Depth recession in 2d to 3d conversions

Andrew Stevens wrote:

Very interesting experiment!

I had little trouble fusing the pairs, but I noticed that I needed to tilt my head a bit to assist fusion. I wonder if this (as well as some of the difficulty fusing the pairs experienced by others) has to do with the uneven heights of portions of the pairs. For example the tallest butte in the Monument Valley pair.

To quickly check pair height alignment in Photoshop, try adding a top layer with a single horizontal line in it across the pair being worked on, as you move the line around it should "rest on" exactly the same features of both pairs. You can do this with a single row selection box as well, but it will be tied to whatever layer is selected, which I find inconvenient exactly half the time.

I'm experimenting with manipulated pairs myself, recently, and am encouraged to see others working with it. I'm emboldened to set up a gallery with a few recent experiments for critique.

Thanks!

Your comments on the general problematic topic of alignment is spot on. The deformations involved in the process quantitatively vary in different regions of the treated copy of the referent image, resulting in a variety of possible decisions relating to optimal alignments. I find it interesting that some viewers have no problems with what I consider severe, unusable alignment, while others are totally intolerant of even what others consider such mold misalignments as to be considered negatively.  Trying to find the middle ground is a real challenge!

I'd sure love to see the results of your work.

Best regards,

Fave