Whenever I see a composited image that doesn’t seem to quite work – my own photographs included – it is always because I am distracted by the editing itself. Case in point is an image I’m working on right now, in which I am editing someone half underwater, but the water line is just too distracting – too white, too big, too in your face. It isn’t blended properly. And because of that, I notice the editing before the concept.

One way that I deal with removing distractions from my edits is to know the order that I like to edit in, and to finish that stage before I move on to the next. My first step is always the actual compositing – literally stitching together various images so that they blend well.

When I was editing this image of my friend Jen Brook in a chateau in France, I had to edit on about eight other images photographed all around her to build out the room she is standing in. Each piece required that I do a couple of things before moving on. First is to blend using a big, fuzzy (soft, 0% hardness) brush to erase where the harsh lines of the edge of the image are so that there were no mistakes where the image was blended. Next was to match up exactly where that new piece of the picture had to go on the underlying image so that everything lined up. This often requires warping (Edit–>Transform–>Warp) so that I can maneuver each piece of the new image into place. Once everything lines up, I click that layer on and off (using the eyeball symbol in the Layers palette) to make sure I’m not missing any pieces. Finally, I retouch the color or exposure of the new piece that has been added, in case the color shifted slightly or the light did.

Once I have all of my images stitched together, and only after they ALL blend perfectly, I move on to lighting. The most common pitfall of compositing is thinking that it stops at blending the separate images together. Compositing must often go beyond stitching because when we enhance light, we take the viewer’s gaze away from certain areas and put it on others in the frame. I like to drastically enhance light, thus doing things like darkening the background and highlighting my subject. In this case below, I selected only the background, feathered my selection (in this case it was about 30px), and then darkened that area using curves and pulling down from the highlights only. This allowed the highlights on the curtains to become dark while maintaining some of the mid-tones, which would include her hair, so that they were not as greatly affected. The result is a more natural fall-off of light.

Working in light and color is rarely necessary in compositing, but it can drastically help the believability of the image. For example, in the above image, by adding blue to the shadows, and by skewing those blacks to be gray, I created a more muddied look in the shadows. This allowed the image to blend together more by taking away some of the sharper details in those areas and creating a painterly look instead. By changing the color of the dress to better fit the color palette, the viewer immediately sees a greater connection between all pieces of the final image. The colors all work together, the light is motivated and draws the eye to a certain part of the frame, and the compositing has no stitching flaws.

Even though my methods may change and evolve over time, my theory on compositing remains the same. I stitch, I enhance or change light, and I enhance or change color. Those three items, in that order, follow me through every edit that I work on.

Are there any go-to methods that you use to create your images, composited or not? Share them below!