How To Create Perfect Garment Images
A repeatable workflow for shooting product images that work first-time on any virtual try-on model — equipment, lighting, settings, post-processing.
How do I create perfect garment images for AI try-on?
For perfect garment images, photograph the garment as a flat-lay against a plain white surface under soft natural daylight. Use a phone camera (modern phones are sufficient — no DSLR needed) held directly above the garment, parallel to the surface. Fill 70-90 percent of the frame with the garment, capture every detail (collar, cuffs, hem, buttons), and avoid wrinkles, hangers, price tags, and watermarks. Export as JPG or PNG at minimum 1024 pixels on the long side.
- Modern phone camera is sufficient — no DSLR needed
- Flat-lay on plain white background
- Soft natural daylight, no harsh shadows
- Camera directly above, parallel to garment
- Garment fills 70-90 percent of frame
- Minimum 1024 pixels on long side
Setup that scales to a catalog
For brands shooting tens or hundreds of garments, repeatability matters. Set up a fixed station: a white seamless background (3x3 ft of white poster paper works), two daylight bulbs (5500K) at 45 degrees to the surface, and a tripod arm that mounts your phone directly above. Mark the corners of the shooting zone with tape. Every garment goes in the same spot, lit the same way, shot from the same height. The result is visual consistency across the catalog that AI try-on relies on for predictable output.
Lighting that beats studio gear
Soft natural daylight from a north-facing window remains the gold standard. If you must shoot at night, two daylight-temperature (5500K) bulbs in white softboxes at 45-degree angles eliminate shadows. Avoid mixed lighting (daylight + tungsten) which produces color casts.
Post-processing checklist
After shooting: crop tightly to the garment plus 2-3 percent of background on each side. Adjust exposure if the white is dingy. Remove any visible dust or lint with a clone tool. Export as JPG at 90 percent quality or PNG. Avoid heavy filters — they introduce artifacts that the AI segmentation network reads as part of the garment.
Frequently asked questions
Do I need a colour-calibrated monitor?+
Can I batch process garment images?+
Drape Editorial is the in-house research team behind Drape Try-On. We test virtual try-on models against real garment photography weekly and publish what we learn.