Why does my lampshade change material when AI turns the lamp on?

Identify what the reference actually proves

A shade viewed by reflected light can look very different when lit from inside. AI may remove a fabric weave, fill a perforated shade or turn opaque metal into translucent plastic. Exposure further complicates the comparison: a bright shade can lose detail simply because the photograph clips highlights. Resolve that source limitation before assuming the real material becomes smooth when illuminated.

Prepare the evidence before generating

1. Record the lamp variant, shade and actual installed light source. Take an unlit view that clearly explains material and construction.
2. With the camera position understood, capture the same lamp illuminated. Adjust exposure as needed to preserve visible shade detail, recording that this is a photographic setting change.
3. Add a detail of the seam, weave or perforation in both states where useful. Keep different bulb configurations clearly labeled.
4. Generate one modest room scene. Compare the material landmarks in the lit result with the actual paired references before approving the surrounding light.

Example creative request and a review matrix

Illustrative request: Place this lamp in a simple room scene in its referenced lit state. Preserve the shade’s real material, seam, weave or openings from the paired photographs. Keep the stand and shade geometry unchanged. Do not turn fabric into frosted plastic, make opaque metal translucent or erase construction under an exaggerated glow.

Make the release decision

Room brightness, color temperature and beam performance need their own verified evidence. A generated interior is not a photometric measurement. Use genuine lit and unlit gallery views when the material transition matters to buying. If the desired glow cannot coexist with readable product detail, reduce the dramatic lighting or choose a genuine photograph rather than approving a different shade.

Download this case’s manual capture card

The CSV below contains this question’s capture steps and symptom-specific review checks. Record your own source filenames, observations and decision in the blank columns. It is a manual worksheet, with no store connection or automatic approval.

Download the capture and review card (CSV)

Reference scope: the listed sources provide relevant photography, media, design or accessibility context. The case-specific checklist is Itemlume editorial guidance, not a vendor guarantee, production standard or report of a completed test.