Is VR dead?

[RCHeliguy]

Don’t believe a word mrtv says (especially about pimax)

Yeah, he said good things about the G2 controllers. G2 battery life wasn't good.

My original Rift would run off a single AA for days. I did use rechargeables. The G2 uses 2 x AA and didn't last nearly as long.

My 3.5 year old Index controllers with built in rechargeable batteries would also last days on a single charge.

 

[gamma123152]

I didn't watch all this, but this guy also commented about the controllers I think in this video:

And here too:

 

[Leynad777]

Agree with the performance part. First thing I do when the 5090 releases is buying that GPU. But I'm tired of the Fresnel lenses and the panels of the G2. So I ordered the Crystal. I expect that I can run it with equal resolution as I run with my G2 which is 3400*3400px per eye 90hz/fps locked(4090 with fixed foveated rendering enabled and 4/8x MSAA / TAA depending on the title) and I expect that this will look better then with my G2. I don't agree that 0.75 "will look horrible", according to flight sim guy the difference between .75 and 1.0 is minor/very small. Of course you rather want 1.25 and 0.75 will look horrible on the G2 for example, but the Crystal has a way high Resolution/PPD so it couldn't be compared like that. But better the Crystal now at 0.75/3400x3400px and later the crystal with 1.25 rr / 4400*5500px or something with the 5090 then living 2-3 more years with Fresnel lenses and outdated panels.

The only times I'm pissed off the fresnel lenses is during the start of AMS2 with the Reiza-logo on black and in the rF2 UI. Otherwise I can't remember I've noticed it ever outside UI and I guess a larger edge-to-edge clarity compare to the G2 helps as well. But that's a good thing about sim-racing. Night racing is not that much of a thing and if so, the glare looks like reflection effects you turned off to make it run At least I had some night-flights in MSFS and didn't notice it and the glare of my Pico 4 is quite similar in annoyance.

I can't find any information what render-quality (0.5 to 2) actually means and so it's what Pimax defined what it is. With the 8KX the difference between 1 and 1.25 is already like between Index and Pico 4. Reducing it to .75 is like going back to the CV1. With the default 1.0 it's not 4k at all and you still have the typical fuzzy visuals in the distance you know from 2nd gen headsets. With 1.25 it's similar to 4k on a 60" TV and what you would expect from a double 4k headset. Just did some testing and switching from high to medium MSAA in AMS2 plus fixed foveated rendering on conservative did the trick to run it on 90 Hz. I rather tune everything down to run 1.25 rendering-quality than running ultra on the default 1.0.

I guess that Pimax is simulating different headsets with this slider for SteamVR, because that's how it looks. The 8KX came out during the 20-series cards if I'm not mistaken (at least the 8k) and so it also needed to run with mid 10-series cards. The Crystal is certainly not made for that GPUs and is mocking HMDs for more recent GPUs with the rendering quality.

 

[metalnwood]

Just saw someone complaining about all the Crystal reviews and how they never mention the controllers, controller tracking, controller battery life, and a number of other topics.

Then someone explained they have read the agreement that reviewers have to sign to get a Crystal to review and they specifically make a large number of issues and items off limits to talk about.

When the newspaper tells me that 'sources close to the xxxxxx say xxxx'

Thats a license to say whatever you want without scrutiny so I will wait to see what people are really saying

 

[Leynad777]

Question: Is there are way to use all the 2D-apps in VR like going with alt-tab out of a game? Virtual Desktop on Steam seems great to use all the 2D-stuff, but VR-games launched inside VD go into theatre-mode instead of VR and if I launch a VR-game inside Steam it quits VD. What a joke.

 

[gamma123152]

I can't find any information what render-quality (0.5 to 2) actually means and so it's what Pimax defined what it is.

It's the render resolution. With the Crystal 1.0 is somewhere like 4350x5100px. You can see that resolution in SteamVR's settings/FPSVR or OpenXR toolkit.

So with .75 it's 3.262x3.825px. It will probably still look okay since that's around 120% which I run with my Reverb G2+4090 currently(3400x3400px).

It differs from headset to headset what that resolution actually is(with the 8Kx .75 is a smaller pixel number then with the Crystal) and how much is actually used for barrel distortion correction and the FoV also matters. With a small FoV(like with the Reverb G2) that same number is much sharper in resolution then when it's spread over a huge FoV like with the 8Kx(especially in the larger FoV mode's). So .75 on the Crystal is a lot sharper then .75 on the 8Kx.

So stating the .75 always looks like going back to the CV1 is not true. It depends on the lenses, the FoV(how the PPD is spread over the lens), the amount of pixels that is used for the barrel distortion and the resolution of the panels.

But yes, you want this as high as possible, that also what I aimed for with the G2. It makes a huge difference to render between 90% or 100% and 120%. And with OpenXR toolkit you can fix the resolution at a specific number, that's what I do for all sims.

 

[RCHeliguy]

We lost EVGA and now ASUS is suspect. What are we going to have left to pick from?

 

[Deleted member 197115]

We lost EVGA and now ASUS is suspect. What are we going to have left to pick from?

Wow, some youtube personality abandoning ASUS sponsorship.
The world must be coming to end.

 

[gamma123152]

Wow, some youtube personality abandoning ASUS sponsorship.
The world must be coming to end.

You can still buy MSI and Gigabyte if you want some proper quality

 

[Deleted member 197115]

Speaking of ASUS, did you guys see this.

https://www.gamespot.com/articles/asus-rog-ally-preorders-are-live-now/1100-6513303/

 

[RCHeliguy]

So none of his points about MB's with beta firmware fixes Asus provides voiding the warranty make any sense?

For $800 and $1000 motherboards, I guess I would expect a bit better.

 

[Leynad777]

It's the render resolution. With the Crystal 1.0 is somewhere like 4350x5100px. You can see that resolution in SteamVR's settings/FPSVR or OpenXR toolkit.

So with .75 it's 3.262x3.825px. It will probably still look okay since that's around 120% which I run with my Reverb G2+4090 currently(3400x3400px).

It differs from headset to headset what that resolution actually is(with the 8Kx .75 is a smaller pixel number then with the Crystal) and how much is actually used for barrel distortion correction and the FoV also matters. With a small FoV(like with the Reverb G2) that same number is much sharper in resolution then when it's spread over a huge FoV like with the 8Kx(especially in the larger FoV mode's). So .75 on the Crystal is a lot sharper then .75 on the 8Kx.

So stating the .75 always looks like going back to the CV1 is not true. It depends on the lenses, the FoV(how the PPD is spread over the lens), the amount of pixels that is used for the barrel distortion and the resolution of the panels.

But yes, you want this as high as possible, that also what I aimed for with the G2. It makes a huge difference to render between 90% or 100% and 120%. And with OpenXR toolkit you can fix the resolution at a specific number, that's what I do for all sims.

I've never said that 0.75 "always" looks like a CV1, actually very much the opposite and if 0.75 on a Crystal really would mean 25 million pixels (triple 4k), I doubt a 4090 could push that with 90 fps and Pimax wouldn't call it render-quality on 0.75.

For a native 2880 x 2880 pixel resolution every pixel density above 4000 x 4000 per eye (factor 1.4) doesn't make sense because there are no benefits. So the pixel-numbers in fpsVR can't be the native render-resolution and bumping them up in SteamVR for example doesn't have a significant enough impact in performance either.

 

[RCHeliguy]

3nm
and... 2-2.6 X performance increase, BUT they aren't reaching those goals at the moment and are increasing the number of cores to meet those goals. And they "claim" that this will be a REAL 2-2.6 increase in performance, not an increase by something like DLSS features.

New architecture and higher clock speeds are supposed to be the main reason.

As always fingers crossed that this is real.

 

[cnhoff]

3nm
and... 2-2.6 X performance increase, BUT they aren't reaching those goals at the moment and are increasing the number of cores to meet those goals. And they "claim" that this will be a REAL 2-2.6 increase in performance, not an increase by something like DLSS features.

New architecture and higher clock speeds are supposed to be the main reason.

As always fingers crossed that this is real.

Lol...Here we go again

 

[RCHeliguy]

Lol...Here we go again

Never! Well surely not! I mean it's not like they've misled us before.

 

[metalnwood]

Not looking forward to it unless they go back to the old pricing model, i.e. prices that might increased but not tied to the performance, just tied to the model it is replacing.

If they stay broadly on the current pricing based on performance its going to be awful for gamers, may be good for commercial users.

 

[Spinelli]

I didn't watch all this, but this guy also commented about the controllers I think in this video:

And here too:

This post of mine has nothing to do with the Pimax Crystal but just generally wondering: How do VR apps measure FOV? People increase FOV of their headsets by modding them, pushing them further into their face, etc. Also, don't all VR headsets have the lenses or screens at different distances to the person's eye from other VR headsets? All that makes a difference on the FOV but how does the game/app know all that data and to adjust so that the FOV is accurate? I could throw on a FOV app on my PC right now and the game will have no idea how big my monitor is nor the distance my eyes are from the monitor. I don't understand.

The above question can also be applied to the world scale. Assuming all VR games are pursuing a 1:1 FOV / world scale, how can this be done without the game knowing the screen and/or lense size and their distance from the person's eyes?

Think about when we use FOV calculators to get 1:1 FOV / world scale when using monitor/s. How does VR accurately and correctly do this?

 

[metalnwood]

This post of mine has nothing to do with the Pimax Crystal but just generally wondering: How do VR apps measure FOV? People increase FOV of their headsets by modding them, pushing them further into their face, etc. Also, don't all VR headsets have the lenses or screens at different distances to the person's eye from other VR headsets? All that makes a difference on the FOV but how does the game/app know all that data and to adjust so that the FOV is accurate? I could throw on a FOV app on my PC right now and the game will have no idea how big my monitor is nor the distance my eyes are from the monitor. I don't understand.

The above question can also be applied to the world scale. Assuming all VR games are pursuing a 1:1 FOV / world scale, how can this be done without the game knowing the screen and/or lense size and their distance from the person's eyes?

Think about when we use FOV calculators to get 1:1 FOV / world scale when using monitor/s. How does VR accurately and correctly do this?

You mention the two factors that are involved, the world scale and the fov.

The fov in the headset is constant, how much you see of its potential is different person to person and mod etc. i.e. the game does not change the fov, the closer you get to the lens the more you may see. like looking through a 5cm hole in a piece of paper at arms length, you only see a small fov, bring it closer and you see more, bring it up to 2cm of your eye and you get a fov like a typical vr headset.. interestingly the scale of the world doesnt change whereever that paper is, just like it also doesnt when you look through the headset, even if you put the headset at arms length and look at an object while bringing the headset towards you, the object you are looking at wont change scale while the apparent fov increases. Not the rendered fov of course, just the snippet you see.

It relies on the world scale being right for the person using the headset, so that their ipd is set correct which will lead to you seeing the correct world scale. As far as the game, all it needs to know is delivered to it in the steamvr/openxr api layers.

When you throw an FOV app on the computer it doesnt know the screen size because there is nothing in the api between the monitor/gpu/os to provide that information. If there was then the fov app would query the additional information to be able to make the calculation.

As this kind of thing is more important to vr working the needed information is made available through the api layers.

 

[RCHeliguy]

The VR Flight Sim Guy is saying in Aero/Crystal AB comparison that the Aero is sharper when looking in the distance. He says either it's the Pimax's Mura or the higher quality lenses in the Aero for this. He also is still saying that he still prefers the Aero over the Crystal.

And once again proving that we still haven't had a review of an actual production Crystal, Pimax is going to ship him a "production copy". I can't imagine that they would go through the trouble of sending him another unit unless there was some noticeable change despite them telling him what they sent him last was 99.999% production. He recently changed that to 99% production.

It would seem Pimax always publically says that it's almost done and yet they are continuing development of the hardware. That's a good thing or bad thing if you have preordered. It hopefully means that what you finally get will be improved over anything to date. However getting an early copy may also mean that you miss out on improvements made with each small production run.

Do his videos say one thing and his comments tell a different story? I am curious what he is still not allowed to talk about at this point that would impact his thoughts. It seems like a roller coaster where he is impressed by the Crystal Visuals and then for some reason that he can't talk about he goes back to the Aero. I'm sensing some frustration. My guess is that he has given them feedback of issues to fix that he can't talk about.

The comments are humorous as well. People really want to believe that Pimax has changed and keep making hopeful comments to that effect.

 

[cnhoff]

I have stopped waiting for new videos from him on the Crystal. His videos, albeit mentioning the Aero being an amazing headset, in essence clearly state he thinks the Crystal has the better package. Those single comments don’t fit into that.

And yeah it is funny how on both the Crystal and Aero side there are those „I want to believe guys“, that take every grain of information and use it to strengthen their point.