How to Shrink a Photo for an Online Form Without It Looking Terrible

There’s a particular kind of frustration reserved for online forms with strict file size limits — the kind that reject your photo with no explanation beyond “file too large,” forcing you to guess your way down to whatever invisible threshold they’ve set. Job applications, visa forms, exam registration portals, and university admissions systems are the worst offenders, often capping uploads at 20KB, 50KB, or 100KB, sizes that are a tiny fraction of what a modern phone camera produces by default.

Getting under that limit is rarely the hard part — most photo tools will let you crank quality down until the file shrinks enough. The hard part is doing it without the photo turning into a blocky, unrecognizable mess, which is a real risk once you’re asking for a 95%+ reduction from the original.

Start with the right photo, not just any photo

Before you touch compression settings at all, the source photo you choose has more influence on the final result than almost anything you do afterward. A few things make a real difference:

Crop tightly. A passport-style photo cropped to just the head and shoulders compresses dramatically better than the same photo with a full room visible in the background, simply because there’s far less detail competing for the same limited byte budget. If the form doesn’t specify an exact framing requirement, crop in before you compress, not after.

Choose a plain background. Compression artefacts — blockiness, color banding, smudging — show up most aggressively in busy, high-detail, high-contrast areas. A plain wall behind a subject survives aggressive compression far better than a cluttered bookshelf or a patterned background.

Start from the original, not a re-saved copy. If you’re working from a screenshot of a photo someone sent you on WhatsApp, or a photo that’s already been through one round of heavy compression, you’re stacking fresh compression artefacts on top of existing ones. Track down the original camera file if you still have access to it — it gives you far more room to compress before things start looking damaged.

Check the lighting. Well-lit, evenly-exposed photos compress more gracefully than dark or harshly-shadowed ones, because compression artefacts are often most visible in shadow detail and underexposed areas, where there’s less original information to begin with.

Work backwards from the size you need, not forwards from a quality guess

This is the part most people get wrong, mainly because most photo editing tools don’t make it easy to do otherwise. The typical workflow — pick a quality percentage, export, check the resulting file size, adjust, repeat — is backwards. You don’t actually care what quality percentage you end up at; you care whether the file is at or under the limit, full stop. Anything over gets rejected regardless of how good it looks.

The more direct approach is to search for the highest quality setting that still produces a file under your target size, rather than guessing a percentage and hoping. This is exactly what dedicated compress-to-target tools do automatically — encode at a range of quality levels, check the resulting size each time, and converge on the best one that fits, typically within about eight attempts. It removes the guesswork entirely and guarantees you’re not leaving quality on the table by picking a setting more aggressive than you actually needed.

When quality reduction alone isn’t enough

For very strict limits — 20KB or 50KB on a real photograph — quality reduction by itself often can’t get there without the image looking genuinely broken. At that point, the more effective move is reducing the image’s pixel dimensions as well. A smaller image has less data to represent in the first place, which means it can hit an aggressive size target at a noticeably higher relative quality than the same content would need at full resolution. Since these forms almost always display the photo at a small size anyway (a thumbnail, a passport-photo-sized box), a smaller but cleaner image usually looks better than a full-resolution image compressed so hard it’s covered in artefacts.

Know which format the form actually wants

Most of these forms specifically require JPG, and for good reason: JPG’s lossy compression is the only practical way to hit an aggressive size target on a real photograph without resorting to a tiny, low-resolution image. If you’re working from a PNG (common if your photo came from a screenshot or a scan saved by default in PNG), converting to JPG first — rather than trying to compress the PNG directly — will get you a much better result, since PNG can only shrink by reducing dimensions, not by trading away quality the way JPG can.

A trick worth knowing if you’ll do this more than once

If you expect to be submitting your photo to several different forms with different size limits in the same application cycle — multiple exam boards, several job applications — it’s worth keeping one well-cropped, well-lit “master” version at a moderate size (a few hundred KB, not your multi-megabyte camera original) rather than starting fresh from scratch each time. Compressing down from that clean master to whatever each specific form demands tends to produce more consistent, predictable results than re-exporting from a different source each time.

The realistic expectation

If a form is asking for 20KB or 50KB from a modern phone photo, some visible compression is simply unavoidable — you’re asking for a 97-99% size reduction, and no tool, free or paid, can make that completely invisible. The goal isn’t a flawless result; it’s the best possible result the size limit allows, which a deliberate, working-backwards approach to compression reliably gets you closer to than guessing at a quality slider ever will.