How it works

Three pieces of plain engineering, all running in this tab. No upload, no server, no AI claims.

1. Exact-size rendering

Each spec in the table carries its official pixel size at 300 DPI (a 35×45 mm photo is 413×531 px; the US 2×2″ is the official 600×600 px). The editor's frame uses that exact aspect ratio, and the crown/chin guide lines are drawn from the spec's head-height range — solid lines at the middle of the range, dashed lines at its edges. Export draws the covered region of your photo onto a canvas of exactly the spec's pixels.

2. Print-sheet packing

Sheets are packed in pixel space at 300 DPI, trying both orientations of the photo and keeping whichever fits more copies; cut lines are drawn on the shared edges. That is how a 2×2″ photo fits 6 on a 4×6″ sheet (2 across × 3 down, exactly) and a 35×45 mm fits 8 (rotated). No margins are wasted, because trimmed passport photos are cut on the line, not inside a border.

3. The KB cap (binary search)

Upload portals set hard limits — 240 KB for US visas, 100 KB for China. JPEG quality is monotonic in file size, so the exporter binary-searches quality (14 steps between 0.05 and 0.95) for the highest quality that still fits your cap, at the full official dimensions. If even the lowest quality is too big — rare, and a sign of a very noisy photo — it steps the dimensions down and tells you plainly that the result suits upload, not printing.

What it never does

It never uploads your photo anywhere (open your network tab and check), never alters the background or the face, and never stores anything between visits. The privacy page has the short version.