Guides/How to Reduce Image File Size Without Losing Quality
How to Reduce Image File Size Without Losing Quality
Almost everyone has hit the same wall at some point: an image that is "too big" — too big to email, too big to upload, too big to load quickly on a website. The instinct is often to assume that making it smaller means making it look worse. In practice, the two are far more separable than that — most images carry a lot of data the eye barely registers, and removing it can shrink a file dramatically while leaving it looking, for all practical purposes, exactly the same.
What actually makes an image file large
An image’s file size comes down to three things: its pixel dimensions (how many pixels wide and tall it is), its format (how those pixels are stored), and how much fine detail it contains. A photo straight from a modern phone camera is often four or five times larger, in pixels, than it needs to be for a screen, a social post, or a printed page of normal size — all those extra pixels cost file size without adding anything you can perceive at the size the image will actually be viewed.
On top of that, formats store data very differently. JPG and WebP use "lossy" compression that intelligently discards detail the eye is least sensitive to, producing dramatically smaller files at a quality that is very hard to tell apart from the original at normal viewing sizes. PNG, by contrast, is "lossless" — it keeps every pixel exactly as it was, which is essential for things like screenshots, diagrams and logos with sharp edges and text, but produces much larger files for photographic content.
Compression: the most direct route to a smaller file
For photos and photo-like images, compressing is usually the single most effective step — and the one with the best ratio of size saved to quality lost. Most compression tools offer a quality slider precisely because the right trade-off depends on what the image is for: push harder for a thumbnail or a quick chat attachment where nobody will zoom in, and ease off for something that will be printed large or examined closely.
A useful habit is to compress at the very end of your workflow, after any other edits are finished — compressing, then editing, then compressing again stacks up generational quality loss for no benefit. One clean compression pass on the final image gets you the smallest file for the least visible cost.
Resizing: shrinking the canvas, not just the data
If an image is going to be displayed at, say, 800 pixels wide, there is no benefit to it being 4000 pixels wide — those extra pixels are simply thrown away by whatever displays it, after costing you bandwidth and storage to send and store them. Resizing the image down to roughly the dimensions it will actually be shown at, before compressing, often produces a far smaller file than compression alone, because you are reducing the amount of data at the source rather than trying to squeeze an oversized image down after the fact.
The two techniques work best together: resize to the dimensions the image actually needs, then compress at a quality setting appropriate for how it will be viewed. Done in that order, you typically end up with a fraction of the original file size and a result that looks, to the eye, just as good.
Choosing the right format for the job
If an image is a photograph or anything with smooth gradients and complex color, JPG or WebP will usually give you the smallest file at an acceptable quality. If it is a screenshot, a logo, a diagram, or anything with sharp edges, flat colors or text, PNG keeps those details crisp where a lossy format would blur or smear them — though often at a noticeably larger file size, so resizing to the dimensions you actually need matters even more for these. And if you have received a photo in a format that the place you’re sending it to does not accept — HEIC from an iPhone is the most common case today — converting it to JPG or PNG first is a prerequisite before any of the above will even apply.
Tools mentioned in this guide