Before I open Photoshop on any new project, I draw it. On paper, with a pencil, like it’s 1987. I sketch the light source, the shadow angles, the rough placement of every element. It sounds slow. It is slow. But every time I’ve skipped that step, I’ve paid for it later, sometimes in front of a client, sometimes just in front of my own monitor at 2am when I’m staring at a sky replacement that refuses to sit naturally in the scene.

The problem most compositors run into isn’t a technical one. It’s a perceptual one. They’re assembling images rather than building a world.

The Invisible Contract Between Light and Belief

When you look at a finished composite, your brain is running a silent audit. It’s checking whether the light on a subject’s face matches the direction of the shadows on the ground. It’s checking whether the color temperature of a sunset sky is bleeding into the foreground or just floating there, disconnected. It’s checking rim light, fill ratios, and catchlights in the eyes. It does all of this in about 200 milliseconds, and if anything fails the audit, the image reads as fake, even if the viewer can’t articulate why.

This is the core problem with most composite work I see: technically clean, perceptually broken. The edges are masked well. The resolution matches. But the light is lying.

Light in the real world is a system. Every source creates a key, a fill, and a bounce. In outdoor scenes, the sun is your key, the sky acts as a soft fill, and the ground kicks up a subtle warm bounce from below. When you pull a subject shot from a studio and drop them into an outdoor scene, you’re bringing studio light into a sunlight world. Those two systems don’t agree, and the composite shows it.

The 5:1 Rule and Why Stock Photos Work Against You

Here’s where I get specific, because vague advice doesn’t help anyone.

A natural outdoor lighting scenario on a partly cloudy day runs roughly a 5:1 ratio between key and fill. That means your key side is significantly brighter than your fill, but not dramatically so. Full midday sun on a clear day can push to 8:1 or even higher. Golden hour drops closer to 3:1 because the sky becomes its own diffused fill source.

Stock photography, by contrast, is almost always shot in controlled studio conditions or with flash-assisted setups that flatten the ratio to somewhere between 2:1 and 3:1. It looks clean and commercial. It also looks completely wrong the moment you place it under a dramatic sky.

The fix is a combination of dodge and burn work and color grading done per-layer, not globally. I work in 16-bit TIFF files at 300 DPI even for web outputs, because I need the room to push values without banding. I use a Curves adjustment layer clipped to my subject, set to Luminosity blend mode, and I manually paint shadow and highlight corrections using a Wacom Intuos Pro at around 20 percent brush opacity. This is not fast. A single subject in a complex scene can take 45 minutes of painting before I even look at color grading.

Color Temperature as a Compositing Tool

Most tutorials treat color temperature as an aesthetic choice. I treat it as structural information.

If your background plate was shot at golden hour, the color temperature sits somewhere around 3200K to 4000K. That warm orange light is bouncing off every surface in that scene, and it needs to be bouncing off your subject too. A thin warm gradient on a Solid Color layer, set to Color blend mode at 10 to 15 percent opacity, clipped to your subject layer, will begin to marry the two worlds together. Then I add a second layer in the opposite color (a cool blue-violet) and paint it only into the shadow areas, because in warm light the shadows shift cooler. This two-layer approach takes about ten minutes and immediately moves a composite from “assembled” to “lit.”

When the Reference Folder Is the Whole Workflow

I keep a reference library that would be embarrassing to count. Tens of thousands of images, organized by lighting scenario, time of day, weather condition, and surface material. Before I build anything, I find three to five real photographs that match the lighting world I’m trying to create, and I keep them open in a second monitor the entire time I’m working.

A few years back I was hired to build an album cover that centered on a figure standing on a pier at dusk, with the light reflecting across moving water. I spent six months before that project even started studying how light actually behaves on water. The way ripples create thousands of tiny angled mirrors. The way a setting sun will streak in a long horizontal line across calm water but fragment completely on choppy water. I shot my own reference, pulled from cinematography books, grabbed frames from films with strong water cinematography. When it came time to composite, I wasn’t guessing. I was translating something I actually understood.

That’s the difference between a composite that impresses other designers and one that makes regular people stop scrolling.

The Single Habit That Separates Consistent Compositors from Lucky Ones

Sketching before opening Photoshop is not a workflow quirk. It’s a forcing function that makes you solve the lighting logic before you’ve committed five hours to a direction. When you draw where your light source is, you commit. When you commit, your compositing decisions stop being arbitrary and start being answers to a specific problem.

The artists who produce consistent, believable composites aren’t the ones with the most plugins or the fastest computers. They’re the ones who understand what light is doing before they ever move a layer.