Need an answer? Check the FAQ — if it's not there, send a ticket and we'll reply by email.
On studio portraits with clear catchlights and well-defined shadows, expect 85-95% accuracy on the major lights (key + fill) and modifier types. Complex multi-light setups with overlapping shadows are harder — the analysis still gets you closer than guessing. Each light comes with a confidence score so you can judge for yourself.
Portraits with at least one visible eye (the catchlight is the AI's strongest signal) and clear shadow definition. Studio work, fashion editorials, beauty close-ups, and headshots are great. Wide group shots and pure landscapes work less well because the lighting cues are subtler.
Typically 8-15 seconds per portrait. Slower for very large source images (they get resized on the way in to keep things snappy).
A rotatable 3D scene of the lighting setup, lighting pattern (Rembrandt/loop/split/butterfly/etc) + ratio, per-light breakdown (role, modifier type, position, color temp), likely camera settings, step-by-step recreation instructions, and a post-processing recipe.
2 free at signup with no credit card. Paid subscribers get 50 analyses per month; the counter resets each billing cycle.
You can save the URL to revisit later, or screenshot any view. PDF export + email-to-yourself are on the roadmap — request it via the ticket form below if you'd like priority.
Each analysis costs us real money on the AI side (the vision model is expensive per call). A subscription keeps the price predictable for you and the service sustainable for us. Cancel anytime — your card stops being charged immediately and your account stays accessible read-only.
We read every message. Replies usually within 1 business day to the email you enter below.