Friday, October 14, 2011

Compositing Multiple Exposures—10/17/11

Patti Thompson Published in Tip Of The Week
Compositing Multiple Exposures—10/17/11
This Article Features Photo Zoom
I’m no expert at panoramas, HDR, or any of those other photographic pursuits that require multiple exposures to create a single high-resolution finished image. But I’m smart enough to understand how making multiple exposures with a camera locked in position on a tripod can be immensely useful for a variety of things.

For instance, if you’re working in an area where people will be walking in front of your camera while you’re shooting—and if you don’t want people walking in front of your camera while you’re shooting—you can always lock your camera down and make two exposures in order to eliminate the errant pedestrians from a final composite image. This type of thing works great. It’s what I call the subtractive use of compositing.

But there’s another approach to multiple-exposure compositing, which I call the additive approach because with it you can add multiple elements to a scene—say, 10 orange traffic cones when you only have one—by layering multiple frames together. You can go even one step further with this approach, and instead of making images that just defy practicality, you can make images that also defy logic. For instance, you can add so many orange traffic cones to a scene that it becomes surreal, especially if they’re falling in an inexplicable downpour around a nice young lady. You can defy logic and add this element of movement to an image, and do it all safely and precisely with multiple exposures. Here’s how.

First, I chose a composition and positioned the volunteer. She had a few traffic cones scattered around her feet, and we snapped a few frames before choosing the best one. Then, with the camera locked in that position on the tripod, the young woman exited the scene and the fun began. With our three available orange cones we created a scene filled with a dozen of them, and we did it by tossing cones one by one into the empty frame and snapping away.

Because the camera was locked down, all of the cones would have the same exact perspective as the original image with the young lady. And because the light didn’t move, all of the shadows and highlights would remain perfectly aligned as well—both on the cones themselves and on the background walls. (This is also the reason why you can’t just photograph one cone and duplicate it over and over. You’ll never remotely match the appropriate highlights and shadows unless the cones are actually flying through the air.)

Once I was satisfied that I had plenty of images of flying cones, I set to work importing them into the computer. I used Aperture to sort the pictures, choosing the ones I thought showed the cones in their ideal positions throughout the frame. I chose cones in positions where they would fill the frame nicely and make it look like an actual rain of hazard cones. Once I’d selected the handful of exposures to layer on top of the original, I opened them all in Photoshop and stacked them one by one above the master image containing the young woman.

I turned off the visibility of all the cone layers with the exception of the first one I wanted to work on. Then I made a rough selection around the flying cones using the polygonal lasso tool and feathered its edges. (Even though the backgrounds should line up perfectly, I still like to avoid hard edges on my blended selections.) With the edges feathered, I clicked the Add Layer Mask icon in the bottom of the layers palette. This uses the selection to determine which areas of the layer to mask out, and which it allows to show through. (Holding down the option key while you click the Add Layer Mask button will mask the unselected area.)

With the unneeded image area masked out of the first flying-cone image, I proceeded to the subsequent layers and repeated the process. As the cones got closer and closer my selections needed to be more and more accurate for masking purposes, so I zoomed in close to work with the lasso tool at the pixel level.

The biggest problem with this approach is that sometimes, no matter how hard we try, we bump the camera. Whether a little or a lot, that subtle movement between original captures can mean the layers don’t line up perfectly. To remedy this, I use my top-secret most favorite layer alignment tool—the Layer Mode option set to Difference. This way when pixels line up perfectly between layers they appear pure black in the Difference layer, which makes it an ideal approach to nudge a layer into place until all the appropriate pixels are lined up perfectly—which you know because the image turns black.

In the end I, of course, saved the layered file in case I ever wanted to go back again and make edits, and then I flattened and resized this version for the web. It may defy logic, but that’s part of the fun. The point is that this a simple and powerful technique—and those are my favorite—to have handy for many multiple exposure compositing situations.
Login to post comments
Subscribe & Save!
International residents, click here.