New headset : Pimax Crystal

For sure - actually a reason I'm thinking the Crystal is probably a better choice than the 12k would be anyway.

Here's hoping that as more of these headsets with eyetracking arrive (there are a few now/soon available), games will finally start natively supporting dynamic foveated rendering which would have more performance impact than eg OpenXR toolkit can achieve with it.

But yeah, I'll be watching AMD's announcement tonight with interest!

Red Matter 2 had it added in for the quest pro extremely quickly. Now it might not be the absolute best implementation ever, but it can clearly be integrated quickly and straightforwardly for at least one platform/build, so I'd hope it's not terrible for most of them.

The benefits of good eye tracking are so immense. Standalone has a good chance of looking great, PCVR can look awesome, smart use of wireless can also potentially allow it to look awesome over wifi.

EDIT - I wonder if Virtual Desktop could integrate eye tracking in terms of streaming the data and focusing mostly on the tracked section?
 
Last edited:
I’m afraid that the 12k will bring any current or near future GPU to its knees in the games I play… my thinking is that the Crystal provides a pretty substantial FoV and PPI upgrade over my G2 or Pico 4, without setting me back 2 GPU generations in terms of performance with dynamic foveated rendering probably offsetting at least some of that extra resolution.

Must admit I’m not a huge fan of that enormous form factor though, how do you get on with the 8KX in terms of becoming what can only be described as a techno hammerhead shark? Do you notice the size of the thing when wearing it?
i'm not so pessimistic about the demands of the 12K although i do see your point. just like the 8KX, the fov/resolution will be changeable to suit the demands of the game you are running. but yeah it's gonna need a 5090/6090 to run at that fov/resolution in most games, but i intend to keep it a few years anyway (so 5090s etc will be out) given the price and i don't expect to see it until atleast the end of 2023.

the size is not as bad as it appears and the crystal is supposed to be a bit lighter than the 8KX. you do see complaints about it but i wonder how many have not got the hmd adjusted properly as proper fit makes a lot of difference. you do notice the size when picking it up and putting it on but you soon forget it once playing a game.
 
Red Matter 2 had it added in for the quest pro extremely quickly. Now it might not be the absolute best implementation ever, but it can clearly be integrated quickly and straightforwardly for at least one platform/build, so I'd hope it's not terrible for most of them.

The benefits of good eye tracking are so immense. Standalone has a good chance of looking great, PCVR can look awesome, smart use of wireless can also potentially allow it to look awesome over wifi.

EDIT - I wonder if Virtual Desktop could integrate eye tracking in terms of streaming the data and focusing mostly on the tracked section?

It’s called foveated transport and the Crystal already will do it apparently... one way they can push such high res and refresh displays over 1.4a. I imagine Guy can probably add it into VD too but there hasn’t been a need to date. Maybe Quest Pro would make it useful.
 
Last edited:
i'm not so pessimistic about the demands of the 12K although i do see your point. just like the 8KX, the fov/resolution will be changeable to suit the demands of the game you are running. but yeah it's gonna need a 5090/6090 to run at that fov/resolution in most games, but i intend to keep it a few years anyway (so 5090s etc will be out) given the price and i don't expect to see it until atleast the end of 2023.

the size is not as bad as it appears and the crystal is supposed to be a bit lighter than the 8KX. you do see complaints about it but i wonder how many have not got the hmd adjusted properly as proper fit makes a lot of difference. you do notice the size when picking it up and putting it on but you soon forget it once playing a game.

My concern (not that it's ever going to happen for me) buying such a top of the line headset now, that can't be fully powered yet, is by the time it can be, there's going to be much better tech out there to help it. Either improved lenses, screens, haptics processing, wireless etc etc, meaning now I can get a 6090 to fully drive it, there's also miles better headsets out there that can also be driven by a 6090.

Equally, if I could afford it without worrying, then why the hell not....

It’s called foveated transport and the Crystal already will do it apparently... one way they can push such high res and refresh displays over 1.4a. I imagine Guy can probably add it into VD too but there hasn’t been a need to date. Maybe Quest Pro would make it useful.
I'll try and look it up, very useful to be used in conjunction with etfr. Especially with the potential of AV1 decoding in headsets, and encoding on graphics cards. The wireless experience is looking to be rather special.
 
@tyke_ thats good to hear, the form factor is the most off putting thing to me after seeing headsets like the NEO 4 and Quest Pro going a very different direction in terms of bulk.

Either way looking forward to reviews, on paper it sounds pretty amazing.
 
My concern (not that it's ever going to happen for me) buying such a top of the line headset now, that can't be fully powered yet, is by the time it can be, there's going to be much better tech out there to help it. Either improved lenses, screens, haptics processing, wireless etc etc, meaning now I can get a 6090 to fully drive it, there's also miles better headsets out there that can also be driven by a 6090.

Equally, if I could afford it without worrying, then why the hell not....


I'll try and look it up, very useful to be used in conjunction with etfr. Especially with the potential of AV1 decoding in headsets, and encoding on graphics cards. The wireless experience is looking to be rather special.

Yeah I think we are finally getting closer to a wireless experience where the trade offs are minimal, if even noticeable… the main problem will be running out of battery lol.

I agree with the above… to some extent you could argue the Crystal and 12K are more “future proof” than some with the wigig modules, face tracking modules, expandable faceplates and removable lenses etc, but then in the face of solid state varifocal lenses and 4k x 4k 10,000 nit uOLEDs on the metaphorical horizon is anything really future proof as such?

the only thing that gives me pause is a possible index 2, but who the hell knows when that will arrive. I might be run over by the number 79 from Horsham before then so I’m more of a carpe diem kinda guy!
 
Last edited:
Yeah I think we are finally getting closer to a wireless experience where the trade offs are minimal, if even noticeable… the main problem will be running out of battery lol.

I agree with the above… to some extent you could argue the Crystal and 12K are more “future proof” than some with the wigig modules, face tracking modules, expandable faceplates and removable lenses etc, but then in the face of solid state varifocal lenses and 4k x 4k 10,000 nit uOLEDs on the metaphorical horizon is anything really future proof as such?

the only thing that gives me pause is a possible index 2, but who the hell knows when that will arrive. I might be run over by the number 79 from Horsham before then so I’m more of a carpe diem kinda guy!

Index 2 is meant to be early next year I believe. But that's just the rumours.
 
Index 2 is meant to be early next year I believe. But that's just the rumours.

I doubt Valve will release anything next year. The bulk of the info about the supposed next Valve headset is from Patents filed a few years ago. The rest is just speculation and conjecture from You Tubers grasping at straws. For all we know Valve might not be working on another headset at all. It wouldn't surprise me if Valve have moved onto other things. Like they always seem to do with any hardware venture.

If you go back to before the Valve Index released. Even almost a year before release we had concrete info, photos of the headset, the knuckle controllers protypes etc. Now we have nothing, just some Patent filings :(
 
I might be run over by the number 79 from Horsham before then so I’m more of a carpe diem kinda guy!
I think in a similar way :D (well, it's how i justify my spending on stuff like hmd's and gpu's etc)
I've reserved a Crystal too now, my resolve to not to ran out in a few days :rolleyes::D
 
The thing I'd like know is whether the dynamic foveated rendering is done at the hardware level and can be applied to all your VR titles, or it only supported in a select number with developer support?

cos what is the use of having massive per eye resolutions only to jump into DCS or MSFS and instantly hit reprojection and large GPU frame render times
 
Ah nuts, just seen the Crystal is Nvidia only…

With the 7900XTX vs 4090 value proposition that’s more than a little annoying.

Get the 16GB 4080? Nvidia has been better for VR up until now and I can't see that changing with this generation. And with their new chiplet design, I doubt VR is any sort of priority for AMD.
 
The 4080 looking like a pretty underwhelming product imo... especially in a world where the 7900XTX exists - can't see myself upgrading from the 3090 to that. If I go Nvidia it will be a 4090FE unless they bring out a 4080Ti in response to AMD.

What you say is true enough, Nvidia has historically been better in most VR games though I suspect a lot of that was due to how they scaled better with higher resolutions (they also performed better vs the RDNA 2 GPUs at 4k than they did at 1440p for example) - we don't yet know how RDNA 3 handles higher resolutions but it seems to be a priority for this architecture with 4k performance being a focus. Need reviews to see how they actually compare in VR this time around...

Where AMD clearly does fall down in VR is the frameworks that Varjo and now Pimax are using for some of the more advanced features like foveated rendering. It certainly isn't the case that AMD can't do it - the openXR toolkit can apply foveated rendering on AMD GPUs for example, but they clearly don't provide the same level of tools to developers as Nvidia do.

We shall see... just mostly annoyed that on a potentially quite long lasting headset (especially with the modularity), one could get locked in to potentially a couple of generations of Nvidia cards or face changing the headset. I don't know who to be annoyed at, AMD or Pimax, but I just know I'm annoyed lol.
 
The 4080 looking like a pretty underwhelming product imo... especially in a world where the 7900XTX exists - can't see myself upgrading from the 3090 to that. If I go Nvidia it will be a 4090FE unless they bring out a 4080Ti in response to AMD.

What you say is true enough, Nvidia has historically been better in most VR games though I suspect a lot of that was due to how they scaled better with higher resolutions (they also performed better vs the RDNA 2 GPUs at 4k than they did at 1440p for example) - we don't yet know how RDNA 3 handles higher resolutions but it seems to be a priority for this architecture with 4k performance being a focus. Need reviews to see how they actually compare in VR this time around...

Where AMD clearly does fall down in VR is the frameworks that Varjo and now Pimax are using for some of the more advanced features like foveated rendering. It certainly isn't the case that AMD can't do it - the openXR toolkit can apply foveated rendering on AMD GPUs for example, but they clearly don't provide the same level of tools to developers as Nvidia do.

We shall see... just mostly annoyed that on a potentially quite long lasting headset (especially with the modularity), one could get locked in to potentially a couple of generations of Nvidia cards or face changing the headset. I don't know who to be annoyed at, AMD or Pimax, but I just know I'm annoyed lol.
I bought an MSI 4090 on launch day, it's given a really nice boost to my VR gaming. Would've preferred an FE but all sold out in seconds. Though I think you'll get 1 if you want 1 and are prepared to be patient, and speedy when drops happen. I didn't consider AMD because of the incompatiblities.

I have a reservation in for the Crystal. I wonder how Pimax will handle the issues with Aspheric lenses like the Aero has.
We'll hopefully soon see on the 19th with roadshow feedback coming through. I think distortions will still be something people can spot in the controlled environment of a roadshow. Would've preferred reviewers like MRTV get a Crystal in their hands at his base but Pimax must not consider it ready for that just yet. I do think the roadshow will be useful feedback though.
 
Yeah no doubt the 4090 would give a welcome boost - just think it’s a bit much money this time around. I thought I was mad for £1300 on the 3090FE.

Only saving grace is that I guess I’d get ~£650 or so from the 3090 sale.

Back more on topic I was hoping we would maybe see a bit more of the crystal at the frontier event last night but they launched the craziest thing instead - the portal.

I’m a bit concerned that pimax are biting off more than they can chew with the portal, I’d rather they just focused on bringing to market the products they already announced… the portal seems like a (overly complicated) distraction.
 
We'll hopefully soon see on the 19th with roadshow feedback coming through. I think distortions will still be something people can spot in the controlled environment of a roadshow. Would've preferred reviewers like MRTV get a Crystal in their hands at his base but Pimax must not consider it ready for that just yet. I do think the roadshow will be useful feedback though.
Quite a bit left to do by the looks of things:
1. MRTVs comments about the non-neutral diopter on the lenses
2. The WiGig module not being on show
3. Tobii eye tracking in but disabled so no automatic IPD adj
4. Presumably no Foveated rendering with the eye tracking off

Lets see if Pimax was listening or just doing lip service and marketing for a product that they know is locked down for mass production
 
Quite a bit left to do by the looks of things:
1. MRTVs comments about the non-neutral diopter on the lenses
2. The WiGig module not being on show
3. Tobii eye tracking in but disabled so no automatic IPD adj
4. Presumably no Foveated rendering with the eye tracking off

Lets see if Pimax was listening or just doing lip service and marketing for a product that they know is locked down for mass production

Also the controllers don't function well at the moment.
 
Back
Top Bottom