Chris Borgman is a New York-based photographer with a knack for crafting wild and whimsical images. He does this with a tremendously creative imagination and help from technology. He often photographs subjects and backgrounds separately, compositing them together in post.
"It all started out very innocently many moons ago," Borgman says. "I was outside shooting a model in a park against trees and a nice blue sky when I wished there were more clouds—nice big ones. So later in Photoshop, which I was just learning, I added a sky that I had shot another day. It was so easy; everything looked seamless. I added the sky, but didn’t get very close or exact around the trees or the model’s blowing hair because I didn’t have to since the original sky was already blue. Later, I found out that when I shot a model against a white overcast sky, it was much harder to add a nice blue sky. So when I started shooting more studio work, I used the same principle—if I wanted to add a sky, I shot on blue. If I wanted to add trees or grass, I shot on green."
From that simple beginning sprouted Borgman’s body of composited work. While his technique may be more refined these days, one thing hasn’t changed—the "green screen" approach.
The technique is called "green screen" because of its connection to the chroma green that prevails in video. (Think of your TV weatherman floating over an animated map.) But Borgman shoots his subjects against a variety of colors. It all starts with a previsualization of the finished image because a seamless blend requires knowing what color the background will eventually be.
"When chroma-keying video," Borgman says, "the use of green, and previously blue, is for the simple reason that most people never wear those exact colors, and programs like Final Cut and After Effects know exactly what they are. But a still photographer using Photoshop can use any color. I try to use a background that closely matches the image I’m dropping in. In a shoot I did for Lipton Tea, I used gray on the floor and blue on the background. I used gray because I was going to create a stone patio, and when models or props are close to the seamless paper, gray neutralizes the color spill. Although it’s what I use, chroma seamless paper isn’t ideal for true plug-in-driven masking. Fabric and paint do a much better job. Paper actually reflects colored light back onto your subject—aka color spill. Fabric and paint help to absorb the spill."
Once the studio is prepped appropriately, it’s time to light. Borgman approximates lighting type and direction to match a background image, but he doesn’t have to limit the way he lights the subject. Hard or soft light, flat or directional, the quality of light on the subject doesn’t really matter. What does matter is the separation between subject and background. Cross-contamination won’t work.
"The lighting is simple," Borgman says, "but you need more room. You have to light the background separately and evenly, no matter what light is on your subject. Shadows on your background will create difficulties for the masking program. It’s a good idea to place large pieces of foam core between your background lights and your subject, just to make sure there’s no unwanted light. And the farther away your subject is from the background, the better. It lessens the chance of color spill and further separates the two different lightings."
Aside from more studio space to create separation, Borgman says you’ll probably need more lights, too—especially if you want to rely on Photoshop plug-ins for the heavy lifting of the mask. If you’re careful, though, you can get away with more rudimentary lighting and a less-than-ideal background.
"If you’re low on lights," he says, "and you have only white seamless and are lighting the model pretty flat, you can get away with using white seamless. Just don’t light it, and make sure your background is one to two stops under your subject. Overexposing the background leads to color spill, and that could contaminate your subject’s light and color balance."
Borgman can’t photograph the subject effectively without first considering the background, of course. That doesn’t always mean he’ll be shooting it himself.
"In many of my jobs, I use stock," he explains. "I don’t always have the budget to fly to China and shoot inside an ancient monastery. So in these situations, I light my subject according to the light in my stock photo. For one fashion editorial, I was shooting models in full evening dresses doing ballet jumps. I had researched sky images online and came across a tornado-chasing photographer. He had some amazing images of all kinds of eerie skies, so I purchased some for the editorial. As I made my background choices, I would study the lighting and map out my lighting placement. It was actually simple—I shot very flat in front and just moved around the accent lights to continue a similar direction as in the background picture."
"On other jobs," Borgman continues, "I shoot everything. When I do, I have more control, but there’s more to go wrong, too. I have to match light, perspective and angle. I should—but I don’t—get very technical here. I kinda eyeball it. I remember the lens and, more or less, the angle.
I’ve never tried to do a composite image too perfectly, especially the more complicated ones. I kinda go for a whimsical approach: Get it close, make it look fun, and then let it go. It’s a little more like illustration."
Whimsy is all well and good, but Borgman takes his postproduction seriously. The key for a successful composite comes from a seamless blend between subject and background. To that end, he starts with Photoshop’s Pen tool for manual selections around all but the trickiest edges.
Says Borgman, "These are the words I live by: The Pen tool is mightier than the sword! The sword is a metaphor for the do-everything plug-in that masks your image with one click or chop. The Pen tool is what I use 90% of the time when I do compositing. There’s no other way to get a perfect mask of smooth pixels. Selection tools have trouble creating a smooth mask, and this makes it harder when blending with the background.
"Photoshop CS5 has some pretty nice features for masking," he continues, "but I’d be lying if I said I’ve perfected them. I always try Mask Pro, Fluid Mask and Primatte along with Photoshop’s built-in tools. These plug-ins are swords and almost never give you a perfect cut. I always have to refine my mask." Adds Borgman, "With fuzzy, furry things like hair, drop your Pen. This is where the masking plug-ins can help. There’s no way to pen out every blowing hair; for this, the Pen tool is too exact. If you’ve lit your subject right and used the right color background, plug-ins offer the best masking options for hair and semitransparent objects like bubbles and smoke."
Do you want to try your hand at greenscreen techniques with a simple setup? Westcott offers an inexpensive kit to get you started. The Green Screen Digital Photo Kit includes a 5×7-foot greenscreen with wall hook, plus software that helps automate the process and includes 100 virtual backgrounds to try. Westcott also offers similar kits that include lights. Estimated Street Price: $69 for the basic kit. Contact: Westcott, www.photobasics.net.
Borgman, expert though he is, still starts every masking process with trial-and-error experimentation. He tests each of his favorite masking plug-ins on a different layer and keeps only the best result. For subtle touch-ups around tricky edges, he uses special brushes, which he finds online, with shapes that create a more natural blend. For the finishing touches, he smoothes the edges of the mask and addresses color and shadow issues.
"When it’s time to get picky about the mask," he says, "I start with a blur/choke technique, then repeat. This is done by selecting the layer mask and applying usually a 1-pixel blur. Then select levels and move the triangles to one side to choke; moving to the other side will spread. This is the old-school way; Photoshop has probably got something in the Selection Refine tool that does the same thing, but I do what I know works. I repeat the process with a 0.5-pixel blur and less movement in Levels. Don’t overdo it; you want to tighten the mask, but not sharpen it. It’s the fuzzy pixels between the subject and background that make a seamless merge. Too sharp an edge, and it looks obvious.
"To get rid of any color from the original background," he continues, "if it doesn’t match perfectly, I try two techniques. One is to use the Hue/Saturation filter to isolate the color in the hair and change that to match the new background. That probably works some places, but not everywhere. Next, I add a new layer above the model and turn it to Color mode. I then select a color of hair right next to the background color and literally brush away that original background. This is why I don’t use white—no way to brush it out, but turning blue or green to brown is easy.
"You can get a perfect mask," Borgman adds, "but if your model was shot with daylight-balanced strobes and your background was shot close to sunset, you’ve got some real tweaking to do. In that situation, I’d probably cool off the background a bit, then warm up the model until it looks convincing. Probably the most important aspect of compositing is shadows, which tie one element to another. If the shadow isn’t convincing, then all your efforts are lost. In compositing, you have to pull the shadow from your studio shot or create one from scratch. Many of my students have had a hard time wrapping their two-dimensional photography minds around a three-dimensional idea. Shadowing—that’s a whole article by itself."
Chris Borgman is a New York-based advertising and editorial photographer. He also teaches photography and advanced Photoshop techniques at the Miami Ad School in Florida. "If you want to learn advertising photography in its truest form, this is the place." Go to chrisborgman.com.