Last week, Adam Savage of Mythbusters fame shared a series of portraits on Twitter, and his followers jumped at the opportunity to insert his face into famous pictures, ranging from movie scenes to posters and everything in between. That’s exactly the sort of challenge I relish, so I had to join in!
Luckily, the portraits are decent source images for editing. They’re well lit, with clear separation between subject and background, and they were all provided in reasonably high-resolution. A flat-colored background would have greatly sped up masking, but these work just fine for a non-professional project.
Picture #1: Zardam
My own personal taste veers in the direction of science-fiction, and I wanted to use films and scenes that other people weren’t using. When that’s the case, the first thing that always comes to my mind is the 1974 film Zardoz. If you haven’t seen Zardoz, it’s a nearly incomprehensible fable about a savage warrior (played by Sean Connery in a mankini) who clashes with a group of immortals overcome with ennui. Part of the plot revolves around a flying stone head, the eponymous Zardoz, who commands the savages as their god and instructs them to murder without mercy. Weird, right?
Let’s see some pictures!
The first image is an original frame from the film (sourced online), and it’s followed by my edited version. The process here was fairly simple and easy to accomplish. Since I was only using Adam’s face, I cut and pasted his entire head onto a new layer, then masked out the unwanted parts using a soft-edged brush to smoothly blend it with the underlying image. I then applied a Curves adjustment layer to get the proper exposure, followed by a Gradient Map adjustment layer using colors sampled from the original, which converted his skin-tones to the proper shades. Next came a noise layer set to Overlay to replicate the film grain, and a clouds layer set to Soft Light to break up the lighting. The final touch in the composite was a little painting with a very translucent brush to match the haze at the top of the head.
I made the logo in Illustrator, based on the film’s own. It’s not terribly well composed (that M is sticking out a like a sore thumb), but it suited the purpose well enough.
Picture #2: Adam the Merciless
For my second picture, I continued the sci-fi theme with 1980’s Flash Gordon. You’ve all seen Flash, right? That’s one film that surely needs no introduction. Rather than swapping Adam in for the film’s protagonist, I decided to blend him with the villain, Ming the Merciless as portrayed by the amazing Max von Sydow.
This presented something more of a challenge because it’s dealing with a realistic subject; something like a stone head gives you a lot of space for fudging and imprecision that simply doesn’t exist when transposing faces between real people. Of course, everyone has an app in their pocket that does roughly the same thing, but without the same level of fidelity and nuance.
The work on this composite was very similar to the Zardam picture, only with several added layers of complexity. I started the same way, by pasting Adam’s face into a new layer, then used a mask to blend the two images. I also repeated the steps of adding Curves and Gradient Map adjustment layers to correct the exposure and coloration. The coloration required some added steps, however, because the white lighting of Adam’s portrait was a poor match for the multicolored lighting on the set. In this case, I added a yellow fill layer set to Soft Light to change the overall warmth, and another in light-blue set to Multiply that cooled the highlights. In retrospect, I probably could have made both changes through Curves, but that’s a lesson learned for next time. After that, there’s also a single Multiply layer where I painted his beard a desaturated brown to blend it with Ming’s beard.
The biggest single challenge is a very small and almost unnoticeable detail, but it was a giant pain in the butt. You see, Adam is wearing glasses, and the glass refraction alters the outline of his face. To blend correctly, I had to paint a new temple, as well as part of the collar that’s visible through the lens. That accounts for maybe 2% of the image’s real estate, but it took up roughly 30% of my time.
Finally, I didn’t care much for the look of the subtitles in the original image, so I painted them out with some cloning, then replaced them with new text.
The overall effect is fairly convincing, though I could have done a better job of matching the lighting. In the original. The original’s mid-tones are redder and the highlights are bluer, but these are differences you’d only really notice when the images are side-by-side.
Picture #3: Forbidden Adam
The last image in this trio comes from 1956’s Forbidden Planet, a genuine classic based on Shakespeare’s The Tempest, but set instead in the depths of space. A team led by Leslie Nielsen is sent to check up on a scientist investigating alien ruins, and drama ensues. It’s a fantastic movie with some of the best matte paintings in the history of film, and if you haven’t seen it yet, fix that post-haste.
I’ll be frank: this isn’t my best work, but I’m including it here because the failures can be instructive. The first mistake was photo selection, both Adam’s portrait and the source image. The original has a very high-contrast, almost painterly style, with a strong key light up and to the left, while the portrait I chose features a key up and to the right. Worse, the original is almost entirely highlight, while Adam here is predominantly mid-tones. A mismatch like this will always make a composite jump out in an unflattering way.
It would have been smarter to use the same portrait from the Adam the Merciless picture, mirrored so its lighting roughly matched, but I was intent on using one of Adam’s goofier expressions. This mistake also forced me to a create new shadow under his chin, and the high contrast there is drawing undue attention. Last, I was sloppy in cutting his hair out of the source photo, and that’s really just down to patience (or lack thereof).
The sum of all these mistakes is a final image that’s fun but doesn’t withstand any level of scrutiny. If this were anything other than a Saturday morning exercise, the work would be unacceptable and I’d have started over from scratch. But instead, it’s here helping teach a few lessons in photo compositing.
So, there you have it — three quick composites and all the gritty behind-the-scenes details. I think the results are pretty entertaining, and these sorts of exercises are great for keeping your skills sharp. No matter how simple a project seems, there’s going to be some small unexpected challenge that forces you to rethink your process and come up with an answer.
I’m planning a step-by-step tutorial for this type of composite in the coming weeks, but I’ll be posting quick and dirty explanations like this more often.
Do you find posts like this useful? Do you have any tips that would improve on what I did? Just want to vent about the headache Photoshop is giving you? Leave a comment and let me know what you think!
Until next time…