really doing anything can be a very compelling experience in VR. Hanging out with other people and not really doing anything? I don't know about you, but that sounds like Home to me.
If managed correctly, I think MMOs and stuff like Home could become pretty huge, because shared experiences are more powerful, since it gives you external reinforcement. Being able to glance over at your buddy, Bob, even if he's a big, purple ogre, causes you to subconsciously accept what's going on around you.
Yup, must be real. Bob the Ogre is here, and he clearly sees it too. Imagine you're walking home late one night and you see a flying saucer flit by overhead. If you were alone, you'd be inclined to disbelieve it, but if Bob saw it too, then shit just got real.
Media Molecule told us they wanted to let us record our dreams. Magic Lab introduces us to Morpheus, god of dreams. Imagine a set of tools from MM that allowed you to create your own little pocket universe, all built with a pair of wands like Anton in those multi-touch demos.
Okay, I need a block of wood here
but it needs to be this big
*puts hands out to stretch block to size* Still not much of a workbench though
*pulls out light saber and carves the block down a bit* You know, I think we should put some mountains here
When you've created your domain, invite your friends over to explore and rule over it with you. Or explore domains others have created. Or jump over in to the Star Wars MMO.
I think stuff like this is what Sony may have been hinting at when they said that "social" was going to be an important aspect of VR.
Sony are in a great position to limit what is supported in VR, and provide a set of expectations for what things should and shouldn't be in VR. Oculus are trying hard to do that too, Palmer has openly talked about needing VR exclusive games, and not retrofitting, but Rift is on PC, and people are going to do it anyway, no matter what Oculus suggest.
/shrug Both sides have their merits. Yes, being inside a curated ecosystem means you get more consistent quality, but being out in the wilds means you get a little more variety and sometimes see some really interesting shit.
No worries about the misunderstanding.
Okay, here's where I'm coming from. Let's say that DK2 is good enough to evoke presence. Abrash says it should be based on his experiments with his prototype and people are reporting it with both DK2 and Morpheous. So Sony baselines VR on PS4 at 1MP and 75 Hz, which is a nice balance between blur reduction and performance demands. Oculus are looking to baseline at 90 Hz a 20% increase in performance demand and 2MP, doubling the already increased demands.
So let's say Ubi release some big VR game for PS4 and PC. They tune the game to look nice on PS4 and run great. Let's say on the PC side of things, the PS4 would be equivalent to Medium settings, with most effects off. But on the PC, thanks to the high specs for the Rift, you'd need a GPU 2.4x as powerful as the PS4 to run those same Medium settings. I took a look, and the Titan is almost exactly 2.4x as powerful as the PS4. It's also
more than $1000 on Amazon. Which do you think will impress potential customers more when explaining why they should buy a $1000 GPU, running the game on Medium and minimal effects, just like the $350 PS4 except at 2MP, or running the game on Ultra with all effects at 1MP?
And what about the poor schmuck who actually
has a PS4-level PC rather than a Titan? He'll need to turn his settings down to Asstastic, if he's even able to run the game at all. All just because they doubled the resolution of the display. That's what I mean by leaving customers out in the cold. Is it not better to have the same resolution with equal or better graphics and reaching a broader range of customers than have only a tiny percentage of your market be able to manage even equal graphics, much less better?
The Steam numbers indicate powerful GPUs aren't terribly common. You dismiss them and assure me the average PC is really quite powerful and getting faster all the time, but that's not much of a rebuttal.
You cite flawed methodology and imply that for some reason Valve may want to hide just how powerful the average PC is, but you present me with no alternate source of data beyond, "Trust me; there's lots."
If Valve do know how secretly powerful the average PC is, maybe we should look to them for recommendations for a minute. If you read Abrash's blog, while he does talk about 8MP at 1000 Hz and stuff like that, he also says none of that will happen in his lifetime, so it makes more sense to discuss what can reasonably be achieved over the next year or two, taking in to account not only the ability to source displays, but also the ability to drive them with reasonable performance and IQ. The reference design he presented was exactly that; a design potent enough to evoke presence in most users, yet with performance demands in reach of the average gaming rig, all for a device releasing in 2015. Would Abrash agree that 2MP panels are superior to 1MP? Without question,
but he doesn't think it's a reasonable design goal for a device releasing next year.
If Oculus set 1MP as their baseline performance, they can offer PS4-level VR to users with PS4-level boxen, and they can offer PS5-level VR to users with PS5-level boxen. And they can maybe even offer PS3-level VR to users with PS3-level boxen. If they go to 2MP, they can only offer PS4-level VR to guys with Titans, and the guys with 7850s will get a gimped experience if they get anything at all.
See what I'm saying? I'm not saying they're not allowed to be better than PS4. I'm saying that visuals will already be taking a noticeable hit to support VR, and at least today, bumping those settings back up a few notches may be more noticeable than bumping resolution, and it's a preferable solution since it's scalable, while a resolution bump effectively raises the performance floor substantially.
Why not release a 1MP display in 2015 as Abrash recommends, build up a sizable and vocal community of supporters, and then release a 2MP model in 2017 for the Titan++ guys who were already running Ultra on Rift 1? It just seems like they're jumping the gun here. I know Oculus have "the best minds in VR" now, but is one of them Tim Taylor or something? Why the rush to double the minimum system requirements? PC will be able to show its superiority at any given resolution, so why set the bar for performance so high?
Don't forget that PC/OR isn't locked at max spec, you can tune it to your liking within the ceiling you have available, that's the infamous strength of the platform.. You are not forced to render at 1440p/90Hz (or 4K/120Hz whatever comes in the near future). If you have less powerful hardware then turn down the details and run it @ 1080p/60Hz then upgrade later for proper presence.
I don't think upscaling is a particularly good idea for VR, especially if you're expecting everyone with less than a Titan to do it.
I disagree with this. This is different than other mediums. The experience is absolutely crucial. If there are other features in a competitor's headset, they can market it all they want as an advantage over the other. However, if that experience is flawed there is a serious risk of disorientation and sickness. On top of that, the sensation of presence is lost and the unique experience of VR is lost.
Why do you assume Sony's experience would be flawed? By all accounts it's quite nice, and it ticks all of Abrash's checkboxes for presence.
This isn't an either/or proposition. Sony are perfectly capable of delivering a solid experience visually, and then strengthening that experience with things that pull you still further in to the simulation, like realistic audio, solid motion-based interaction, and strong community ties.
