Compress image to 20KB

Twenty kilobytes is among the tightest limits you will meet, and it is almost always attached to a small image: a signature strip, a thumbnail portrait, an identity photo in a fixed box on a form.

At that size the pixel dimensions matter more than the quality slider. A signature scanned at 800 × 200 pixels reaches 20 KB comfortably and still reads cleanly. The same 20 KB spread across a 4000-pixel-wide photograph gives roughly one byte for every two hundred pixels, which no encoder can make look good.

So if a form asks for 20 KB, check whether it also states dimensions. It usually does, and matching them first makes the size limit easy.

Encoders do not take a file size as input — they take a quality setting, and the size falls out of it. Getting to a specific number therefore means searching.

The tool encodes your image at a mid-range quality, measures the result, and then moves the quality up or down depending on whether it came in under or over your limit. It repeats that, narrowing the range each time, until it finds the highest quality that still fits. That is a binary search, and it converges in about eight passes.

If nothing fits even at minimum quality, and you have allowed the dimensions to change, it scales the image down by a fifth and searches again. If you have locked the dimensions, it stops and tells you the target could not be met at that pixel size — rather than quietly handing you a file that will be rejected.

Size caps on government portals, exam boards and job applications are almost always about storage and bandwidth across millions of submissions, not about image quality. A form that accepts a hundred thousand applications at 5 MB each is storing half a terabyte of photographs.

The limits are usually enforced strictly and unhelpfully: the upload is rejected with a generic error, and rarely tells you by how much you missed. That is why it is worth targeting a little under the stated cap — 95 KB for a 100 KB limit — so that any difference in how the server measures the file cannot push you over.

The quality number in a JPEG or WebP encoder is not a percentage of anything. It is an index into a set of quantisation tables that control how aggressively detail is discarded. The relationship to file size is steep and non-linear.

For a typical 12-megapixel phone photograph the curve looks roughly like this:

• Quality 95–100 — near-original appearance, but the file is often only 20–35% smaller. Rarely worth it for the web.
• Quality 80–90 — the sweet spot for photography you care about. Typically 50–70% smaller with no artefact you would notice at normal viewing distance.
• Quality 70–80 — the sweet spot for ordinary web images. Usually 70–85% smaller. Artefacts appear only if you zoom in on flat sky or skin.
• Quality 50–70 — visible softening and blocking in smooth areas, though thumbnails and background images survive it well.
• Below 50 — obvious blocking, ringing and colour banding. Useful only for placeholders.

Diminishing returns set in fast at the top of the range. Going from quality 100 to quality 90 usually halves the file for no perceptible change; going from 90 to 80 saves far less and costs slightly more. If you only remember one number, use 80.