Eyestrain discussion / Problems and Solutions

Also with glasses for almost 40 years. But I see well in distance, only really need them for reading. In Pimax I don’t use glasses at all and all is clear enough - I can see SDE with both eyes so the resolution can’t really get better than that.

But yes, maybe for us who wear glasses IRL the brain is already trained to deal with possible imperfections.

2 Likes

I don’t know how representative that is, I have no issues at all, and in all the pimax meetups with hundreds of backers, it didn’t show much as an issue, nor reviewers noticed any problems, and you don’t see it mentioned much here.

Still I hope pimax can help in that regard, but stating that this is a massive issue for everyone is simply not true.

3 Likes

Interesting, so some one with normal 20/20 is possibly not using the sweet spot of the lens but slightly out side of it.

1 Like

Hey @risa2000 sorry to bring this subject up over and over again, but you are saying the wrong IPD causes the geometry distortion.
In your opinion, what is the best way to set up the ipd/eye to lense distance?
Should I tune the HMD IPD same as my measured IPD which is 66,5mm and then, with the different thicknesses of the facepads work the eye to lens distance to get the picture as clear as possible, or should I work the eye to lens distance first to minimize the outer edge distortion and THEN tune the HMD IPD to get a good picture regardless the real measured IPD ? :exploding_head:

2 Likes

Don’t know about the rest but in this I would say so. After all it is not numbers you are after (and those might not be calibrated properly anyway) but good and comfortable image.

I don’t know what my IPD is, I just tried all the IPD range and settled for the one where I get good & comfortable view. Also tried various head-strap settings to get good fit. I’m probably lucky though that included facepad distance works for me Ok.

2 Likes

I am wearing glasses and I have to wear them in the headset, so my options to change the facepad thickness (and comment on it) are limited. What however I can say based on the design:

  1. If you set the IPD on the headset to your real life IPD then the lenses optical axes should converge on your eyes, if:
    a) Your eyes are at the exactly same spot Pimax used when they calibrated the model,
    b) Your headset is correctly calibrated.

  2. If your eyes are not exactly at the right spot, because either a) or b) above does not hold, you can try to compensate for that by, either:
    a) Changing the facepad thickness
    b) Changing the IPD set on the headset

Increasing the facepad thickness means you are moving the point of views convergence in front of your face and is roughly equivalent (in terms of the geometric distortion) to decreasing the IPD set on the headset. The same goes in the opposite direction, if you make the facepad thinner, it is roughly equivalent to dialing in bigger IPD.

The relation can be expressed by the formula:

d_ipd = -2 * d_pad * tan(10°) =>
d_ipd = -0,35 * d_pad

where d_ipd is the difference in perceived IPD and d_pad is the difference in the facepad thickness. For example, if you add 10 mm to your facepad thickness, it would make the perceived headset IPD smaller by 3,5 mm.

Because you have two different means to reach the correct convergence point, it means you can also use them to counter each other. I.e. increasing the facepad thickness, while increasing the IPD setting on the headset keeps the perceived IPD the same, but increases the eye distance from the lenses (and vice versa). This principle can be used to fine tune the eye to lens distance, without changing the perceived IPD.

Anyway, what I would suggest is:

  1. Dial-in the real IPD and see how it goes.
  2. If not comfortable try some range around the original value, i.e. +/- 2 mm
  3. If not comfortable, try 1) and 2) with thinner padding (if possible)
  4. If not comfortable, try 1) and 2) with thicker padding

In general, closer you get the eye to the lens, lesser (geometric) distortion you get. What you suggest (trying the thinner pad first) also makes sense, it is just that with the thinner pad, you may already be too thin.

What is also confusing, Pimax shipping the facepad of different thickness, even though the thickness has direct impact on finding the right IPD. It is as if they did not know, or did not care, but neither is very encouraging.

One thing to keep in mind though, the steps above are supposed to help you primarily with the geometric distortion, but not with getting the sharp image in both eyes. The latter can only be eased by reducing the eye-lens distance, but not completely eliminated. And if you decide to eliminate it by bringing the lenses too close (and ignoring the IPD setting), you will introduce the geometric distortion, which may have some other unwanted consequences.

8 Likes

Yeah. So the IPD adjustment is basically pointless or actually the IPD reading is.
This has a design flaw written all over it. The lenses distance to the users eyes should be adjustable and adjustable FIRST before the IPD reading would make sense.
Thats atleast my understanding.

Well yeah whatever, maybe the God almighty eyetracking will be reality some day.

I would not call it pointless, it should give you a ballpark value to start with. But I agree that there are so many additional variables, which Pimax seemed to ignore, and which make the finding the right spot an adventurous task, especially when you do not know, how your face compares to the Pimax one.

Good observation. This would definitely help to make the setting process more predictable.

3 Likes

StarVR one is using eyetracking to set the distortion profile. I was hoping that the GDC interview with @PimaxUSA would have discussed using eye tracking module to do this. But nothing about eye tracking at all :frowning:

3 Likes

So you CAN get it clear in both eyes, but then you see distortion which leads me to belive the distortion profile that is set with the IPD value dialed in on the HDM is not correct and needs to be fix?

For me my Near ipd is 61, also my left eye pupil is 2mm off compared to my right eye Im hoping that if I push the HMD farther out that it will give me a better picture. (is not that bad right now, but something is off) I have this 3D printed face mod coming (https://www.reddit.com/r/Pimax/comments/b12vcw/pimax_5k8k_facial_interface_for_narrow_faces/) hoping it will help

2 Likes

Yes noticed this mod on Discord & added it to Davebobman’s Table of Content Wiki

in my case the eyestrain its much less than lenovo explorer i dont know why. The only thing that its not so good its the center vertical line that can disappear with a good headstrap a ticker foam.

1 Like

Lenovo explorer has fixed lenses with only soft ipd adjustment.

Pimax new headsets like others with it have hardware ipd. Now there is also balancing in pimax software side in balancing between them.

No, I can get clear image in both eyes but at the cost that the geometry is completely screwed up, because I have to set an IPD which is much smaller than my real IPD. If I set the correct IPD there is no major distortion, just the image is not completely clear in both eyes. Why this happens I have already explained in my other posts.

4 Likes

@anon74848233, @Sean.Huang Please answer. Why if i measure with a ruler the distance between the centers of the lenses is MINIMUM 68 mm (when displayed 60mm)? Vive has minimum 60mm and other HMD has correct distance between centers of the lenses.
With my 70 mm IPD i also must set 64.5 mm (and distance between lenses has 70 mm).
It’s VERY VERY important question. If lenses not may be less than 68 mm, than give correct software IPD adjustment and i think for different IPD adjust you should correct distortion profile.

Would it be possible to design lenses that have the sweetspot axis angled at 100° to the lens plane? If that were possible, I could imagine the periphery becoming even more blurred.

My understanding is that it’s because of the slanted screens; the distance between the centers of the lenses is not the IPD.

4 Likes

This would indicate that the correct IPD for you would be when the HMD has a clear image but the geometry (distortion) profile is wrong and dose not match the IPD (sweet spot) of the lens.

1 Like

I would really like to know this! They are so silent! They designed this hmd, I believe they already know the answer…

1 Like

No. It would not indicate any such thing.

When I talk about the clear image here I mean clear image of the panel pixels. In other words, I cannot see clearly the pixels (or the panel matrix if you want). This has nothing to do with the distortion profile. The distortion profile makes sure that the geometry rendered by the game into perspective projection is perceived the same way through the lenses (which would otherwise add a non-perspective warping).

In fact, for being able (or not) to see the pixels, I do not care what is displayed on the panel. But I thought you already knew that (https://community.openmr.ai/t/pimax-is-there-a-plan-to-fix-eye-strain-and-ipd-issues-to-make-the-5k-and-8k-headsets-more-useable-for-people-with-less-than-60mm-ipd-small-ipd-160-fov-lens-replacements-and-adaptors/16925/312).

My point is that the image gets blurred because I am not looking through the lens center but through the outer ring. And I am not looking through the lens center, because the headset was designed this way.

3 Likes