Does your world bend? I mean, you know how the world looks to you as you glance around so the question is, does your world bend?
Evolution has resulting in three visual benefits to our species: 1. Binocular vision enables us to judge distance, 2. Wide angle of view means we can quickly scan a large area, and 3. Our brain corrects what our eyes see so our perception fits our preconceived notion of what the world ought to look like.
You aren’t aware of that last one? Brain does a pretty good job, doesn’t it.
One frustration about photographs is they show exactly what the camera saw, not necessarily what we were seeing at the time. The classic perspective issue is photographing skyscrapers with regular cameras – the resulting pictures show a building behaving like a railroad track in that the sides slope toward each other, receding to some infinite point outside the picture frame. Yet while standing there we don’t see that. Well, our eye sees it and the image of it is recorded on our retina, but our brain “corrects” the image so we perceive the sides as being straight because in our world buildings don’t behave that way. We need specialized photography equipment to make this correction that our brain does on the fly.
That’s a typical example in just about any photography book. I ran across another just this week while playing around with another form of image, the panoramic. We’ve all seen this type of expansive photograph but do we ever think about whether what we’re seeing is what was actually there?
The old fashioned way of making these images was to literally piece together different pieces of film to make one larger image. You had to work with what you had and make them all fit together. In the digital world there is software that does that for us, stitching together multiple files and blending them into a seamless wide-angle photograph. And that’s where the question comes to play.
In Photoshop you are given layout options on how you want these multiple images stitched and blended together. The two most common are called “Perspective” and “Reposition.” According to Adobe, here’s what these two options are doing:
Perspective – Creates a consistent composition by designating one of the source images (by default, the middle image) as the reference image. The other images are then transformed (repositioned, stretched or skewed as necessary) so that overlapping content across layers is matched.
Reposition – Aligns the layers and matches overlapping content, but does not transform (stretch or skew) any of the source layers.
As an example here are the same images run through Photoshop’s panoramic merge routine, each choosing a different layout option.
As you see, they result in different images so which one represents reality? I was standing there and I can’t remember what I saw. You think a straight road is the correct version (the road in front of the camping area is indeed straight) but other elements don’t fit that. The trees in the middle of the lower image were actually closer to my camera than the ones on the end so they should look as they do – larger. That fits with my expectation. But it appears that in order to have the trees in their proper relationship the road has to curve. How could my eye see nearer/larger trees and yet still see a straight road? In case the odd shape of the top images leads you to believe it’s been cropped, notice it has the same visual elements as the lower image. All the information is there, it’s just shaped differently.
I enjoy panoramics as a way to portray lots of detail. The reason I was making this image was to play with a new piece of equipment to help make better panoramics but now it’s got me confused. Which perspective is the real one?