Eyestrain discussion / Problems and Solutions

events
issue

#665

…which is exactly what Risa demonstrates, in a configuration some of us determine fairly reasonable, given the engineering constraints, which guide the design parameters.

…but once again: The sweet spot is not something you “look at”; It is a location in space, where you (or to be specific: “the pupil of your eye”, is.

It also optimally expects you to look along the axis of the lens - perpendicularly through its centre.

…and, also again, there is nothing magic about the centre of the lens, that somehow makes everything clear, regardless of which direction from which you are peering through it.

Bringing the lenses closer together could indeed offer more stereo overlap, *IF there is more screen behind them (…and, with canted screens, it is not too far away in front).

I would certainly appreciate more overlap, and more clarity, but… (as it happens, btw; Given my low IPD, there is less room for more overlap, than for somebody with wider, since I’m closer to the inner edges, where the screens impose on one another’s space.)

A little story about different tradeoffs (numbers not properly recalled, and taken out of thin air, to spare myself the research :P):

So you had the HTC Vive and the Oculus Rift CV1.

Both these devices have a 1080x1200 pixel screen for each eye, but they look very different.

With the Vive, these 1080 pixels horizontally are spread over just under 100-ish degrees for each eye, and when you use the device, it gives you almost 100% stereo overlap.

With the CV1, Oculus balanced things slightly differently: By giving each eye only 80-ish degrees of field of view, their 1080 pixels didn’t need to stretch as far, which blessed their device with a 20% resolution (not in bitmap size, but in real world measurements: Pixels Per Degree) advantage over the Vive, at the cost of a more claustrophobic experience.

To ameliorate this situation, the Oculus engineers decided to skew the lens-to-screen alignment, and the camera frusta for the left and right eye game camera, so that the left eye could see 5-10-ish degrees farther to the left than the right eye, and vice versa.

This way they sacrificed some stereoscopic overlap for a greater overall FOV, both eyes taken together. This still left this resulting rectangular-ish field of view, remaining 70-ish degree stereo, and two times 10 mono added together, fitting entirely within the round stereoscopic portion of the Vive’s, but it was an improvement of sorts, allowing them to keep the higher resolution, and with the slightly offset lenses, it felt to most users as if the Rift had a significantly larger clear area part of the view, than the Vive, since at least one eye was in focus out to the side, even though the other was beginning to blur – this was strengthened by a simple matter of relativity: X degrees of 80, can be perceived as more than the exact same X of 100.

Oculus’ marketers also came up with the notion of presenting the FOV of the CV1 as the diagonal of that rectangular view, rather than the width.

…you may be picking up a parallel or two, about now. :stuck_out_tongue:

I haven’t made any proper measurements, but I am pretty sure the binocular overlap of the 8k/5k is even worse than that of the CV1. For some reason this still doesn’t feel as extremely uncomfortable to me as that of the Rift’s. My speculation is that this is because whereas the per-eye views in the CV1 ends abruptly, as a vertical edge which is sharply in focus (whilst the other eye keeps going for 5-10 degrees), the (even earlier) corresponding edge in the 5k+ “fades” out of focus, so that the terminator isn’t as jarring – something of a silver lining to a severe weakness. :stuck_out_tongue:


#666

This is exactly the benefit in figure B and advantage over figure A I was referring to by suggesting smaller lenses

Bringing the lenses closer together with smaller lenses would undoubtedly be a an ideal cost effective solution for low IPD users with no real lens redesign required, to get clarity and an image in both eyes when looking left to right whilet also increasing stereo overlap for people unable to see one lens when looking left or right due to having low IPD.

I appreciate all each of your inputs so far on this subject once again.


#667

I take it by “smaller lenses”, you refer to the taking a nail clipping out of the inner edge of the existing lenses, and fitting them back in…?

I am about to go to bed, but will leave you with the suggestion that you consider what, if anything, you would be giving up, if you did this.

Other than that: A pair of engineering matters, off the top of my head: You would need to adjust the barrel distortion, so that it takes the (EDIT2: pupil swim) skew into account (something we have been hoping to see coupled with eye tracking, in any case), and whether there is in fact more screen to see toward the nose.

There have been many suggestions for how to gain clarity across the entire view (EDIT: Not just Pimax - these things used to frequently crop up on the old Oculus DK1 developer forums), without radically changing the engineering solution, but sticking with the simple eye box construction: Curved screens is a common one, that would allow the display to conform to the field curvature of the lens - at least on one axis, depending on how the flexible display bends. A patent expired some months ago, about placing a fused fiber optics faceplate onto a regular flat screen, “Piping” the image up to a curved surface - this would easily allow it to be curved on both axes. Progressive lenses, to flatten their field curvature, etc…

Anyway: G’day… or g’night, for me. :7


#669

There’s been some debate recently as to the cause of the IPD measurement discrepancy issue on this forum. @risa2000 has been nice enough to provide an explanation as to those discrepancies in these two threads: here (Some thoughts on the IPD discrepancy ) and here (Clarifying Near IPD x Distant IPD confusion .

In my view Risa’s explanation does not adequately serve to explain the cause of the IPD measurement discrepancy. It serves as a useful tool to understand how cantered lenses work that’s for sure. It doesn’t explain why theres a 10mm less software ipd reading when by geometry and divergence it should result in additional measurement added to software ipd reading to compensate for the reduction in the space between the centre of the lenses which a divergent display would cause over a 180 degree flat planed VR lens.

Risas own diagrams illustrate this if you look at the fact the flat planed diagram he shows has a greater IPD range than the divergent diagram IPD that he shows.

Risa’s explanation relies on the centre of the lenses not being the focal points. My explanation conversely relies on the centre of the lenses being the focal points for the eyes.

For deliberation between both points of view you have to ask:
Why even have the smallest circle fresnel rings indicating the centre of the lenses if that’s not actually where the sweet spots are ??

Much more likely everyones looking at the inner edges of the sweet spots which fall within 65mm IPD since 70mm is the true minimum, and the diameter of the sweet spot is 2.5mm-5mm in each direction from the middle of the smallest fresnel rings.

There’s no assumptions being made, the data speaks for itself. Theres a discrepancy of 10mm extra measurement between the physical lens distances and the on screen IPD setting, on screen IPD being 10mm less than physical distance.
Mathematics and geometry are saying it should be a smaller lens distance created due to divergence of panels and lenses, compared to a 180 degree plane and that should be accounted for with extra measurement on the software IPD reading to give an accurate reading, not a reduced software IPD reading.

The middle point of a line on a 45 degree angle is shorter than the middle point of a line on a 180 degree angle, which is why extra software IPD measurement should be added to compensate for the divergent displays to give an accurate 180 flat planed vr IPD reading (which corresponds to real IPD), not by reducing the on-screen IPD reading. This is my explanation for the IPD measurement discrepancy being 10mm less on screen than the 70mm physical minimum distance between the lenses on Pimax.

I believe the measurement discrepancy is caused by Pimax opting to reduce IPD to increase lens real estate as much as possible, and why the minimum distance between the centre of each lenses is 70mm, whereas most other 100 fov consumer vr headsets have a 60mm minimum IPD and IPD values which correspond to the users IPD measurements.

If anyone’s willing to chime in to explain how do Risa’s posts account for a lesser on screen IPD reading when it should in fact have a greater on screen IPD reading to compensate for actual IPD measurements since divergent displays result in a lesser IPD than a 180 degree flat-planed vr lens, seeing as they’re the only two explanations that make sense.

Does Pimax want to chip in and offer a solution for customers and backers that purchased the product with one of the main assumptions informing their purchase was that reviewers and product information said 55mm or 60mm IPD was the minimum required to use the headsets?

Some can’t use at 65mm, what about them? Even after extensive fitting adjustments.
As far as I see it everyone under 70mm IPD - 55mm IPD require a solution and address by pimax and until a solution is provided customers in those IPD ranges should be entitled to a refund who can’t use the headsst for even short periods or at all due to intense eye strain and discomfort.

I propose that the IPD issues of discrepancy and being unable to get a clear image in both eyes, and eyestrain issues could be solved by shaving off 5mm off the inner parts of each lenses towards the nose and creating a rubber adapter for the smaller 160/190 fov lenses, which sits at the outer peripheral part of the lenses, which will allow for compression of the adapter when IPD is adjusted outwards to the maximum value and an accurate on screen reading in comparison to real IPD measurements.

@Sean.Huang @Dallas.Hao @mozi @PimaxUSA An official response would be appreciated as to whether or not my abovementioned suggestion for a smaller lens design will be considered and implemented as a purchasing option and a replacement lens on store, as a smaller lens of -5mm off the inner side towards the nose and an outer periphary rubber compressible adapter would entirely solve the three problems.

    1. IPD measurement discrepancy issues
    1. Eye strain related issues.
    1. Customers within the 55-65mm IPD ranges would be able to get clarity in both eyes at the same time.

I’ve brought both of these diagrams to approximate scale so my explanation behind the discrepancy can be easier illustrated.

Risa’s suggested focal points, which I suspect to be towards the outer edges of the sweet spots of both lenses.


#670

Just one note, stereo overlap is not defined by the lenses position, but by the (views) geometry at which the game renders the views for the left and the right eye. How is the view geometry defined by Pimax for its headset (5k+ at least as I do not have 8k) I have explained here (All the different FOVs of Pimax 5k+).

In other words, bringing the lenses closer to each other does not change the (angular) overlap.

When defining the views projections, Pimax uses the similar approach @jojon describes above for Oculus, i.e. the views for the left and the right eye are not symmetric around vertical axis, but extend more outwards than inwards (at the cost of having smaller stereo overlap). The exact figures (angles) are in my post I linked above.


#671

imho there is also another thing
when you dial in a different ipd then the one the HMD osd gives, you will look in a different way through the lens then it was done in factory setup and the distortion profile might not fit anymore and whatever you do with the foam it never will in the way it was meant to be


#672

Not disagreeing that anyone who buys something wearable based on advertised parameters and finds that the fit is not right for them will be disappointed. The thing is that marketing such a product with a only one or two variables (out of many) advertised is totally within acceptable practice. For example, in order to not complicate the buying process too much, it is assumed that a specific shoe advertised as size 9 1/2 will fit a large percentage of the population, but certainly not all who might go by size 9 1/2. This is because there are other parameters that can affect fit: width, amount of arch in the sole, etc.

As a consumer, your expectations should be that if the fit isn’t right, you can return or refund and not be left hanging by having paid for a product you can’t use/enjoy. As long as Pimax is willing to do this wrt problems with eyestrain or discomfort, you’ve go nothing to complain about. I believe they’ve stated that they will treat their customers right in this respect.

Of course future iterations of their product line could introduce design changes to accommodate a wider range of people, but whining that “it doesn’t fit right and I expected it to because it said size X and I’m size X” is an unreal expectation (IMO) for a product such as this which has so many parameters which need to be considered for things to be “perfect”. Yes, it could be closer to ideal, but so could your physiognomy.


#673

You need C. Being looking straight you would be looking through the same inside ring when looking straight ahead. Ie ring number 2. When looking to the left your left eye looks on an angle through the center of the lens while the right eye will say might be looking through ring number 4.

This would be the case for most ipds requiring a more complex distortion calculation & will have trade offs. The greater thr Asymetry between eyes the more refinement needed. Ie soft ipd adjustments to correct absolute pupil position vs equal eye symetry.


#674

That’s exactly what I found. I’ve set my headset to 62.3 which seems “best” and my (self-measured) IPD is 63.

I agree. I think there is something more than just IPD which is causing the discomfort that some people are experiencing. Luckily, I’m able to use my 8K for hours at a time without any discomfort.


#675

I wonder if a software ipd changer could promote the wearer to have it closer there face ?


#676

I have done a lot more testing, also with one or two more users. My impression stands. If I have my IPD set at around 65 which is my natural IPD that is where edge distortions are smaller and is also where my eyestrain goes away. If I try to go right down to the minimum and try to force the lenses to 59ish then I am just nearly but not quite able to get both lenses to their ‘sweet spot’ but it is f#$@ing awful, distortions all over the edges and I develop a headache within an hour as well as feeling cockeyed for the rest of the day.

I would guess that if my ipd was lower, like say 59, the issue there would be that converging the view would be more toward the inside of the lenses and so might be a bit blurrier but my advice would be for people having eyestrain issues to try setting their natural IPD and then play with positioning to try to get both eyes centered as much as possible and also noting tilt angle and vertical position. Spend a few hours with it and see if you still have eyestrain (try to ignore the fact that it is not super sharp). Would be interesting to see if that is just in my case or if it also applies to others. I can do a 5 hour simracing session now with zero eyestrain or ‘something is off’ feeling, and I would still not trade this headset for anything else.

The thing this headset excels at for simracing is sense of speed. Maxing out on a long straight in a 60’s era open wheeler with this FOV gives a sense of speed unmatched by anything else and for me the tradeoffs are totally worth it.


#677

Dont get what you are saying, is that not figer B, draw on a bit of paper and take a photo and upload it if you cant use windows paint.


#678

Now this is interesting as I don’t have problems with distortions now no matter what the IPD is set to, maybe a tiny bit on the outer edges. I started out with IPD at max and worked my way down untill the image is sharp as I can get it, could still do with maybe 1-2 below the minimum of 60 on my 5k+ but distortions don’t get any better or worse with any IPD for me.


#679

How does one ascertain whether those with IPD issues have either a faulty headset or whether it’s just a variance in people’s perception It seems to me if all these people don’t have issues, then something might be wrong with the headset. At least two people returned their headsets, so hopefully Pimax will confirm this either way.


#680

After some research on the web :slight_smile:

looking for more data, pm me if you got, I will update

Average IPD in the world

Global Average

  • American / Indian : 65.1 mm
  • Asian/Pacific Islander : 63.2 mm
  • black : 65.6 mm
  • Hispanic : 63.5 mm
  • white : 62 mm
  • other : 63.3 mm
    Survey from around 4’000 people (Source)

#681

That meshes with my experience too. Trying to get crystal-clear vision straight ahead seems to cause distortions. It appears that the Pimax engineers “balanced” a number of factors and that setting your headset IPD to your “real” IPD is correct, even though some visible areas will not be precisely in focus, because you aren’t looking through the center of the lens. The Pimax is a unique design, never attempted before, which leads to some complicated issues.

Moving the existing lenses closer would not appear to fix this “IPD” issue. It would simply shift the discussion to another complaint.


#682

I have to recognize i have a big problem now with my left eye and very fatigued. I thought i had set up everything right after hours of experimenting, positioning facepads ipd etc., but now that i actually play longer sessions of one hour+ i get severe eyestrain in my left eye.
Not long ago an optician mentioned my left iris is about 3 mm further from my nosecenter then the right one. and also a few mm higher.
With the origenal facepad of 11mm i get severe distortions, if i make it thicker with extra padding i could almost eliminate it. But i see the middle edgeborder of the lens on the left eye wich breaks immersion. I can adjust my pupils that i don’t see it but that is very tiring.
Also thickening the facepad on one side doesn’t help. i think i have tried everything
But the worst is the eyestrain. I’m even considering selling it while i was so enthousiastic in the beginning.
Curious to know your opinions!
Can this assymetrical thing be fixed with eyetracking in the future?
Does anyone see the inside of the lensborder when looking straight ahead?
Maybe also pimax engineers can say something about this, anyone has tagnames? @PimaxVR @PimaxUSA


#683

Mine is left 2 mm further from my nose. Right now I ended up 3D printing a facial interface mod that a person posted on reddit. Which is helping, I also shift the hmd to my left a little (I have more room around my nose to shift with the facial interface mod installed)

This mod comes in one piece or 2 pieces. I printed out the one piece version, but some one else said they were going to try just the lower piece to see if they can get a better vertical position


#684

oke thanks i have seen it, i have a printer,and i noticed it’s for a narrow face, getting thicker on the sides, while it has to be tinner on my face i think. But i can give it a try.
I did try that to shift the headset a bit to the left. It worked on my lenovo but with this one it doesn’t help me


#685

This is why we need an ipd software offset perhaps per eye to correct position rendering for pupils.