Pimax IPD Image method 0.1. Don’t use “FAR IPD”, need “NEAR IPD”

Pimax 8K Series
,
1

Hello everybody,

I want to share with you my tests about the IPD in Pimax, because a need your help to create a better method to find it.

I create a image to use instead of pimax_default.jpg (\Program Files\Pimax\Runtime folder), as start screen, and adjust IPD. But actually is a difficult method… sorry, with your help I want to make it easy.

What I discovered:
-When you change the IPD, you barely change the distant distance focus. Then you can try in Pimax VR home to stare to a star… relax your eyes, like your are seeing a real star, and find if you see it double. Then you can change the IPD to test, but the range is huge.
-Is almost impossible to achive a one over the other sweet spot. Are huge and distant, and it doesn’t have the same measure that IPD, like others HMDs.
-I tried the method of adjust first my left eye (closing right eye), moving the Pimax in parallel over my face until I see the text sharp, and then close this eye and open right, and use IPD to adjust. But I found that with this method I need less IPD… almost 55mm or less. Usually I use 69mm in Vive.

Then I found two methods to adjust IPD, because I found that with this HUGE LENS the IPD must be something more wide.

1- First of all. If you look to infinity in SteamVR home, with large FOV, and change the IPD… I fell how the perimeter of my sight expands and contract. I assume that it’s the reason why there are people who feel the objects in the perimeter are bad located. Then I tried to find the IPD where everything is correct. But it was difficult… around 66-67mm.

2- Second, and difficult too. I use the image that I created, and tried between 66 and 67mm, staring on the red dot. I found that with 66.4 everything get relaxed, and the horizontal/vertical lines in my perimeter feel “clean”. If I use differnts IPD I feel that the lines get broken, or If I try to stare to the point, my eyes tried to relax and see everything double. Test that, to imagine a distant object.

Please, give me your impressions. I will try to create a better measure image or method, like Oculus, to find a correct IPD setting for this HUGE lens and WIDE sweet spot.

One important thing. You would think that the IPD in Pimax doesn’t matter because you see almost the same trying differents IPD. But I found, for me, that the difference between 66.4 and 66.5 is significant. 66.5 gives me eyestrain, and 66.3, but 66.4 is perfect. With this huge lens, you need to find the EXACT point. Everything else seems the same.

BONUS: I found an app in my iPhone X. “Eye Measure”. It uses FaceID camera to find my “real” IPD. And it says exactly 66.4 (like I have in Pimax HMD). But my IPD is 65mm (measured with the exactly “holes in paper” test method). I don’t know if it is a coincidence.

BONUS2: I tried again with the iPhone X app. It gives me two IPD. Far and Near. Far is the IPD I use in Vive, 69mm… and near is the IPD I have in Pimax, 66.4mm. Again… a coincidence? I’m trying to study it.

Finally. Everybody talk about that. Pimax doesn’t use “far ipd” like other HMDs. It needs “near ipd”. If you go to a optometrist, you need this two values, and use the “near” option, instead of “normal ipd”.

pimax_default
pimax_default2560×1440 1.07 MB

[Album] imgur.com
20 Likes
2

A lot of great info here already.:+1:
Just to be on the safe side, the HMD shouldn’t be in inactive mode for too long because the image can get burnt into the panels. A few hours don’t matter, but a longer period of time it could happen like to 2.5k owner reported(maybe less danger with CLPL than with OLED).
Hopefully Pimax will give us a proper calibration option in PiTool.

2 Likes
3

Awesome. Haven’t tried it but if you look in 5k 8k directory @neelrocker created s n Oculus like ipd setting.

4

Yeah! Thanks! The thing is I find IPD in Pimax works different that Oculus and Vive.

2 Likes
5

There is a recent post replacing the default.jpg for setting ipd. Within the last 12hrs that may help as well.

6

Thank you Yen for all your work and effort on this. I also have had difficulty finding a best ipd adjustment for my Pimax HMD and I spend way more time than most in it by far and what would be considered healthy, so finding a solution would be great.

Unfortunately and usually I try to never be discouraging or negative and remain supportive in someones efforts but I have to say here and I admit I could be proved wrong, but IMHO because of the pupil changing position constantly unless looking straight forward and the design of the wide FOV lenses you will never be able to reach a perfect IPD adjustment for all scenarios with this hmd using these methods.

What you are doing here will only work if you limit yourself to looking exactly straight ahead like other hmds that have narrow FOV. Just as soon as you look to your sides your ipd adjustment will be off. Now, to reduce some eye strain I tend to look forward and turn my whole head in the direction I want to look.

. So, I feel the best you can do is find a close approximation to your actual IPD measurement and to what your are using the Pimax for reading text use smalller IPD looking to your periphery use larger IPD. Thats why a balanced IPD is best IMO.

Just my two cents worth, I hope you or someone else can find a solution. However, I think a permanent solution will only be achieved in software using eye tracking.

edit: replaced foveated rendering with eye tracking. Eye tracking is what I should have originally wrote. foveated rendering doesn’t apply to this fix.
2nd edit: I should have proof read and enjoyed a bit more coffee before commenting. :pensive:

4 Likes
7

@Davebobman should add this to your troubleshooting.

1 Like
8

yes i notice that too, on the vive i use 70 ipd if i go higher or lower i notice a difference, in the pimax 5k+, it doent matter if i went from 65-70 its still look the same to me

1 Like
9

Yes! I talked about that in other post. An option to adjust IPD based in eye motion and foveteated system.

1 Like
10

Finally, I found that “near ipd” is the correct in Pimax HMDs. There are people talking about that in Reddit and all with the same conclusion.

3 Likes
11

Done
https://community.openmr.ai/t/table-of-contents-wiki/10851?u=davebobman

4 Likes