This is an interesting idea. I see two potential problems though:
Looking at the 10 degree mark in the virtual world does not guarantee that your eye has this angle in the real world as well. It would, if you already placed the eye into the right spot.
When you look 10 degree sideways, your pupil moves ~2mm in that direction. So if you find perfect match for this particular eye orientation and then look straight ahead, your IPD will be 4mm shorter than what you set it for.
But I guess this idea is original enough to try anyway while compensating for the potential pitfalls mentioned above.
Pimax headsets should have the hardware IPD setter calibrated, so if you dial your real IPD in, it should set the lenses position correctly. But from the reports, it looks like Pimax did not really think it through in the current design.
I guess @Kapucha read something I wrote on reddit. :7
…and I stand by it.
The problem with the pre-calculated values we see in the headset, is that they assume one already have the right amount of eye relief: Optimally one would set the lens spacing dial so that the number shown in the HMD matches one’s real IPD, and then move the HMD closer to-, or farther from one’s face, until one hit the sweet spot (…and that’s not even touching on the matter of whether the HMD sits level on the user’s face :7).
…but the p8k/5k offers, infamously, no form of lens-to-cornea distance adjustment whatsoever.
Assuming the projection is rendered and presented correctly by the HMD, with great verity, one should be on the axis of the lens, when the sharpest part of the view is in the centre, and symmetrically round, rather than elliptic in any way; …which I mean the described procedure helps with – it puts you there, even if one have too much or too little eye relief.
Preferrably, I suppose this would be tested, instead of with the prescribed background, with a… let’s say: “calibration” pattern, that makes it easier to see that the falloff in sharpness and geometrical fidelity is equal on both axes, and not getting fatter in any direction, or otherwise warping… :7
@jojon I think I understand Your method and I also think that HMD IPD display label is reporting IPD assuming user is in some distance from lenses defined by face pad and the shape of the face. I will try this method during the weekend. Thanks again for this brilliant idea.
