Here’s how I went about this remix.
- Starting in Pixlr’s advanced option, which looks similar to Gimp or Photoshop, I first uploaded three images (Bella, Edward, and the painting).
- I used the Wand tool to cut around the heads of the figures in the painting, deleting them and creating an empty space. In retrospect I probably didn’t need to do this, since the new heads would have covered the old.
- Then I cut the head from the image of Edward, clicked on the painting, created a new layer, and copied his head onto the painting.
- I realized at this point that I didn’t have the sizes correlated for each image, because Edward’s head was tiny and I couldn’t figure out how to make it bigger. So I deleted the layer I’d just created, went back to Edward’s initial image, enlarged it until his head was proportional to the gent in Wood’s painting, and went through the process again.
- I did the same process with Bella. Through this process I saved at multiple points with different filenames so I could backtrack easily if necessary.
- At this point I used the pointer tool to move the heads into position. Edward’s looked pretty good, but Bella’s was more problematic because of the lighting. I probably should have stopped and found another photo at this point… but I didn’t.
- I “flattened” the image, essentially merging all the layers into one. The I began to finesse the image to make it look less fake.
- I used the “magic brush” tool highlight areas that needed color, and then used the “clone” tool to copy a nearby section and reproduce it. This took some practice and backtracking before I got the hang of it.
- I tried adjusting Bella’s skin tone and highlighting, but after a long time messing with it, I still didn’t see it getting much better. At this point I saved again, and then reopened the image in Pixlr’s “efficient” option.
- I used the Touch-up and Airbrush tools to keep working on Bella. Then I played around with Color, Color Splash, Contrast, and the Effects option. A more sepia toned look seemed best.
Overall, I was fairly pleased with how it turned out, with the takeaway that next time I’ll spend more consideration on matching the original images for lighting.