DK2 FOV downgrade from DK1?

[jrd3d]

Sorry if there is already a topic for this that I couldn't find from doing a search, but I was wondering if the downgrade in the Field of View from the DK1 to the DK2 should be of any concern?
The DK1 states a Field of View of 110 degrees and the DK2 states a Field of View of 100 degrees.
I know it's only 10 degrees less, but does that mean that the edges of the screen will be more noticeable?

[pewzie]

I just dont understand why some people on here have bought any of the Oculus units.

You either want to be part of the beginning of an epic journey into what will be one of the most revolutionary and used forms of technology in the future or you dont.

It staggers me that some people dont seem to grasp what we are at the beginning of here. This will be a long and monumental journey and I am so glad I get to experience it at the beginning.

I was there at the true beginnings of personal computing with the likes of the ZX Spectrum, Commodore 64 etc and I would not swap my time with them for a second regardless of how my incredible spec PC performs now.

Its an experience. Enjoy it for what it is, which is a solid start that luckily we can all pretty much be a part of commercially.

Look forward to the trolling 🙂

[amenconi]

"brantlew" wrote:
I believe the specs for DK1 were listed at around 110 and the DK2 at 100. That's about right. If you view at comfortable eye relief (approx 10mm) then you probably will not see a difference between the headsets. If you like to have your eyelashes brushing the lenses then you will notice the screen edges on the DK2 more than the DK1 (appearing as black bars on the sides). The DK1 was overspec'd in this regard and you pretty much couldn't see the edge even with your eyeballs touching the lens.

The seldom discussed fact about FOV is this. INCREASING FOV COMES AT HUGE COSTS TO RENDER SPEED. Why - because the pre-warp render texture size increases exponentially as FOV increases and also additional geometry must be processed. In fact, render speed is much more affected by FOV than by final resolution of the output device. Go into your Config Util while rendering the Demo Scene on DK1. Now play with the eye-relief slider while the demo is running (this is easier now with Direct Mode). You can see the rendering radius increase or decrease along with the slider. That's a function both of the lens distortion and the oversized screen of the DK1. The same thing on the DK2 will produce less variability. Render target size for DK2 is more-or-less constant, but is highly variable on DK1. So here's the interesting thing - the maximum render target size on DK1 is GREATER than on DK2 which leads to a rather counter-intuitive fact. At the same frame-rate it is actually MORE expensive to render out to DK1 than DK2. The higher render requirements for DK2 are almost entirely a result of the higher frame rate rather than the higher resolution.

Engineering for VR is a constant set of trade-offs. It's easy to say that it would be better if FOV was 120 or 140 degrees. Of course it would feel better if you can achieve it. But those values would pretty much melt all your computers because we are talking about render target resolutions of probably 6k x 4k or larger!



Edit: A clarification here since Reddit seems to be jumping to conclusions... Render target sizes are not fixed. The SDK provides "suggested" render target sizes to achieve lossless pixel sampling in the center, but developers are free to choose whatever render target size they want. Render targets can be reduced to trade performance for image fidelity. But in general - increasing FOV increases rendering costs and is yet another tradeoff associated with high FOV along with aesthetics, mechanical design, optical design and distortion, optical density, and part availability. The DK2 is what it is mainly because of physical constraints and visual tradeoffs. A positive consequence of that is that rendering costs are also kept in check.

Brant, thanks for giving us an official response on this.

As someone that feels that limited FOV is one of the greatest barriers to immersion I am a bit disappointed to hear it was lowered a bit. However I understand the tradeoff made and even more important than immersion is to remove sim sickness as much as possible.

I'm fairly sure I'll be happy with a small FOV reduction considering all the other improvements made and will reserve judgement either way until I have my DK2 in hand to test with.

With DK1 I used A cups with the eye relief about half way out to accommodate my glasses, thinking now maybe I can get contacts and bring the screen closer to regain some of that lost FOV.

Can you comment on the planned FOV for CV1? I think the greatest concern now considering your comment is that there will be another FOV compromise due to another jump in resolution. Honestly I'd be surprised if this is what you guys ended up doing but then I would have also assumed the DK2's FOV would have at least stayed the same from DK1.

Though this is a negative topic, overall I'd like to say I think you guys are doing great things and appreciate access to the kits you're building. Whatever the FOV, thanks for all the work everyone at Oculus is pouring into this technology.

[n00854180t]

I'd still like to know if the CV1 is going to be even further crippled in FOV than the DK2.

Trying to claim this is "entitlement" is nonsense trolling bs. These issues will affect development. They need to be hashed out now, not one month before CV1 drops, as is the case here with DK2.

[davidjc]

I have been nothing but a huge promoter of Oculus, but I am sorry I don't believe they should be the industry leader or setting the standard when they are treading backwards on such an integral and important feature. I don't buy it was done to save our GPU's from rendering. It does not fly with me.

The fact is FOV was something that was considered deficient in DK1. I am not sure where Oculus's priorities are, but they have missed it big time with the DK2. FOV should have under no circumstances gotten worse. And frankly if the CV1 is not improved a great deal I won't be buying it. Hopefully there are other companies out there to get some good hardware out there as well.

Thew problem is, that Oculus is setting the industry standard with game makers in particular. It's already hard to get them to implement any kind of VR at all, and now the only VR we will have in a lot of titles is VR with crippled FOV. This makes us fully dependent on Oculus, and makes it very hard for competitors to get in the market, they will be confined to specs and SDK that Oculus has been promoting.

It would have been better if another company had played this role, one more comitted to keeping core immersion factors as a priority. I am sorry I do not see why we had to lose 10 degrees of FOV.

[menione]

I find it very unhealthy for the community to call people simply giving their opinion as “Entitlement”. Most people that have commented on the low FOV seem to be only giving feedback as to what they believe will be a better product.

Is it possible that Oculus compromised on the FOV to keep costs down? Sure. But I thought one of the reasons why a Development Kit exists is so that they can get feedback. People are simply trying to express that FOV are one of the compromises they would prefer not to make.

And while Oculus is still developing CV1, this is probably the best time for opinions such as these to be heard.

[RirtualVeality]

"oxmyx" wrote:
The entitlement in this thread is astonishing.

Guys, do you honestly believe Oculus chose to lower the field of view, when they just as easily could have provided a larger field of view? No? Well, then maybe, it isn't so simple after all.
VR is, at this point in time, all about making compromises. And you've got to make an awful lot of compromises for a $300 product, which will only be manufactured in small quantities (compared to, say, smartphones; and even compared to smartphones, the DK2 is extremely cheap).

DK2 was all about positional tracking and low persistence, and they've delivered on these things. But why oh why didn't they also increase the field of view? Well, maybe it was simply not as high a priority compared to the other things. It certainly doesn't prevent developers from creating games and content for VR and becoming acquainted with VR.

Soooo it's fair to say they CHOSE to lower the FOV to compromise for performance? Most people are aware of why they did it. That doesn't mean they have to like it and it has nothing to do with any "entitlement" as a few people have been saying. The consumer (it's not a consumer product DERP DERP!) helps to shape the final product, so if they're not happy about something they have every right to let them know. Relax.

[VRCitizen]

"jngdwe" wrote:
Out of curiosity, who here could see the edges of the screen some times in the DK1? I certainly could, at times. I use the C cup lenses with the rift as close to my eyes as it will allow. It is STILL like wearing some goggles, so I doubt this is a huge issue in the DK2.

I have DK1 and aside from the standard binocular effect, I have always been able to see the edges of the actual display as well, appearing as a kind of vertical letterbox effect. I wear A cups with my eyelashes brushing the lenses, and when staring straight ahead the black bars are quite prominent in my periphery and a bit hard to ignore sometimes. When I try to look at them however, they are greatly reduced or disappear entirely (haven't used my rift in a while).
EDIT: I forgot that when one of the more recent SDKs came out and they changed the warping shader, or whatever, it reduced the black bars presence quite a bit.

The idea that DK2 might have larger black areas to contend with is kind of a bummer, but we'll see how it goes. I'm not going to scream heresy, gnash my teeth and declare DK2 to be The Antichrist, but I certainly hope CV1 will eliminate the bars, leaving me to battle the annoyance of the binocular effect only.

What's funny and slightly frustrating to me is that some people that I've demoed the Rift to say they didn't notice any blackness at all, not even the binocular effect. How that's physically possible I have no idea.

[Anonymous]

I think it's fair to say that the change in FOV is due in part to the smaller screen, which was obviously chosen to accommodate a smaller form factor. No one here could deny that the DK1 was pretty big. I wouldn't sell something so ugly and fat as a consumer product.

[fenrok]

"davidjc" wrote:
I have been nothing but a huge promoter of Oculus, but I am sorry I don't believe they should be the industry leader or setting the standard when they are treading backwards on such an integral and important feature. I don't buy it was done to save our GPU's from rendering. It does not fly with me.

The fact is FOV was something that was considered deficient in DK1. I am not sure where Oculus's priorities are, but they have missed it big time with the DK2. FOV should have under no circumstances gotten worse. And frankly if the CV1 is not improved a great deal I won't be buying it. Hopefully there are other companies out there to get some good hardware out there as well.

Thew problem is, that Oculus is setting the industry standard with game makers in particular. It's already hard to get them to implement any kind of VR at all, and now the only VR we will have in a lot of titles is VR with crippled FOV. This makes us fully dependent on Oculus, and makes it very hard for competitors to get in the market, they will be confined to specs and SDK that Oculus has been promoting.

It would have been better if another company had played this role, one more comitted to keeping core immersion factors as a priority. I am sorry I do not see why we had to lose 10 degrees of FOV.

As sad as it is but he is right. FOV should never be trade off cause DAMN FOV IS THE MOST IMPORTANT FOR IMMERSION CAUSE IT'S DAMN VIRTUAL REALITY WE TALK ABOUT. I'm sorry to say but i'm really disappointed. What wil happen with cv1? wil it be with resolution 3400 x 1800 and FOV 30 cause rendering? So why not just drop virtual reality and start making high definition flat displays? I'm really disappointed at technical choice Oculus made in dk2.

[LucidReality]

"menione" wrote:
I find it very unhealthy for the community to call people simply giving their opinion as “Entitlement”. Most people that have commented on the low FOV seem to be only giving feedback as to what they believe will be a better product.

Is it possible that Oculus compromised on the FOV to keep costs down? Sure. But I thought one of the reasons why a Development Kit exists is so that they can get feedback. People are simply trying to express that FOV are one of the compromises they would prefer not to make.

And while Oculus is still developing CV1, this is probably the best time for opinions such as these to be heard.

This. Exactly. If we can't voice our displeasure of a change in design, then what's the point of giving feedback? There are many legitimate developers who are disappointed with their choice of FOV, and by coming here and expressing this, there's the possibility that Oculus will lean more towards a higher FOV after seeing the amount of people who aren't happy with this decision.

Of course, I'm not trying to defend those people who are here simply whining without even trying it out