Surface texture: matte, satin, brushed metal, glass
"Premium finish" is a vibe — Midjourney, Ideogram, and ChatGPT will each interpret it differently, and the result is a guess. A prompt that names the surface (matte powder coat, satin anodized aluminum, brushed steel, polished glass) hands the model a reproducible spec for how light falls on the object. The falloff and micro-highlight of cohesive light on a textured surface is the product-still signature — the cue that tells a viewer the object is a real manufactured thing, not a render.
Write the surface first, then the modifier, then the direction. Matte reads under a soft overhead box because the highlight rolls evenly; satin needs a strip box or a strip with a 30° grid to throw a defined specular; brushed metal wants a hard raking light to amplify the grain; polished glass wants the softbox to disappear entirely so the read is the edge-light and the refracted highlight, not the source itself. Midjourney, Ideogram, and ChatGPT all reproduce the texture read when surface + modifier are named together rather than left at "premium."
Light source direction: side, modifier, ratio
A product prompt that says "studio lighting" hands the model a default it will average across thousands of catalog shots. A prompt that names the direction (camera-left, camera-right, above, behind, below), the modifier (strip box with a 30° grid, soft snoot, bounce card, bare-bulb, beauty dish), and the ratio (key-to-fill 4:1, 2:1, or 1:1) gives the model a lighting setup it can place behind the subject.
Direction comes first because it sets the angle of incidence on the surface. Then the modifier, because the modifier shapes the falloff. Then the ratio, because the ratio is the difference between a soft catalog shot and an editorial hero shot. Side-light at 4:1 reads as a deliberate product pose; soft-light at 1:1 reads as e-commerce. Midjourney, Ideogram, and ChatGPT all reproduce the direction + modifier + ratio when they're named in that order, not when the prompt stops at "moody product."
Background: explicit fill, distance, negative space
"Clean background" reads to the model as white. A prompt that names the fill (white seamless, painted paper, gradient wash, black void, brushed surface, concrete tabletop), the distance from the subject (subject sits 12 inches out from a 6-foot sweep), and where the negative space lives (above-the-rim, side, wraparound seamless) gives the model the actual product-stage layout.
The negative space is the brief, not the empty pixels. A porcelain mug photographed bottom-right of frame with overhead negative space reads as editorial overhead. The same mug photographed dead center on the same sweep reads as catalog. The fill stays the same; the difference is the framing and where the negative space sits. Midjourney, Ideogram, and ChatGPT reproduce the negative-space read when the prompt names "subject positioned bottom-right, negative space overhead" — not when the prompt stops at "product on white."
Lens compression: focal length, aperture, distance
Focal length is the second biggest product-photography lever after the modifier. 35mm at f/8 is the wide catalog framing — front-to-back resolved, the product sits in the room. 50mm at f/5.6 is the standard product framing — natural perspective, slight background softening. 85mm at f/4 to f/11 is the editorial framing — subject compressed against the backdrop, depth controlled. 100mm macro is the texture framing — surface grain fills the frame, focus stacked. 135mm telephoto is the tight compression framing — the bottle reads against a soft gradient wash.
Aperture controls how much of that framing survives depth. f/4 dissolves the backdrop. f/8 holds the product sharp while softening the background. f/11 to f/16 renders every specular edge sharp — chosen for stop-and-stare chrome or cut-glass. Distance is the third lever: it changes the perspective and the size of the negative space around the subject. A good product prompt names all three.
When the brief moves past product toward the human element — editorial framing at 85mm, aperture as the depth lever, distance as the proportion lever — step into the portrait prompts guide for the aperture + focal length deep dive.