Has Pimax fixed the lens sweetspot issue on the 12k QLED?

There is a problem with the lens positioning on the 8k X and all other Pimax Vision series headsets. Another user posted an illustration of this problem back in 2019. The problem is kind of hard to explain so let me link the image first.

https://community.openmr.ai/t/eyestrain-discussion-problems-and-solutions/16925/663?u=shinytomb

Pimax falls into the A illustration (albeit potentially worse than is shown). As you can see from this illustration, unlike other headsets, the angle at which you are viewing through the lenses is farther away from the sweetspot. This results in the user viewing one screen’s image outside of the sweetspot at times and instead through the outer fresnel rings. This is a pretty annoying problem. Especially for lower IPD users, since it makes one side of the image blurry in certain circumstances.

Pimax says they have increased the IPD range on the 12k from 60mm minimum to 57mm. The sweetspot also appears to be somewhat closer to the edge of the lens, but I’m still a bit concerned. Their use of an aspherical lens for the sweetspot might not be doing them any favors in this case, as aspherical lenses have smaller sweetspots compared to fresnel. Can anyone at Pimax confirm that this problem has been eliminated? @PimaxUSA @PimaxVR @PimaxQuorra

3 Likes

It’s too early to say

1 Like

I brought this question up here during MRTV’s livestream Q and A (Vincent Tuminello), I received no reply.

So… Since before 8k/5k kickstarter units started shipping, concerns have been raised about the ability to adjust eye relief, especially given the canted screen design, where varying eye relief distance moves the lenses back and forth “past” your eyes to the sides, altering the whole projection setup - something that was never addressed; One could even argue that the “comfort kit”, that eventually came out, even made the headset less adjustable, clamping it in place, hard.

(And this should actually be expanded into more fine-grained, fitment adjustability, so that you could accomodate wide and narrow faces, and different cheek/brow depth relationships.
(Only ones I’ve noticed that may have addressed this somewhat are Varjo (if I understand what reviewers are saying about its collection of headstrap knobs right), and maybe Sony (I believe the PSVR may more or less “float free” in front of your eyes, rather than relying on the friction between one’s face and a gasket?)).

IPD adjustment alone can only get you so far, and even if they have managed to make new lens give you a clear view across the entire view, and from any point in the eyebox, hitting the sweet spot can still only remain optimal, especially to be right for the peripheral fresnel portion. (/me also think IPD adjustment should be separate for each eye, to accomodate at least horizontal facial asymmetry.)

So from what it looks like on video, we still have the “fried egg” design, basically with the difference that the “yolk” part is now collapsed into zero fresnel segments, instead of many, and it looks to me it is still biconvex, so we still have a transition area between the inner and outer zone - the distortion profile is not a smoothly contiguous arc, but a more complex spline, and we still have parts of the near eye surface, that slopes away.

I don’t know how many have seen Valve’s little logotype for their Index headset… It is a disc depicting an eyeball, in front of two half-discs, representing the HMD’s two monoconvex lenses, stacked, per eye (that’s not to say the lenses necessarily do have a spherical profile like the discs - they could be parabolic or whatever, for all I know - it’s just stylisation for the logo after all, but they are monoconvex - the side toward the eye is flat), going away from the hybrid design from the HTC Vive, with fresnel on one side, and a conventional lens surface on the other.

Whilst the hybrid approach let you use the front refractive surface to compensate for certain abberations, I do believe the simplifiction it is to get rid of it, probably has benefits in reducing pupil swim, as well as user-side reflections in the lens. It also lets you get your eyeball a millimeter or two closer to the lens… :9

We’ll see what Pimax come out with, and hey - maybe they will even have managed to suss out dynamic distortion correction, using the built-in eyetracking, but maybe it could be worth making a bit of noise about this whole thing right now, whilst there is a year to go, and there may still be time to in any way influence things, to prevent bad history from repeating… :7

(Wonder whether the maybe-prototype-but-more-likely-mockups people were wearing in the video are hand-sculpted samples, or CGI, or if tooling could perhaps already be done…)

4 Likes

I still think they should just make a simple “spacer” system which would allow a user to choose between 2 or three different distances and include this with the headsets.

I.e. a minimal distance cowling plus various “spacers” between cowling and pad.

That way You wouldn’t really need different thicknesses of pads either and could just go with a “medium” thickness which is soft enough providing a good comfort and then focus on just having a few types of “outer material” available (“Index type fabric”? Velour? PU leather?).

2 Likes

Little foam “Lego bricks” would be a very simple, flexible and cost-effective way to build up one’s own personal face gasket, but undoubtedly a fiddly and time-consuming proposition for the user - especially if the device is shared by many. :9

(EDIT: It looks like foveated transport may actually be set in stone as a thing for the 12k, by the way, from what people discuss elsewhere on the forum - that was another thing we were asking the silence about back in the day, for the 8k kickstarter – maybe one could take that as a good sign for other things as well… :7)

1 Like

I did hear that the 12k is supposed to have lens distance adjustment so maybe that will help. I was watching Sadly it’s Bradley’s stream yesterday and he had a specs spreadsheet that showed it. Hopefully someone who has the problem with the current headsets will be able to test it out in January.

2 Likes

I didn’t mean “lego bricks” - Just “spacers” in between the thin/minmal cowling and the gasket/pad… :slight_smile:

This would allow a user to choose between let’s say a thin, medium or large spacer (or even no spacer)… :slight_smile:

Not much different than with the original Vive which had a thick and a thin gasket/pad, only this would just be a spacer between the gasket/pad and the actual cowling instead… :slight_smile:

1 Like

Yes - I just took that notion and picked it into even smaller discrete fragments - go voxel gasket! :7

2 Likes

How about an inflatable cushion?
Pump it up to get some distance between the eyes and the lenses!
:full_moon_with_face:

2 Likes

That could also work, but I’d think it would be better to keep it as simple as possible (there’s enough complexity already)… :slight_smile:

I mean, if none of the official and included spacers work, just print Your own custom one… :slight_smile:

1 Like

No amount of printing custom parts for the picetag on the 12K HMD should be needed

4 Likes

:7

1 Like

I don’t disagree. Just saying it’s a possibility… :slight_smile:

If they include three different “spacers” or “interfaces” I would think that would cover most if not all already… :wink:

Thing is, I’m not sure if there’s space enough for the usual “eye-relief” but of course that would be even better… :slight_smile:

What they really need to focus on as well is that stability of the HMD, looking at the renders they still use the same kind of hinges to attatch the headstrap.

This is just not going to be good enough for that price tag.

3 Likes

Yeah, they need to go back to the drawing board and make fundamental changes to the headstrap, not just tweak around the edges. Some kind of forehead strap like the Varjo headsets for comfort should probably be included. Otherwise it’s likely going to fall in to all the same problems as before.

1 Like

The presentation was really impressive, besides the one hair on the long haired guys jacket which was innocuous but still slightly bothersome (Maybe only for people who have OCD who get distracted because they zone out from the main point in the video, from hyperfixating on that hair on the blazer that they just want to not see there, as it may have been very hard for some people to ignore what could have been at times a considerably strong urge and desire which some people may have had from watching the video, to pick off said offending hair in question off of the blazer, as if they were there in person)

also white subtitles on a white background made it difficult to read at times whenever the subtitles were not on the presenters body, and the fact you guys were late by over half a day was not good for professionality’s sake and maintaining a good impression of organisation. (although to be fair in this day and age of covid everything is delayed so delays are to be expected and warranted, so no real criticism there.))

Otherwise the show was worth watching and I don’t have any more constructive criticism on the presentation. I do have two small suggestions for the headset though, but first I have a couple of questions for the Pimax team.

I was glad to see there’s going to be more binocular overlap and also that you’re working on some projector device for a 6000 PPI/70PPD display.

@PimaxQuorra @PimaxUSA

  1. Is the 70PPD projection/reflection mirror display in Pimax 12K QLED? Is this achieved by using some kind of focus display like Varjo in the middle of the display, or is the 70PPD for a device after the 12KX?

  2. And does the Reality headset have a physical minimum hardware IPD of 55mm? (Asking since there’s more binocular overlap and the 8KX can physically only go down to 60mm without using software adjustments for those of us at 60mm IPD or less)

I just have two suggestions please hear me out they aren’t that complex nor are they too design heavy or expensive.
There’s still no eye relief depth adjustment to minimise the need of custom face foam thicknesses, and no extreme IPD mod to get extra minimum IPD range from what I saw. (unless the binocular overlap is achieved with the extreme IPD mod but you guys didn’t advertise extra IPD range)

@PimaxUSA can you guys look into these two things since it should just be a matter of plastic adjustments in the housing?

a) put the face foam on a collapsible/expandable plastic fitting which can be adjusted on the headset by moving a dial to bring the eyes closer or further away from the lenses

(Index, Vive, Vive Pro 2, Reverb G2 all have this feature, so there’s no excuses to not have this feature seeing as it’s a simple plastic modification if Pimax really wants to have a “plug and play” headset which requires ease of access and easy user experience, a “no compromises” headset, comfortable means very lightweight and a “one size fits all” aka adjustable headset with no modding, extra accessories or 3rd party products required to make the headset work (it should work out of the box for 90% of users at that price tag with no modifications or accessories needed just to be comfortable and to even use the device properly))

b) implement the extreme IPD mod by reducing the length of the worm mechanism behind the plastic housing involved with moving the lenses when the IPD dial is adjusted.
Check here for info https://www.reddit.com/r/another_extreme_ipd_mod_8kx/

And here for the original mod
https://www.reddit.com/r/Pimax/comments/bq4q8f/extreme_ipd_mod/

1 Like

During Pimax’s frontier presentation, there was as graphic of the lenses. During this segment, we see that both the lenses bezels are pretty much coming into contact with each other. Or half a mm away from coming into contact with each other. This is very similar to the “Extreme IPD Mod”. So, perhaps Pimax took their users advice. I was told that Tobii is helping work on the IPD aspects of this headset, so that’s a good sign. Even still, the “Extreme IPD Mod” only gives about 4mm of extra range.

Assuming Pimax didn’t make any changes to the general position of the lens sweetspot that still leaves the two lens centers about 66mm from each other (according to one such modder). However this is an entirely different lens system, so it’s possible they’ve made other changes which could alleviate this problem. One change which would guarantee this problem is improved would be moving the center of the lenses closer to the edge where your nose sits. It’s kind of difficult to tell if this change has been implemented, since the only shots we have of the lenses are 3d models.

2 Likes

I’d have to imagine the increased stereo overlap (and the FOV increase overall), as well as the reduced minimum IPD, stems from greater “magnification” by the lenses, letting you use a smaller display panel for same FOV, and thus making room.

If it is true that the new lenses give you significantly better both edge-to-edge sharpness, and sweetspot, than the old ones (just a tiny pinprick at the very center, and then blur, for some of us; Others experience differently), it may not matter that the canting inherently sets the lens centers farther apart than your IPD - it should (knock on wood) look crisp even when looking diagonally though the lens, “entering” a bit next to its center.

3 Likes

Eye relief was one of the things I was hoping to see so I’m a little disappointed to see yet another headset without it. Stacking foam is a ridiculous and janky solution for a 2400 dollar headset. Valve had an eye relief adjustment on the very first Vive 5 years ago…this should be a solved problem…

Nevertheless if this is my only criticism then that tells that the event was quite successful.

3 Likes