In addition to my previous thoughts on adding my image into a sequence, I decided to marry that idea by having not the whole image step forwards and backwrds but just the parts where my shadow fell, so the closer part of me was to the work those pixels under that shadow would be pixels from futher down the timeline. Complicated to explain yes(i will try again with a more concrete example soon).
I decide it would be beneficial to have some video where the basic structure stayed the same so I started looking for time-lapse imagery. I came upon a sequence of a tree taken over a year, this seemed like an ideal candidate for some tests.
Explanation take 2. – the tree is set in a wintery scene, it branches are barren and snow lies on the ground. As a person walks towards the image of the tree their form appears in the image. But not as a captured video likeness as a silhouette. As the viewer moves closer there image becomes larger but silhouette is becoming more defined it holds the scene from a different season. Pushing forward until the tree is fully leaved.
Maybe the video will help.
I found this work to be quite enjoyable to interact with. It looked interesting but as it was an almost random piece of footage it was really just a bit of a gimmick. For this to work for me there should be an idea behind the footage I treat like this. But as a proof of concept it worked.