Selective depth of field

Start with the right photo.

A scene with a clear foreground focal object and more distant background.

What to expect

Keep the coffee table and its small potted plant sharply focused, and progressively soften the sofa, rear plant, armchair and back wall with natural lens depth of field.

Keep an eye on

Illustrative depth separation; fine leaf boundaries may require review and this does not recover optical depth measurements.

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.

Depth separation should follow object boundaries and distance. Fine leaves, hair and transparent objects are difficult boundaries and deserve a close review.

Review and refine the result

If blur spills over the focal subject, correct the edge locally. Avoid a uniformly blurred backdrop when near and far elements need different softness. This treatment cannot recover a real depth map.

A related direction, and why

/MACRO — Macro-style framing changes the apparent viewing scale, rather than merely softening the existing background.

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.

Keep the coffee table and its small potted plant sharply focused, and progressively soften the sofa, rear plant, armchair and back wall with natural lens depth of field. Illustrative depth separation; fine leaf boundaries may require review and this does not recover optical depth measurements.

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.