Digital vs Print — How to Submit an ID Photo

Upload portal or printed photo? Pixel sizes, file limits, JPEG vs PNG, printing on a 4×6 sheet at 300 dpi, and why photographing a print never works.

Published 2026-09-27

Two different objects with the same face

A “photo” for an ID submission is one of two different objects, and mixing them up is where most failed submissions start:

  • A digital file — judged in pixels and bytes. Badge systems, student-ID portals, visa applications, professional profiles.
  • A physical print — judged in millimetres on paper. Some schools, gyms, government counters and legacy processes still want one.

The same source shot can produce both — the maker outputs exactly that pair — but the rules for each are different enough to deserve their own checklist.

Digital submission: pixels and bytes

Upload validators read three things, in this order:

  1. Pixel dimensions. The portal either states them (“square photo, minimum 350×350 px”) or implies them with a shape (“ID photo”). The safe defaults when nothing is stated: 600×600 px square, or 413×531 px for a 35×45 mm card shape — that’s 300 dpi at the physical size.
  2. File size. Byte windows are common: LinkedIn caps at 8 MB, visa portals often enforce both a floor and a ceiling. A normal JPEG from this tool sits comfortably inside almost every window — the failures come from 20 MB phone-camera originals at one end and over-compressed messengers at the other.
  3. Format. JPEG is the universal expectation. Submit it unless the portal names another type; if you hand PNG to a system that converts silently to JPEG, you’ve surrendered the compression quality to their defaults.

One myth to drop: the dpi field in your file means nothing to an upload validator. DPI is print metadata — a 72 dpi tag and a 300 dpi tag on the same pixels produce identical files. What matters is the pixel grid, which is why the maker gives you exact dimensions rather than dpi settings.

Physical prints are ruled by physical measurement, and the failure mode is scaling, not pixels:

  • Print at true 300 dpi, 100% scale. The 4×6 sheet from the maker tiles your photo at exact physical size — eight 35×45 mm prints on one sheet — with dashed cutting guides. Print it “actual size”, never “fit to page”; a 2% shrink makes a 35 mm photo 34.3 mm, and picky issuers measure.
  • Beware kiosk “enhancements”. Lab software that auto-corrects colour, crops to fill, or sharpens will quietly break the size or the background. Turn off auto-everything; you want a dumb print of a correct file.
  • Verify with a ruler. Measure one photo off the sheet before cutting the rest — 35×45 mm means exactly 35×45 mm. If it’s off, the print ran at the wrong scale; fix the print settings rather than the file.
  • Paper choice is minor but real. Matte or lustre photo paper reads better under badge lamination than high gloss (glare, fingerprints), and some counters reject glossy prints that reflect under their lights.

The trap to avoid entirely

Never re-photograph a print to satisfy a digital ask. Phone-snapping a paper photo injects four defects at once: moiré from the paper’s halftone dots, specular glare off the surface, keystone skew from the hand-held angle, and a second generation of compression. Automated validators are specifically trained to catch it. If you only have a print, a flatbed scan is the tolerable path — but shooting a fresh digital photo against a wall is easier and better.

Decision summary

Asked for a file? → exact pixels, JPEG, inside the byte window. Asked for a print? → 4×6 sheet at 300 dpi, actual size, matte paper, ruler-check. Asked for nothing specific? → the digital conventions cover you almost everywhere, and the requirements guide shows which formats are standards and which are convention. When the channel allows both, submit digital and keep the sheet as backup — portals don’t smudge.

Frequently asked questions

What pixel size should my digital ID photo be?

Whatever the portal states — and if it states nothing, the safe conventions are 600×600 px for a square upload and ~413×531 px for an ID-card shape (35×45 mm at 300 dpi). Bigger is not automatically better: some portals set *maximum* dimensions as well as minimums, so follow the numbers on the upload page when they're given. Our presets output exact sizes for exactly this reason.

JPEG or PNG — which should I submit?

JPEG unless the portal explicitly says otherwise. It's the expected format virtually everywhere, its compression suits photos, and the file lands in whatever byte window the portal set. PNG is lossless but produces files several times larger — useful only when a portal asks for it or when you're archiving a master copy. If you submit PNG to a portal that silently converts to JPEG, you lose control of the compression quality.

Does the dpi number in my file matter for uploads?

No — this is the most persistent myth in photo submissions. DPI is a metadata hint for printing, not a property of the image; upload validators read pixel dimensions, not the dpi field. A 600×600 px file tagged 72 dpi and one tagged 300 dpi are pixel-identical. DPI becomes real only when the photo hits paper: there, 300 dpi at final size is the print standard, which is why our sheet output renders at true 300 dpi.

How do I print ID photos so they come out the right size?

Use the 4×6 sheet download — your photo tiled at true 300 dpi with cutting guides — and print it at 100% scale, never 'fit to page'. Kiosks and lab software that auto-crop or 'enhance' will break the physical size. After printing, measure one photo with a ruler before cutting the rest: a 35×45 mm photo should measure exactly that. Matte or lustre paper handles ID photos better than high gloss (glare, fingerprints).

Can I just photograph a printed photo for a digital submission?

Don't — it's the most common self-inflicted rejection. Re-photographing a print adds moiré from the paper texture, glare from the surface, skew from the camera angle and a second generation of compression. Portals and validators flag all four. Always go back to the original digital file; if the original is lost, a flatbed scan beats a phone photo every time — and shooting a fresh photo against a wall beats both.