25 checkpoints to run before every render. The questions that separate a composed image from a screenshot.
Composition is not a creative mystery. It is a set of decisions made before the render button is pressed. Most renders look flat or uninteresting not because the lighting failed, but because the composition was never deliberately built. This checklist gives you 25 specific questions to answer before you render. If all 25 pass, the composition is working.
Run through every item before your final render. Not during modeling, not during lighting. Before the final. Each item is a question. If you cannot answer it confidently, that element needs work before you render.
Before anything else, the viewer must know where to look.
One element must be visually dominant. Not two, not three. One. If you cannot name the hero subject in two seconds, you do not have one.
Desaturate the render completely. The subject must still be identifiable without colour. If it disappears into the background, the lighting or tonal contrast is insufficient.
The subject should sit at or near a rule of thirds intersection, not dead centre unless centred symmetry is a deliberate compositional choice. Dead centre with no deliberate intention reads as amateur.
Rim light, tonal contrast, depth of field, or atmospheric haze must separate the subject from whatever is behind it. A subject that merges into the background does not exist as a subject.
Leading lines, light direction, converging perspective, and compositional weight should all point toward the hero. If other elements compete for attention, they need to be darkened, softened, or repositioned.
A flat image has no foreground, midground, or background. A composed image has all three.
Foreground, midground, and background must be visually distinct from each other. If the eye cannot separate these zones, the image reads as a flat illustration rather than a three-dimensional space.
A foreground element, even a simple dark shape at the edge of frame, creates the sense that the viewer is standing inside the environment rather than looking at a photograph of a miniature.
Distant objects should be lighter, less saturated, and less contrasty than near objects. This is aerial perspective. Without it, the background competes with the subject and the image loses depth.
The viewer must be able to read how large the environment is. A human figure, a recognisable object, or a scale reference embedded in the scene communicates this. Without it, the scene could be any size.
If depth of field is used, the focus point must be on the hero subject. Background blur should increase progressively with distance. Foreground blur, if present, should be stronger than background blur.
The camera is a compositional tool. Every setting is a decision.
A horizon at 50% of the frame height is the weakest compositional choice. A high horizon emphasises the ground and scale. A low horizon emphasises the sky and drama. Choose one deliberately.
Wide lenses exaggerate depth and scale. Long lenses compress space and create cinematic subject separation. The focal length chosen for a scene should serve the emotional intent, not default to a neutral 50mm.
Check every edge of the frame. Floating objects half inside the frame, tangent lines where elements barely touch the edge, and distracting bright areas at the corners all weaken the composition without being immediately obvious.
A camera at ground level makes environments feel massive and overwhelming. Eye level feels natural and grounded. High camera positions create a god's-eye view and reduce the sense of scale. The camera height is a storytelling decision.
A tilted camera without a purpose reads as an error, not a style choice. If the camera is tilted, it must communicate something: instability, tension, or disorientation. If it does not communicate anything, level the camera.
Before colour exists, the image must work in greyscale.
There should be true blacks in shadow areas and true whites or near-whites in the brightest highlights. An image that sits entirely in mid-grey values has no contrast, no drama, and no sense of light.
The eye goes to the brightest area first. That area should be the hero subject or directly adjacent to it. If the brightest area is a background element or a sky patch unrelated to the subject, re-light or darken it.
Pure black shadows with no detail at all flatten the image and destroy the sense of atmosphere. Shadow areas should be dark enough to feel like shadow but retain enough detail to feel real. The fill light controls this.
A dark subject needs a lighter background zone immediately behind it. A light subject needs a darker zone. This localised contrast is what makes the subject pop without requiring dramatic lighting changes.
Light areas attract attention. The progression of light areas through the frame should follow a path from entry point to subject. If bright areas are scattered randomly, the eye wanders without landing anywhere.
The last five questions before the render button.
Any environment longer than 10 metres of visible depth should have some form of atmosphere: fog, volumetric haze, or at minimum a subtle aerial perspective grade in the composite. Without it the scene looks like a miniature.
Empty areas of the frame are not wasted space. They provide visual rest, emphasise the subject by contrast, and communicate scale. If every area of the frame is filled with detail, nothing has emphasis.
Large, medium, and small elements should coexist in the frame. A scene with all elements at roughly the same visual size reads as flat and repetitive. Scale variety creates visual interest and communicates the complexity of the environment.
A well-composed image can be described simply: "low camera, hero in left third, dark environment, single god ray from upper right". If you cannot describe it simply, it is probably not clearly composed. Complexity in execution, simplicity in concept.
Name it. Isolated. Overwhelmed. Peaceful. Afraid. Curious. If you cannot name a specific emotion, the composition has not done its job. Technical correctness without emotional intent produces technically correct images that nobody remembers.
If all 25 items pass, render. If any item fails, fix it first. A render that fails the checklist will require a re-render after the client or audience sees it. Fix it now.
Two more free resources from the Environment Artist Starter Pack:
10 chapters. 12 render passes. 14 Photoshop composite layers. The full production pipeline from reference to final image. Maya Redshift and Unreal Engine 5. Free to read online, or $29 for full access with all render pass images.
Read Free Guide Full Access, $29Available for freelance projects, collaborations, and commissions. Based in Karachi, working globally.