A PPD of 42 should be very good for anybody who wants to use the headset in a VR desktop application, it’ll be amazing for watching movies web browsing, Cad work, if they ever launched a stereo 3D full color pass-through plate, I could see that working really well for AR applications. There are a lot of uses, and the best part is iyou have the choice.
1 Like
good point i didnt think of game textures might show more flaws that you couldnt see before
1 Like
I know with the aero that it’s 35ppd in a center ring of view only, then as u go out from the center the ppd drops, to something like 27ppd (iirc) at the edge. Maybe the crystal has similar, it’s 42ppd in the center only and it drops as u look away from the center, it may drop to 20ppd at the edge for example. It might explain the numbers, no?
2 Likes
iirc 8kx is 21, G2 is 24, arpara is 32.
3 Likes
if i buy the crystal i intend to make use of 42ppd. I wouldn’t upgrade from an aero to just have the same clarity. But i can see why if i never already had an aero, if i was buying the aero or crystal, then crystal would be the better proposition on paper.
I game at 35ppd in all of the games i play regularly (alyx, walking dead, ams2, ac, ets2, ats, msfs etc), the only 1 where fps is a cause for concern is msfs (but i just accept low fps/less smoothness and still use 35). Gpu makes all the difference (not cpu) at such high resolutions, my heavily oc’d 3080ti just about manages for the most part. 4xxx series will deffo be needed for 42ppd imo.
3 Likes
MSFS is an outlier. It’ll melt your GPU but even if there were no GPU limit, it still wouldn’t be able to achieve anywhere close to 90fps due to being CPU bound.
2 Likes
Yeah, as a Marketing Director. A premium product which the Crystal undoubtably now is in the class of (even prosumer range) should be premium right across the marketing mix. Ie everything to do with it: the software, the headstrap, the build quality, the packaging, the comfort, the warranty, the support.
It definitely shouldn’t be a janky fisher price toy that you have to heavily mod and has loads of broken bits. Lol.
8 Likes
I’m hoping the support for it will match as well. Even the best companies will inevitably have some products with problems and so having good support is vital for when that does happen. And either baked in or as an optional purchase I’d hope Pimax might consider a warranty longer than 1 year for something this expensive that’s supposed to be future proof enough to last for a few years.
3 Likes
Pimax has directly said that the 42 PPD is only in the center of the lens. The “narrow” lens concentrates PPD at the center while sacrificing PPD at the edges. The “wide” lens has the PPD more evenly distributed across the whole display.
This may not be the same thing as what you’re talking about with the Aero. What you describe with the rings sounds like foveated rendering. Whereas what Pimax is doing is a lens distortion effect.
Note that the narrow lens is also sacrificing total FOV as well versus the wide lens. Going down to 110 degrees from 125 degrees horizontal. It’s the combination of narrowing the total FOV and concentrating pixel density at the center that reaches 42 PPD.
6 Likes
I guess we might as well make sure to put it down in writing, rather than just leaving it to the implicit common: “As we all know…” understanding, that all lenses distort the image, and every “modern” HMD has this pincushion distortion, which concentrates resolution in the center.
That is part of this consumer-focussed VR “revival” era; When it was determined that given we had reached a point where we can afford to compensate for distortion in software, we could skip expensive and heavy correctional optics, and just use a single cheap lens, and you got people like Palmer Luckey duct-taping simple HMDs together in their spare time.
Of course, how much distortion each lens has can vary.
4 Likes
Oh, that would be really funny if Pimax asked 2,499$ for the 12K but then tried to get away with a paltry 1 year of warranty.
Sure, I’ll dish out that kind of money but let them off the hook after a year… especially considering the reputation Pimax have in terms of QA…
How is this calculated by Varjo ? At the end, you have a given number of pixels on the displays and the FoV is clear too. One unknown remains the percetage to which the display is actually utilized - it will never be 100%, but if there was a greater difference between Varjo and Pimax on that part it could influence the consideration too.
1 Like
Right. So rather than fighting this distortion, Pimax is intentionally accentuating it to magnify the center of the image. Since it’s an aspheric lens, it’s possible to do this selectively. Something like varifocal glasses. So theoretically the design could be both intentionally distorting more in the center and also distorting less around the edges.
I think this is pretty ingenious. As it’s a way to produce the same kind of effect that Varjo gets in their commercial VR headsets (albeit only 42 PPD rather than 70 PPD) with a second higher density display projected in the center but at much lower cost because there is no second display. It’s just a different shaped lens and changes to the software correction profile. Brilliant!
I actually wonder if part of Pimax’s intention here is to dip into the commercial market with a VR headset which is much cheaper for a mid-capability if the full 70 PPD isn’t really needed in the application. $1900 versus $3400 + $1500 per year is a significant savings.
OR, Pimax did what they always do: exaggerate their numbers (like their FoV numbers). To me that sounds more likely (it’s not just Pimax though, most companies exaggerate their FoV)
1 Like
I think one of the really smart moves with this approach is that it will get the effect of a VR-3 focus display, but unlike Varjo’s 70ppd foveated display, this will work with any and all software out of the box.
42 PPD should be enough that at least part of the display viewed through the lens should be able to resolve all the detail of a 720p display. If the backlight and strobe tuning is really good, this could be an LCD that looks very close to what an old tube looked like in motion, which will be worth it, just for legacy content.
I already know New Retro Arcade Neon with CRT shaders and other emulators are going to be amazing on this thing.
I know you and others mentioned a lack of texture detail in current VR apps, but that can be fixed by developers, or community based texture packs. At least now with the crystal, we could see a developer lead push to improve the quality of those games meant to run on Quest 2 era hardware.
2 Likes
720p quality at 42ppd cant be right,most people on aero discord if not all said its similiar to a 1440p monitor,just booted gmod in 720p,and then on my 8kx at its full res(gmod has a vr mod),8kx is way sharper inside,than what i see on my 32 inch monitor at 720p.
3 Likes
Good to hear, I was going off of armchair and in my head math, and thinking in terms of how a 720p small monitor, like 24 inches would look.
I am happy to be underestimating, because 42PPD will be better.
3 Likes
Actually, I would say that Varjo has been the most guilty of this by far. I have another post on these forums where I was calling out Varjo for straight up false advertising with their FOV claims. Yes, other companies post numbers that are higher than you’ll really see, but Varjo’s were substantially farther removed from real world measurements than those of other companies. The Aero’s FOV claims appeared to be just plain made up.
The thing is, these other companies (including Pimax) are not just making up these numbers. They’re just calculating them differently than we as consumers would calculate them. It’s analogous to car manufacturers claiming horsepower based on readings with the engine on a stand with no accessories. It’s a real figure, but it’s not what you’re going to see at the wheels. And if one company is posting bigger numbers this way, the others have to follow suit.
I think the FOV being claimed by these companies is the FOV of the display panels themselves relative to eyes in the most ideal positions. The use of the panels for the measurement would be where “diagonal FOV” comes from, which is a completely useless measurement for VR because lenses are round. Real world diagonal FOV is smaller than horizontal FOV. So it’s all kind of silly until a real standard for how FOV should be measured in VR comes into existence.
But… if you compare claimed FOVs between the various manufacturers, you’ll actually have about the right idea as to how they stack up against each other. Just make sure you always compare real world measurements against real world measurements, and claimed FOV against claimed FOV, and you’ll be good… except for Varjo. Their claims for the Aero put it in Valve Index territory, and yet in reality it has the smallest FOV of any headset on the market… even smaller than the Quest 2… and not by just a little.
The really interesting thing is that this changed with a firmware update on the Aero. Suddenly the FOV of the Aero got bigger (?!!). That seems like it should be impossible. But I think what it means is that Varjo was blacking out the edges of the lenses and not using those pixels because they were too distorted. Much like normal mode on the Pimax 8KX just blacks out pixels at the periphery. The firmware update supposedly fixed the distortion problem on the Aero substantially. And I think they must have opened up using those additional pixels and widened the FOV to what it was actually capable of because they were able to get acceptable results at the edges now.
I actually haven’t seen anyone re-measure the real world FOV of the Aero post update. And I’m really interested to see how much it actually expanded. Is its horizontal FOV actually close to the Index now per Varjo’s claims? The vertical FOV surely still isn’t.
This is a relevant question to raise again because it’s entirely germane to the discussion about the FOV(s) of the Crystal and what’s possible there and how it compares to the Aero… as well as the Index and Quest 2.
3 Likes
I think this may be an error in NextGen’s video. Specifically, I think this came from the MRTV interview about the 12K, and he misremembered that as applying to the Crystal release.
I don’t quite get your Point
, it’s basically an 8kx on less FOV so more ppd with better quality lenses… If you liked the 120ish degrees and want all the details sharp in there, then this is your device, if not well 8kx or 12kx… And ofcourse it’s not going to help if you still have a Voodoo Card in your system. 3080ti or better.
For the lens part it’s like magnifying, or zooming on a Beamer, more area more pronounced pixels. Only difference here is that you can consider the changes in the rendering pipeline and provide more or less FOV.
And since Pimax makes HMDs seems natural we are talking about that… Sure a 6090 won’t hurt, but that’s another story.
1 Like
Point is squeezing same number of pixels into a smaller FOV doesn’t improve everything overall, sure you may have a preference, clarity vs. FOV, but it’s a zero-sum game unless you bump the GPU. And you can increase the resolution to a billion subpixels, but it won’t improve your game you’re running at 70% Steam VR all that much - think people who play 1080p game with their potato systems on a 4k monitor. Just saying: All things being equal, anyone purchasing a reality series HMD without at least a 3090 has their priorities out of balance.
1 Like