Rift CV1 stats confirmed: 2160x1200@90Hz. Oculus recommended card: "GTX 970". Frontier, it's time to optimize!

Do we know (pretty sure we do not but..........) if the screens in CV1 are full RGB or pentile?

if they are full RGB then actually the real world resolution boost (on anything other than green screens) will actually be a lot greater than what the simple numbers would suggest.

Whilst I would have loved 1440p 90hz screen(s) I must admit I am really not concerned. I am confident that it will be way better than DK2.

indeed, I suspect 1200P RGB screens will actually be better in the real world than a 1440P pentile display.

perfect example here

http://images.anandtech.com/doci/4165/Bayer-vs-PenTile2.jpg
 
Last edited:
Interesting note - extremely high frame rate gets you a sort of temporal anti-aliasing that substantially increases perceived resolution.

There is a resolution/persistence threshold for this to happen with micro-movement of the head as opposed to large rotations, CV1 hits it.

This makes a huge difference with content that is rendered at resolutions higher than the headset (movies in VR cinema, text layers, etc)

You can sort of simulate this by moving your head around while peering through the mesh of a microwave door, oddly. You're right that it's dependant on persistence. Actually, persistence becomes a real mixed blessing.. You don't want too much of it, or you get "smearing", and you don't want too little, or you break the visual system's saccadic masking, and cause perceptual strobing. There's a very definite sweet spot in persistence, in relation to lower framerates. Obviously, as you increase the framerate, a display which runs at a higher per-frame persistence proportion would produce less smearing, though (as there's less movement per pixel across the retina per frame).

It's all a bit like the headaches that we have when thinking about shuttering in cameras, especially when shooting fast-moving subjects like sports.

Meanwhile, I am really looking forward to diminished SDE in whichever headset replaces my slightly ramshackle DK2 :)




Whilst Id hoped for a better res I still think some FPS wizardry will make the resolution look higher then it actually is. I still think we'll be suitably impressed

Yep, throwing framerate at things is often the smarter thing to do- a point that has been sadly missed in the spatial resolution wars in video over the last few years. We're finally seeing standards work happening with high frame rate and high dynamic range stuff- but not until after people were made very nauseous indeed watching things like Super Hi-Vision showing moving shots :)
 
Last edited:
I currently run the DK2 with its unfinished drivers at a higher resolution than this with no problem (SLI GTX 970), so I'm not worried. I've got 2 USB3.0 ports on the top of my case so that should get me covered.

Still, I'm a bit surprised by the resolution, I thought it would be 1440p. I guess with better pixels and less screen door effect, things should be more readable even if the resolution isn't that much higher than DK2.
 
I currently run the DK2 with its unfinished drivers at a higher resolution than this with no problem (SLI GTX 970), so I'm not worried. I've got 2 USB3.0 ports on the top of my case so that should get me covered.

Still, I'm a bit surprised by the resolution, I thought it would be 1440p. I guess with better pixels and less screen door effect, things should be more readable even if the resolution isn't that much higher than DK2.
If it's RGB as well it'll look a LOT better than the DK2. Just moving to RGB basically adds 30% more subpixels, combined with 25% higher res it will look very good. Pretty excited about high fill as well, no VSync will help a lot if you drop under 90 FPS.
 
Most of the stutter, when I get any, is in station, but it's not particularly bad; or when there are shed-loads of NPCs (I.e. like when there was the multi-spawning issue when 1.2 dropped), which is most likely CPU related.

I run at ultra with GeForce Experience-controlled super sampling in Extended mode, my specs are in my sig, as you can see, my CPU isn't exactly new. Yes I'm running Two 780tis in Sli but I'm not convinced Rift rendering and ED scale terribly well anyway.
 
you're full of baloney

come over and have a try and judge it for yourself.

The only stutter I got is when I plotted a course on the galaxy map near Sag-A and activated hyperspace.
Apart from that it runs smoothly. Shadows run smoothly in the cockpit.
No complaints from me.
 
come over and have a try and judge it for yourself.

The only stutter I got is when I plotted a course on the galaxy map near Sag-A and activated hyperspace.
Apart from that it runs smoothly. Shadows run smoothly in the cockpit.
No complaints from me.
As we have said.. this is because you're in unpopulated systems, and the lag/studder is directly related to the number of ships in the instance, whether they are on screen or not.
 
Double Titan x o/c and judder still occurs when loading all the ships and planets. Res sites can judder etc even with low settings
 
Have people with dual cards tried disabling the 2nd and running on just 1 card on the lowest possible graphics settings? The judder could be from multi gpu frame pacing.
 
come over and have a try and judge it for yourself.

The only stutter I got is when I plotted a course on the galaxy map near Sag-A and activated hyperspace.
Apart from that it runs smoothly. Shadows run smoothly in the cockpit.
No complaints from me.

Go to a resource extraction site.
 
oh yeah.. and the resolution specs are painfully disappointing. It is only going to look marginally better than the DK2 at 1080×1200 per eye instead of 960×1080. It will basically look the same. We can only hope they did a lot of work to the lenses to minimize the SDE.
And this generates a question for myself: Extra resolution WILL have an impact on performance. Regardless of "how much perceptual difference" it will make to my eyes, my R9 290 is still going to have some extra work to do with the CV1. So,,,,,,,,,,

I am left wondering if I should just grab a DK2 when I can afford it and just forget about the CV1 until I can also afford a beefier GPU?

And if the CV1 is released at a price point that is about the same of the DK2, I would think it would force a lower price on used DK2's.

Or am I missing something?
 
spot on Riktar, before i upgrade my 2x680s cards i wanted to know when the CV1 was coming out. Sure i could get the new 980Ti's but whats the point.
We know that the SC1 is coming out Q1 2016

This also happens to be the same time that the new 16nm GPUs come out (currently on 28nm), personally i feel the two will marry up nicely. For CV1 i am planning on buying 2 x Nvidia Pascal GPUs (top end), that should net a system that can definitely run 2160x1200 at 90fps. And lets be honest, you are going to want to run super sampling on that head set. By that time im also hoping DX12, SLI VR drivers and screen door effect are all getting better.

For now, DK2 is freaking awesome, its pretty cheap and really, its a long wait to CV1... Get it and have a ball, learn about the ins and outs of VR and seriously prepare to be stunned...
 
As I was saying:
-High-tech stations (the ones with foliage)
-Supercruise while loading planets
-any RES

Stutter like crazy no matter the settings. Many people even with 980 have that problem, it's clearly an optimization issue on Frontier's part.


Completely agree.

You can have two titan X's and you will get that.

Lets hope they patch for SDK 0.6 get direct mode and positional timewarp running.
 
It's probably networking code, also the guys at Oculus VR forums wrote that Frontier adds an intentional delay into the game loop depending on the number of objects, somehow to reduce the network traffic because of that (sorry) horrible P2P networking they use instead of normal (and way more reliable) server based networking.

You can have a quad Titan Z setup on a multi CPU mainboard, it will still stutter.
 
As we have said.. this is because you're in unpopulated systems, and the lag/studder is directly related to the number of ships in the instance, whether they are on screen or not.

yes, you're right.
Back from my trip, and indeed it's getting worse now.

I guess I forgot how it was after being away for 2 months :rolleyes:
 
Same train of thoughts here, Riktar.

I bought a DK2 from ebay last week (sadly, it will only arrive on tuesday or wednesday) because I realized there is no way I can afford a CV1 *AND* the hardware needed for it. Interestingly the price for used DK2s seems to have increased slightly since the CV1 specs have been released (I've been monitoring for some time, wondering whether to bite the bullet or wait for the consumer version). My hardware is probably a bit low end for E:D and the DK2, but upgrading it to an acceptable level in the next months should still be less costly than a full CV1 upgrade.
That way I'll still have the DK2 to keep me happy, until such times where I either make enough money that I can just throw it around or the hardware becomes much more affordable.
 
He also said -


Interesting note - extremely high frame rate gets you a sort of temporal anti-aliasing that substantially increases perceived resolution.

There is a resolution/persistence threshold for this to happen with micro-movement of the head as opposed to large rotations, CV1 hits it.

This makes a huge difference with content that is rendered at resolutions higher than the headset (movies in VR cinema, text layers, etc)
You can sort of simulate this by moving your head around while peering through the mesh of a microwave door, oddly. You're right that it's dependant on persistence. Actually, persistence becomes a real mixed blessing.. You don't want too much of it, or you get "smearing", and you don't want too little, or you break the visual system's saccadic masking, and cause perceptual strobing. There's a very definite sweet spot in persistence, in relation to lower framerates. Obviously, as you increase the framerate, a display which runs at a higher per-frame persistence proportion would produce less smearing, though (as there's less movement per pixel across the retina per frame).

That effect is also noticable on our DK2s. I tried explaining it before as being like seeing through the gaps in a moving fence. I suppose you could also say this.. The old CRT television sets only has a resolution of 1 x 1. The single dot would start at the top left of the screen and scan along and down down it changing its brightness depending on the picture.

This does lead to a very strange phenomenon. Text is easier to read when it's moving :)
 
Last edited:
Well I'm a hobbyist game developer with the unreal engine 4. And when it comes to VR there isn't a set of do's and don'ts, as of yet. Just remember VR is just starting to come available for everyone. Even the big game developers are working what works and doesn't, ti get the best for VR without causing simulator sickness. We won't get the best VR experience for another 1 to 2 years.
 
Back
Top Bottom