Oculus Rift Discussion

[isiolia]

The Wikipedia article has a general progression of the hardware. Suffice to say, there's a significant hardware improvement over each iteration. A DK1 has an effective resolution of 640x800 per eye, with a single actual panel. The CV1 is two separate 90hz 1080x1200 panels (same as the Vive has).

There are other aspects that were improved on over the course of it, such as size/weight, auto-screen off to preserve the LCDs, and so on.

Aside from image quality, a lot really comes down to what SDK revision something was compiled for. Far as I know, the DK2 remains supported by the current 1.3 SDK for development purposes, though it's not officially supported on the consumer/release side. In turn, most of the release stuff apparently works fine on it for the time being, but it doesn't have to. Oculus's wording is basically "will stay functional as development kits this year", so I would kind of expect support for it to drop in a future SDK release.

The flip side is that CV1 shipped with the 1.3 SDK, and does not support anything earlier. While, practically speaking, it just means that something has to get recompiled with the new version...as it stands right now, there are a fair number of demos and mods out there built for the 0.8 runtime (or earlier) that will only run on DKs.

While that is kind of annoying, I think it's mostly a transitional thing. Some of them are getting redone for CV1 support, and any current/future releases will target that too.


SteamVR stuff is different, and it does support the Rift - the hurdle there is more that a lot of it is built with room-level or at least motion controllers in mind, and Oculus hasn't shipped their hardware for that yet.

[noiseredux]

so I couldn't really get a DK2 and expect to be playing all the new Rift games? Or I could... but only count on it for a year?

[isiolia]

I don't think you could really count on it exactly. For the time being, the current drivers are going to recognize the hardware. However, since it's not supported on the retail/consumer side, a developer could potentially develop and test all on a CV1 and release something that doesn't actually work on the DK2...and Oculus would be fine with that.

More that, if you had a DK2 sitting there, it wouldn't be a paperweight...yet. It's only a matter of time though, so I wouldn't invest much in one. Might retain a slight niche for playing a small window of VR stuff I suppose, but I suspect it's only a matter of time before someone makes a compatibility layer or something.

[noiseredux]

it's tempting to try to grab one in the $200-$300 range... if I felt like I'd be able to use it for a while. But I'd be pissed if a game came out next year I couldn't play. :\

[noiseredux]

so I've made a couple lowball bids on a DK2, but no luck yet.

I'm still also curious about Razer's OSVR Hacker Dev Kit... I can't really tell how compatible it would be with Oculus (CV1) and Vive games... if it is at all. (?)

Like, if I knew I could play all the VR games on Steam (which is maybe... cuz isn't SteamVR compatible w/ OSVR?) then I'd be much more likely to go for it.

EDIT: Actually, if I'm reading correctly it would seem that the OSVR kit is functional with SteamVR stuff that doesn't require full room tracking. So I don't think I can explore Ikea. :\

Hmm. Apparently Elite Dangerous works with it also... That makes it pretty tempting.

[Krejlooc]

Re: OSVR

I actually have a HDK on order but haven't received one yet. For the general consumer, I'd say hold off on the purchase at the moment because there is supposed to be a major revision coming soon that will be a significant upgrade, but is supposed to keep the same price point. For someone dying to try VR at the moment, updates be damned, OSVR isn't the worst solution out there. A big leg up it has over the DK2 is that it does 360 degree positional tracking from it's outside-in camera tracking solution - the headset itself has a small box on the back of the strap that also includes IR LEDs to track. With the DK2, if you turn away from the camera, it loses tracking, and depending on the application this manifests in everything from your character snapping back to being upright, to the screen fading the black, to drifting off into infinity. Very annoying. One major perk of the DK2 over OSVR, however, is that the Oculus 1.4 SDK enables asynchronous timewarp across all oculus devices (save the DK1) which massively improves performance. This doesn't work with OSVR.

I actually much like the proposition of OSVR, I think the market needs a $300 headset. Picking up an HDK probably isn't a bad idea in and of itself, but support is kind of wonky at the moment. Officially, SteamVR supports OSVR through a plugin, but it reportedly performs poorly. Given the support the Sixense stuff has been getting in steamVR lately (the razer hydra and unreleased Sixense STEMs are completely compatible with the Vive controllers at the moment) I think it's just a matter of time before compatibility becomes good enough to wholeheartedly recommend, but everything I've read tells me it's not there yet. But that is the ultimate goal - complete interoperability among all headsets. Currently, at this very moment, the OSVR-SteamVR plugin is completely broken, so you can't run anything on the headset. This sort of happens from time to time as the plugin gets updated, and should be fixed in the next few days.

Where things get tricky is the Oculus SDK stuff. The Oculus SDK applications can currently run on the Vive by using ReVive, an injection driver that uses LibreVR, an opensourced variant on SteamVR (not to be confused with OpenVR, the steam-free version of SteamVR that is not open sourced). When the OSVR-SteamVR plugin is working, you can get Oculus SDK content working on the HDK using ReVive, but it isn't officially supported so, should an Oculus SDK update break ReVive, you might be waiting a while for it to become workable again. As I said earlier, Asynchronous timewarp only works on DK2 and CV1 at the moment - SteamVR itself has it's own reprojection method that works a bit differently and isn't nearly as smooth, and doesn't turn on with Rift apps at the moment. You can really see the lack of timewarp in games like Chronos and Lucky's Tale, which lean heavily on ATW to keep the framerate up to 90 fps. Even on my powerful dev PC, with ATW there are parts where those games become really choppy and drop down to what feels like 40 or 50 fps (and this is extremely noticeable when you're dropping from 90 fps) because ATW isn't enabled. The developers behind ReVive said it should be possible to get ATW going in their injection driver, but no time table has been set (it's still a very early project).

A bunch of chinese headsets are in the pipeline that is based on the OSVR spec, which is a good thing, and sort of ensures a market for this stuff will eventually form. My only hesitation in recommending the OSVR HDK is, again, that a better model is supposed to be announced in just a few months and the experience is supposed to be very "hackerish" at the moment. If messing with drivers and spending a lot of time getting your stuff to work doesn't scare you, nor the prospect that the driver could break randomly for a few days here and there, and you don't really care about updated specs, then I say go for it.

FWIW there is another headset coming out of China called the Pico that was announced a while back supposedly for the equivalent of $300. It uses the same screens as the HTC Vive and the Oculus Rift CV1, and is a "dumb" headset (i.e. no computer components in the headset). It uses optical inside-out tracking like OSVR and will include 2 PS Move-like wands for hand tracking. It's big selling point is that the controller it comes with packs a snapdragon CPU inside so it can run google cardboard apps "wirelessly," or you can tether it to a PC to presumably run OSVR apps. No release date has been set, and you know how chinese product announcements are (grains of salt) but it looks appealing.

I bought an HDK specifically to test this stuff out because I was curious. When it ultimately arrives, I hope I can give an update. I really like the low end market for VR. Obviously the high end stuff is super exciting, but I make most of my money in the low end, I've developed several apps for google cardboard because it's easy to distribute to people. If the tech is going to proliferate, it'll be on the low end of the price spectrum, which is why PSVR and Cardboard and GearVR and OSVR and all these low end solutions really matter to me. I just don't want to recommend such a headset to someone looking for a plug and play experience. I personally ordered my HDK as a low end option for me to test my applications on to see if they'll run on both high end headsets (I have a CV1 and a Vive) and low end headsets.

hope this helps, let me know if you have any other questions.

EDIT: About elite dangerous and SteamVR - the steamVR version of Elite Dangerous is really messed up at the moment, and has been for quite a while. No time table on when they'll fix it, but it runs at a noticably lower resolution than the rift version. This is likely because of a subtle difference in the way Oculus SDK and SteamVR's compositor works - the rift technically has a slightly greater diagonal view, and it appears they are merely scaling the image down for the Vive. This normally isn't a problem, but for Elite Dangerous, which uses a tiny font, it makes much of the menu unreadable.

I wouldn't recommend basing any non-oculus purchase on elite dangerous. All this sounds like I'm really pro-oculus, but truthfully I hate a lot of the practices they are doing in early VR with regards to market segregation and hope all these issues become ironed out.

[noiseredux]

man, ask and ye shall receive! Thanks for your feedback; I knew you'd be the guy to ask!

Personally I love the concept of a more unified marketplace. That is, pick whichever headset you want and you can play VR GAMES, rather than pick a Rift to play Rift games and a Vive to play Vive games and so on. So hopefully that is the way things turn out. Because I agree, market fragmentation is not a good idea if we want this tech to actually stick around. And yeah, we def need these lower end headsets to help make that happen as well. The point is, if I were to get a $300 set and actually use it a lot - it would make the idea of spending more on a better set a lot easier to stomach.

Question: Would you consider the OSVR HDK a better route than trying to get hands on an OR DK2? I'd think the OSVR would have more longevity?

I know you said a new OSVR set is planned soon - and for that same price point - how much do you know? I mean... are we talking about an announcement in the next couple months but then it could be another 6 before we see it for sale? If that's the case, I may well just be impulsive and if the next set seemed like something worth upgrading to I'd imagine the HDK would at least retain enough value to recoup some of the investment.

I'm feeling slightly impulsive, but trying to be smart haha. To a degree it's not even like there's any one game I'm dying to play (Elite Dangerous did look awesome though...) as much as I'm just really interested in the tech. I might find myself trying a virtual movie theater or desktop just as likely as am I to explore the VR games.

Do you have an idea of when your HDK is expected to arrive? I'd certainly love to hear your first hand experiences.

[Krejlooc]

man, ask and ye shall receive! Thanks for your feedback; I knew you'd be the guy to ask!

Personally I love the concept of a more unified marketplace. That is, pick whichever headset you want and you can play VR GAMES, rather than pick a Rift to play Rift games and a Vive to play Vive games and so on. So hopefully that is the way things turn out. Because I agree, market fragmentation is not a good idea if we want this tech to actually stick around. And yeah, we def need these lower end headsets to help make that happen as well. The point is, if I were to get a $300 set and actually use it a lot - it would make the idea of spending more on a better set a lot easier to stomach.

Question: Would you consider the OSVR HDK a better route than trying to get hands on an OR DK2? I'd think the OSVR would have more longevity?

I know you said a new OSVR set is planned soon - and for that same price point - how much do you know? I mean... are we talking about an announcement in the next couple months but then it could be another 6 before we see it for sale? If that's the case, I may well just be impulsive and if the next set seemed like something worth upgrading to I'd imagine the HDK would at least retain enough value to recoup some of the investment.

I'm feeling slightly impulsive, but trying to be smart haha. To a degree it's not even like there's any one game I'm dying to play (Elite Dangerous did look awesome though...) as much as I'm just really interested in the tech. I might find myself trying a virtual movie theater or desktop just as likely as am I to explore the VR games.

Do you have an idea of when your HDK is expected to arrive? I'd certainly love to hear your first hand experiences.

I don't know too much about the supposed revision - I run a dev group out of houston that meets every few months, and we are planning a conference for this fall, and razer is one of the groups coming out. In a causal, off-the-record conversation they mentioned the revision but not much else. razer, if it's not obvious, is one of the companies involved in the OSVR consortium.

Looking up completed listings for DK2 on ebay, it seems they're going for about $550-600 at the moment? If thats the case, then the DK2 is absolutely not worth that price, I would never pay that much for a DK2 when for just a couple hundred more I could get a Vive brand new. Vives actually ship pretty quickly. If your options are a $550 DK2 and a $300 OSVR, I'd go with the OSVR every time.

that said, I cut my teeth in VR with the razer hydras, and after about a month with the vive, I can say that motion controls in VR are as big of a game changer as the headset itself. It sounds funny to say, as when I say motion controllers people inevitably think of the wiimote and how inaccurate it was when you were tilting it left and right or shaking it to trigger a canned animation. motion controls in VR are nothing like that. What the wiimote was, is closer to gesture recognition - outside of when the camera on the wiimote was facing the sensor bar (inside-out positional tracking) it had no idea where it was in the room and had to rely on IMUs to read changes in acceleration. This is hugely inaccurate. Motion controls in VR are 1:1, the solutions for the PS Move and Oculus Touch (outside-in positional tracking) and Vive (inside-out positional tracking using lighthouse beacons to triangulate position) ensure that, essentially, the controllers will always know precisely where they are in the room and their exact orientation. This is huge. I can't express enough how much this changes the interaction in VR. Playing VR games with a gamepad feels like you're playing just normal games in VR, but using motion controls feels like an entirely new experience. It is an enormous leap.

What makes this so tricky is that if you go the OSVR route, what motion controllers you'll have available is sort of limited. Admittedly, you'd run into the same problem if you went with the DK2 - the Oculus Touch controllers communicate directly with the CV1 headset, so you can't use the touch controllers with any other headset. And the Vive controllers communicate with a breakout box that is only sold with the Vive itself (and controllers aren't even available separately yet, if ever). So you'll have to get creative with your motion controller solutions. SteamVR and OSVR actually have a motion controller abstraction layer in their APIs which is supposed to make cross-compatible motion controllers a thing in the future, but the thin market at the moment means the reality is pretty slim. Leap Motion is a kinect-like adapter you can connect to the front of your headset to track your hands, and it has a plug in that'll transform your hand tracking into a that motion controller abstraction, so you can treat your hands like motion controllers themselves, but it's laggy and you have to have your hands in front of you. One of the biggest joys in VR is, for example, firing your gun to your left while looking to the right to dodge some bullets (something that lots of Vive games take advantage of right now). Not possible with Leap motion at the moment.

There is also supposed to be a PS Move plugin that I admittedly am completely clueless about, so I can't tell you if it works at all, let alone how well it works. But if it does work, it is assuredly the cheapest, best motion controller solution for someone a budget.

The best supported motion controller solution for OSVR at the moment is the razer hydra, but those go for stupid amounts of money at the moment, and the experience is rather terrible because they are pretty imprecise, subject to interference, drift, and worst of all they are wired. The Sixense STEMs are eternally delayed and, while reports from conventions say they're much improved over the hydra and are pretty cool and comparable to the Vive controllers, they're also like $500.

Keep in mind, these issues would be just as prevelent with a DK2. People often say "I'll just stick to seated cockpit games, or play VR movie theaters" but trust me, you will want motion controllers. Even if you don't want to stand up, I guarantee you will still want them. they are enormous game changers. The lack of motion controls are the single biggest problem with CV1 at the moment IMO. Several games try to get around the lack of motion controls by treating the headset itself like a motion controller, and that feels weird.

As for my OSVR HDK, I'm not sure when it is scheduled to arrive, they ship kind of in weird increments and waves. I actually only ordered it a few days ago, so it's not like I've been waiting a long time.

[isiolia]

I've dabbled some with the Leap Motion velcroed to the front of my Rift, which is pretty neat, but a bit inconsistent. I think Razer is going to push on that some, since they seem to be making a replacement front that holds one.

I agree, motion tracking and control is going to be a huge part of a lot of VR experiences. Possibly something that might get limited to "bigger" setups too though, you think? Seems like you want two cameras and then some for proper tracking.

Generally though, I think VR isn't yet at the point where trying to cobble stuff together on the cheap is necessarily a good idea. The potential experience possible today is astonishing, but it's also dependent on having a lot of pieces all working together in concert at a fairly high minimum performance level (or better). That's what a Rift or Vive get you, from the companies that have been working towards it for years.
Other stuff is still not inexpensive, and can easily add up if you starting wanting to add to it.

[noiseredux]

it kind of sucks, because I'm interested in exploring the tech but there's a ($) wall. I mean, I could afford to buy a Rift or Vive right now if I wanted. But I just can't seem to justify that price. On the flip, there's lower cost alternatives (OSVR being the prime one it seems), but there's at least some indication that I won't really be getting the same experience. So I don't know... I go back and forth on the whole thing. But I think it just annoys me because it's kind of something going on in the gaming industry that is being talked about everywhere (especially if you're into PC gaming) and I can't even have a first-hand opinion. :\

#givethebabyhisbottle, I know. And I also know that prices will fall as tech gets improved upon. But as a nerd I can't help but be curious what I'm missing. I knew I shouldn't have re-watched The Lawnmower man...