Terrain and the strange things it does

[bknight]

Wow so strange how peoples perception differs, it shows an un-tethered astronaut outside the space shuttle with the word NASA above.

To see this click on it and make it larger, try looking past the image for a few seconds, once it comes into view it's hard not to see it.

I'm too right brained for this activity, but since you mentioned astronaut, I can recognize the golden(?) visor.

[gillianren]

"Just see it" is not helpful advice.  I've been told how to see them, and I've never been able to.  In this one, I can tell what it's supposed to be, but my brain won't work the trick.

[JayUtah]

No two people have eyes that are exactly the same distance apart, so it's not surprising to me that 3D effects (whether red-blue or polarized) would work better for some people than for others.  I'm sure the image offset is chosen based on the most common distance...

Actually the baseline distance varies a lot and is often much wider than human eye separation.

But you're onto something.  Human depth perception is actually a combination of many dissimilar factors, only one of which is retinal disparity (the difference in the image seen by the left and right eye).  Physiology contributes to two other factors -- convergence feedback and dynamic parallax.  For nearby objects, the degree to which your eyeballs have to turn inward to bring the object of attention into the same position in the field of view provides neuromotor feedback.  Similarly in the real world we constantly move and turn our heads almost imperceptibly.  Almost.  Because the eyeballs lie some distance from the axes of cranial rotation, this motion provides subtle relative movement of objects in the field of view.

For most depth perception we rely upon a vast memory of seen objects and details and a knowledge of their expected sizes.  Their relative size in the field of vision is ironically the strongest depth cue.  The others generally work only with objects not more than a few dozen meters away.

But since depth perception is a mixture of different phenomena that are affected differently by various artificial 3-D encoding schemes, the degree to which each is successful varies according to individualized reliance on one type of depth cue over others.

[Glom]

No two people have eyes that are exactly the same distance apart, so it's not surprising to me that 3D effects (whether red-blue or polarized) would work better for some people than for others.  I'm sure the image offset is chosen based on the most common distance...

Actually the baseline distance varies a lot and is often much wider than human eye separation.

But you're onto something.  Human depth perception is actually a combination of many dissimilar factors, only one of which is retinal disparity (the difference in the image seen by the left and right eye).  Physiology contributes to two other factors -- convergence feedback and dynamic parallax.  For nearby objects, the degree to which your eyeballs have to turn inward to bring the object of attention into the same position in the field of view provides neuromotor feedback.  Similarly in the real world we constantly move and turn our heads almost imperceptibly.  Almost.  Because the eyeballs lie some distance from the axes of cranial rotation, this motion provides subtle relative movement of objects in the field of view.

For most depth perception we rely upon a vast memory of seen objects and details and a knowledge of their expected sizes.  Their relative size in the field of vision is ironically the strongest depth cue.  The others generally work only with objects not more than a few dozen meters away.

But since depth perception is a mixture of different phenomena that are affected differently by various artificial 3-D encoding schemes, the degree to which each is successful varies according to individualized reliance on one type of depth cue over others.

This all takes me back to a year ago and the report I wrote on full flight simulators and the difficulty in replicating visual cues. I added an appendix which discussed why 3D movies suck basically due to these issues.

[jfb]

Wow so strange how peoples perception differs, it shows an un-tethered astronaut outside the space shuttle with the word NASA above.

To see this click on it and make it larger, try looking past the image for a few seconds, once it comes into view it's hard not to see it.

FWIW, I have a much easier time with the full-size image vs. the thumbnail - I'm able to lock in on the 3D effect almost immediately.  The thumbnail doesn't overlap between my eyes properly, no matter what distance I view it from, so I only get a partial 3D image of the orbiter.

The image has to be sized such that when you look "past" it, the elements of the image "overlap" between your two eyes, which is where the 3D effect comes from. 

Again, the offsets are chosen based on works for most people viewing it an assumed distance - outside of those assumptions, all bets are off.

[raven]

I wear glasses, and something I've noticed is that that this particular image goes out of focus when I try it with glasses on and it easily snaps back to normal view, while, without my glasses, it's in focus, quite a bit  past when other images in my field of view of that depth are starting to get fuzzy, and I can hold it fairly well, even noticing a parallax effect as I subtly move my head from side to side.