3
$\begingroup$

I think that VR / AR experts will know the answer to this.

At what vergence angle are "infinitely far away" objects such as star field textures drawn in head-mounted VR / AR displays ?

I believe I have read that the stars in the night sky are so far away that our human visual system cannot distinguish the degree of vergence required to look at them from "parallel vergence", where the view directions of the two eyes are parallel. That is, the stars are so far away that in this particular sense, they may as well be infinitely far away.

In my experiments with anaglyph displays on laptop screens, I have found that it is very uncomfortable to diverge my eyes to the point of parallel view directions, in order to resolve star field textures that are "infinitely distant", i.e., in which the left and right images are separated horizontally by exactly my interpupillary distance.

In case this might be due to measurement errors causing the separation on screen to very slightly exceed my actual interpupillary distance, I have experimented with reducing that separation one mm at a time, until the degree of vergence became tolerably comfortable. However, this required a separation difference that was significantly smaller than my actual interpupillary distance, tending to refute the "measurement errors" hypothesis.

Is the discomfort likely to be due more to the well-known contradiction between vergence and accommodation (focus) that is a drawback of conventional stereoscopic displays ?

(EDIT As well as the muscular discomfort, I noticed that the star field texture was blurry, as if my eyes were refusing to focus on the screen. Perhaps at this extreme of vergence, the physiological connection between vergence and accommodation forces accommodation at infinity ?)

In VR headsets, does the greater focal depth alleviate this problem, i.e. is the limit of comfort for vergence closer to parallel, in this case ?

I have read that parallel vergence should be avoided in stereoscopic displays, because it is unnatural for our visual system, but I have not read any explanation as to why (almost) parallel vergence is perfectly comfortable for viewing stars in the real world. Might it also be to do with the fact that the solid angle subtended on earth by a star in the real world is very small ?

Are some of my assumptions wrong ?

EDIT:

I moved the most important part of the question to the top, and modified the title to emphasise it.

EDIT:

I have cross-posted this quesion on the gamedev stackexchange, here: https://gamedev.stackexchange.com/q/205830/145133

Improve this question
$\endgroup$
5
  • 1
    $\begingroup$ Without more specific information on the nature of the discomfort you've experienced, it's hard to tell what could be the cause. It seems plausible to me that it is indeed related to the well-known accommodation-convergence mismatch, so perhaps try reading subjective descriptions of that effect and see if they match your experience? $\endgroup$
    – waldyrious
    Commented Apr 8, 2023 at 5:14
  • $\begingroup$ Thank you - that's a great suggestion. I'll do that. I would describe the discomfort as an ache, seeming to come from the muscles involved in vergence, but I am sceptical of my ability to identify the source precisely. $\endgroup$
    – Simon
    Commented Apr 8, 2023 at 17:48
  • $\begingroup$ For the record, I'm not an expert in the field and don't know off-hand how people describe the accommodation/vergence mismatch. That said, my lay interpretation is that that seems to be what you're describing. It seems plausible to me that the sensation of discomfort is coming from forcing an eye orientation different from what the focusing mechanisms are triggering (or vice versa). $\endgroup$
    – waldyrious
    Commented Apr 9, 2023 at 8:54
  • $\begingroup$ "the left and right images are separated horizontally by exactly my interpupillary distance": what do you mean ? For points at infinity, the interpupillary distance plays no role, the stars are seen at the same places (in the same directions) in both images. $\endgroup$
    – user1703
    Commented Apr 11, 2023 at 15:36
  • 2
    $\begingroup$ @YvesDaoust Yes, and in order to simulate that on an anaglyph display (a nearby object), the images must be physically offset by the interpupillary distance. $\endgroup$
    – Kevin Reid
    Commented Apr 11, 2023 at 15:50

0

Reset to default