Image generation

A product still-life prompt that writes texture, light and lens on separate lines

A commercial still-life prompt for a perfume bottle. Subject, material, set, lighting, lens and composition each get their own line, so swapping in another product means editing only the first line. Written for image models such as GPT Image and Midjourney.

ModelsGPT ImageMidjourney

The prompt

Commercial product still life: a square frosted-glass perfume bottle with a gold metal cap, no text or logo anywhere on the bottle.
Material: fine frosted grain on the glass; the liquid inside is pale amber with slight visible refraction.
Set: resting on a wet black slate slab with a few scattered water droplets; plain dark grey background, nothing else in frame.
Lighting: a single softbox from behind and to the left as a rim light tracing the bottle's edges, with a white bounce card filling in gently from the right.
Lens: 100mm macro, f/4, shallow depth of field, focus on the line where the cap meets the shoulder.
Composition: product in the right third of the frame, slightly below center, generous negative space on the left.
Style: photorealistic, restrained premium advertising look, low saturation, 16:9 landscape.

The usual problem with product-shot prompts is one sentence stuffed with adjectives: “premium, textured, cinematic lighting, stunning.” The model receives a pile of moods, and every generation comes out different.

This prompt gives each line one job: subject, material, set, lighting, lens, composition and style. That pays off twice:

  • Easy to adapt. For a watch or a skincare jar you change the first two lines. For a darker mood you touch only the lighting line, so the rest holds still and your results stay comparable.
  • Easy to debug. When a render goes wrong, you can tell whether it’s the reflections or the framing and edit the matching line.

Details worth keeping

  • “No text or logo anywhere on the bottle”: image models like to invent garbled lettering on packaging, and ruling it out up front saves retries.
  • Saying where the light comes from and what it hits is far more reliable than “cinematic lighting.”
  • Naming the focus point keeps the shallow depth of field from blurring the product itself.
  • The aspect ratio sits on the last line. Midjourney users can drop it and use the --ar 16:9 parameter instead.