Comment by drzaiusapelord | Hacker News Reader

Comment by drzaiusapelord | Hacker News Reader

Parent Full thread

I would like to know what kind of computer and graphics card I need to make this thing work. Yeah, a cheaper VR headset pricepoint is nice, but not if you're just unloading all the expensive components onto the buyer via a $1200+ PC purchase. Not to mention, as someone who suffers with motion sickness with traditional games, how will the Oculus work for me? Have they solved this riddle? Will they ever?

This is also why I see Sony having great success with its Morpheus headset. The PS4 is cheap as-is ($399) and if you already own one, its a no-brainer to go with their product. I'd be really surprised if Oculus ever goes mainstream. It still looks like an uber-nerds hobbyist toy like a, say, 3D printer, which technophiles have been telling us for years would be mainstream by now and in "every kitchen."

Carmack isn't stupid. He didn't sell to Zuck because he thought he'd be moving 10's of millions of product every year. He sold to Zuck because he knew that the shipping date doesn't really change things for Oculus. Its still a very rough road to mainstream acceptance and economic competitiveness. I can afford $299 for the headset, but I'm not interested in buying a super gaming box to make it work. Nor am I interested in being a beta tester for a technology known to be a bit wonky, cause motion sickness, have questionable 3rd party support, and Facebook ownership/integration/driver and API's TOS.

This is what on top of what Valve is doing, which may just steal Carmack's thunder as they have everything they need to make this work: the store, no relationship with unrelated businesses like social networking, large customer base, community goodwill, game publisher relationships, a decade plus in the game selling world, etc. If both the Valve set and the Oculus set go for $299, I can't think of any reason to even consider the Oculus.

You'll need a "gaming rig". That's the simplest way I can put it. Don't buy an "ultrabook" or an "all in one" or some $500 laptop to use Oculus. Whether it's a PC (cheaper) or a notebook, it needs to be gaming-focused one.

In terms of specs I'd buy a GPU with at least a 3TF performance, and even that it's probably going to be "mediocre" for the next 2 years. 2TF should work, too, but might not last you too long, especially for new games. Obviously, the higher the performance the better, so if you can afford to pay thousands of dollars for a gaming rig knock yourself out. It will last longer.

You might also want to keep some accessories in mind, such as the Virtualix Omni or those full body controllers, which could cost you an extra thousand dollars.

Everyone I know has a gaming rig of some type, but not anything that can do 1080p per eye at 60fps per eye, which is where this stuff actually starts to work correctly and what Oculus is designed to use (less motion sickness, better 3D illusion, etc).

I don't think a lot of people interested in this technology realize how much hardware they will need to buy to make this work. You're looking, at minimum, a $400-500 graphic card. Assuming you have the CPU heft to handle everything else. If not then there's another $400-500 for the chasis, mb, high-end cpu, high-watt ps, fast 8-16gb of RAM, etc. And that's for a fairly mid to low-end VR box. That could be $1000 out of pocket right there on top of whatever the Oculus costs.

In theory, with DX12 timed for a 2015 release it could ease the burden enough to make the PC upgrade easier. From what I've gathered DX12 will be significant in terms of efficiency, so new games which support it may be viable.

It's also worth considering whether new games that come out will take VR into account in terms of graphical complexity. If it's clear VR is a viable marketing strat for a game or it gets to a point where big games are expected to support it, it could mean a developer will compromise scene complexity for it. I'm okay with that.

Well, it depends entirely on your fidelity of course.

For example, just last night I played a game of Supreme Commander. That was one of the first games I know of that supported dual monitors, and it does it well. But I digress- the point is that I had it maxed out on two 1080p monitors and it was running at 60fps the whole time.

So yes, you will need a big beefy expensive graphics card to play the newest AAA games maxed out at 60fps. But you could always just turn the settings down, or play older games. I'm also willing to bet that when the VR market takes off later this year/early next year, that the indie scene will heartily embrace it. And typically most of those games run fine on "normal" hardware.

>For example, just last night I played a game of Supreme Commander.

I'm not buying a VR headset in 2016 to play a game from 2007.

If you're not happy with first generation then don't buy it. I guarantee that the second and third generation will be even better. Some guy whining on HN about a first generation product adds zero value. That first generation iPhone had lots of critics too, and rightly so.

Bottom line is that a real VR product ships in less than 12 months.

The Vive ships earlier, so Oculus will actually be second fiddle.

That's fine. One thing that I am certain of: today's $500 video card is next year's $250 video card and the next year's default upgrade from on-motherboard.

I had to turn down numerous sliders, including the Oculus Quality slider which made screen door pronounced, but I was able to play Elite at 70fps (you want 70 on the DK2) on a Retina iMac bootcamped into Windows. It tore or jerked occasionally, like in Lave, but it's not nearly as demanding as you say. Minecrift did even better.

My experience was very smooth and playable for hours. On a Mac. With laptop graphics. I'd definitely build a machine around a 980 next, but the tech is perfectly accessible.

If you're getting shitty graphics, turning everything off or lowering to low levels, low framerates, and screen door effect and jerking, etc uh I would rethink what it means to have an acceptable VR experience.

If anything your sub-par experience is proof that I'm correct. Without excellent hardware, the experience is terrible. Just because "it kinda worked" doesn't mean its going to sell or impress anyone.

You were arguing a barrier to entry, not quality. My point was there's not a wall in front of this stuff as you say. You moved the goalposts to "acceptable VR experience," which is a different topic and largely subjective.

I also never said the experience was terrible or "kinda worked." Quite the opposite, in fact, I rather enjoyed it even at a lower quality because I understand the limitations of things. Outside of Lave I stayed right at 70fps and could turn my head from left to right without even a millisecond of delay. It was actually the most fun I've ever had in PC gaming. But you're right, because I didn't have everything turned up to maximum, I should rethink my enjoyment.

Yep! I bought all the stuff you said, but skimped slightly on the CPU/motherboard (2 years old) and I'm having a terrible experience. A universe of judder :-( I'm ready to sell the whole thing.

I'd be interested to know how simple you can make the graphics and still have a good experience. Isn't there an inherent problem with VR that if you try to compete with state of the art non-VR graphics, you're always going to need a more powerful machine for VR than you need for non-VR (e.g. larger field of view, two screens, higher FPS)? The problem will only increase as the headset Field of View increases.

Very simple graphics with flawless 75fps+ and very little lag is highly preferable to dropping frames in a complex scene. In fact Minecraft is a case in point - it works great in VR!

I agree overall but at the same time, it really depends on what you already have. Sure, if you own a $400 PS4 and whatever other items you typically buy with a game console, another $300-400 for a HMD is an easier sell.

But I don't have a PS4. What I do have is a solid, modern computer since I need one for all sorts of work and play situations. I can either spend $400-500 on a PS4 and any typical addons or I can spend $400-500 on a powerful graphics card for my computer. And with graphics cards, performance goes up and price comes down every year.

The other angle I'd need to consider is what you can do with the whole package (computer + Rift or PS4 + Morpheus). Morpheus will work on the PS4 and I'll be able to buy PS4 content to use with the Morpheus. The Rift will work on any capable computer and I can either buy games/content from Steam or from other sellers or I can go to Oculus Share and download all sorts of experiments and demos and games from all sorts of developers. I can even fire up the free version of UE and start building environments to walk around in via the Rift immediately. The power to create and share content shouldn't be underrated, especially at this early stage.

I think that in 5-10 years time, if this hasn't gone the way of 3D TV (something cool and easy to find but ultimately not that big a deal), the costs will be low enough and content will be abundant enough that it won't matter whether you use a black box under the TV or a silver box under your desk. But during this early time in the market, access to demos and creative tools will be a big deal.

When someone develops a VR teleconferencing app with depth cameras and live streaming, the betas and first release won't be on some Sony or Microsoft game console. They'll be on Windows, OSX, and Linux. When you want to use a sweet HOTAS for your VR flight simulator or enable VR support in the config of some new game, you won't be able to do that on a Playstation or Xbox.

All that stuff is more than enough to get me to lean toward a $500 GPU upgrade over a $400 console purchase. I think until VR stuff isn't early adopter/enthusiast territory by definition, the more packaged solutions will be lacking. Once it's commonplace, it'll be more a question of ergonomics like PC/console gaming is now.

Carmack is an employee, primarily writing code. The company is owned and run by other people.

He had a huge amount of buy-in/ownership when he was brought on board and certainly had some level of executive decision making. I doubt he just gets a 1099 for $market_wage and goes home. He's swimming in stock and other perks. You don't hire him to sit locked in an office all day and not have say in your organization.

You hire him because he's been writing 3d graphics code for longer than literally anyone else on the planet.

His is the CTO. That is not a coder monkey position, yes he probably codes and supervises a lot but he's C-level and that means executive decision making.

John Carmack, CTO of Oculus, probably has a few responsibilities that could be viewed as "running the company", at least in part.

He stated publicly that he didn't have much if anything to do with the facebook acquisition.

http://vrfocus.com/archives/14917/nvidia-ceo-vr-going-future...

Reply on news.ycombinator.com