Low-key interior lighting

Start with the right photo.

An interior with one obvious window and recognizable furniture edges for selective illumination.

What to expect

Relight the room in a restrained low-key style with one narrow soft beam entering the existing left window, illuminating the sofa edge and coffee table while the rest falls into deep readable shadow.

Keep an eye on

A single view cannot establish measurements or hidden construction. Check furniture count and window alignment.

Make it yours, step by step.

1. Adjust the recipe to describe your subject and preferred background.
2. Copy the complete recipe, then open ChatGPT.
3. Upload your photo in ChatGPT and paste the recipe. The short code alone is not an official command.
4. Inspect identity, text and proportions. Ask for a targeted correction if needed.

Choose the change deliberately.

Decide which part of the subject deserves the limited light. Deep shadows can simplify a scene, but the main identity cue must remain visible.

Review and refine the result

If the result looks merely underexposed, request a defined key light and readable midtones on the focal area. Check dark edges against the background before adding more contrast.

A related direction, and why

/DRAMATICLIGHTING — A hard directional-light study can create strong contrast without placing most of the scene in darkness.

These are editing suggestions, not additional test observations. Read the practical guide or use the before/after checklist.

Where this effect helps — and its limits.

Relight the room in a restrained low-key style with one narrow soft beam entering the existing left window, illuminating the sofa edge and coffee table while the rest falls into deep readable shadow. A single view cannot establish measurements or hidden construction. Check furniture count and window alignment.

Input advice and use cases are editorial guidance, not additional test results.

Notes from recorded tests.

These notes describe the named runs, which may use a different subject from the comparison above. They do not guarantee results on another input or a future model. See our testing method.