How to Compress Photos Without Losing Quality

A photo taken with a recent smartphone weighs 4-10 MB. For an online form that accepts 500 KB at most, for a website that needs to load fast or for an email with ten attachments, that's too much. The good news: you can compress photos without losing quality you can actually see — not by magic, but because most of those megabytes hold detail the human eye can't distinguish.

In this guide you'll learn how compression works in plain words, the three levers that determine a file's size, the recommended quality values for every use (web, social media, print, archive) and the complete step-by-step procedure. At the end there's a list of the most common mistakes: avoiding them matters as much as knowing the technique.

There are two ways to make a file smaller. Lossless compression works like folding a shirt neatly into a suitcase: it takes up less space, but when you pull it out it's exactly as it was. It reorganizes the data more efficiently without throwing anything away — that's what PNG does. The limitation: on photographs it doesn't shrink much, because millions of subtly different pixels "fold" badly.

Lossy compression is more radical: it genuinely deletes information, choosing what the eye notices least. Our visual system perceives brightness differences well but small color variations poorly: the JPG and WebP algorithms exploit that weakness and discard exactly there. That's why a photo can drop from 5 MB to 400 KB and look identical: what's missing is detail you would never have seen anyway. "Without losing quality", in practice, means without losing perceptible quality — the goal of this entire guide is finding the point where the file is smallest and the difference invisible.

An image's final size depends on three factors, and it pays to know all of them because they act independently:

• Compression quality (1-100): how much the algorithm is allowed to discard. From 100 down to 85 the file often halves with no visible difference; below 70 the artifacts start showing.
• The format: at the same perceived quality, WebP weighs 25-30% less than JPG, while PNG on a photo weighs up to ten times as much. The choice of container matters as much as the compression itself.
• The pixel dimensions: the most powerful and most ignored lever. File size grows with area: a photo 4000 px on a side has four times the pixels of one at 2000 px. If the image will be viewed on a full HD screen, every pixel beyond 1920 wide is pure dead weight.

The winning strategy combines the three levers in the right order: first reduce the pixels to what's needed, then pick the right format, and finally tune the quality. Relying on the quality slider alone — as most people do — forces you down to low values that ruin the photo, for a result resizing would have delivered for free.

There's no universal quality value: it depends on where the photo will end up. These are the reference points professionals use:

One clarification about the numbers: the scale isn't linear. Between 100 and 85 the file shrinks dramatically and the eye sees no difference; between 85 and 70 the savings continue but skies and skin start smearing slightly; below 60 the blocky artifacts become obvious. When you're torn between two values, pick the lower one and check the comparison: if you can't see a difference at 100% zoom, there isn't one.

Compressing in the wrong format is wasted effort. If the photo is headed for a website, converting it to WebP before optimizing starts you off with a 25-30% head start over JPG: it's one step with the tool to convert to WebP. If instead the recipient explicitly requires JPG — print labs, business systems, older software — stick with the classic via the convert to JPG tool and compensate with a slightly higher quality.

The case to spot instantly is the photographic PNG: if someone sends you a 3 MB photo as a PNG (it happens all the time with screenshots or downloaded images), converting it to JPG or WebP at quality 85 brings it down to 300-400 KB with an invisible difference. Conversely, do not convert graphics with text and flat colors to lossy: those live better as PNG — if you need them in the lossless format, there's the tool to convert to PNG — where lossless compression is actually more efficient. For the full picture on each format's strengths and limits, see the dedicated guide to JPG vs PNG vs WebP.

Here's the workflow I apply to any photo that's too heavy:

1. Resize to the pixels you need. Open the tool to resize your photo, enable "Lock aspect ratio" and set the long side based on the destination: 1920 px for full-screen web, 1080-1200 px for social media and blogs, 800 px for email. Hit "Apply size".
2. Compress with your eyes on the result. Move to the tool to compress images: load the file, drag the "Quality" slider and watch the Original/New comparison side by side, with the estimated size updating in real time. Start at 85 and go down until you notice a difference.
3. Check at 100% and download. Inspect the critical areas — skies, gradients, sharp edges — at actual size, then hit "Download".

Everything happens locally: the images are processed by the browser on your own device and never pass through any server, so the procedure is suitable even for scanned documents and private photos. With a little practice, the whole operation takes less than a minute per photo.

To understand the relative weight of the three levers, let's follow a real photo: a smartphone shot at 4032 × 3024 px weighing 6.2 MB as a JPG at original quality. Goal: publish it on a blog.

• Quality only: recompressing at quality 80 without touching the pixels gets you to about 1.8 MB. Decent, but the file stays big because all 12 megapixels are still there.
• Resizing only: bringing it to 1920 px on the long side (quality unchanged) lands around 1.1 MB. On its own, this lever beats the previous one.
• Both together: 1920 px + quality 80 = roughly 350 KB. Seventeen times lighter than the original, indistinguishable on screen.
• With the format switch: the same image exported as WebP drops to around 250 KB.

From 6.2 MB to 250 KB with no perceptible loss: that's the potential of the full method. The numbers vary from photo to photo — images packed with fine detail compress less than ones with large uniform areas — but the proportions between the levers hold. And if the photo is going on social media, also check the ideal dimensions in the social media image sizes chart: they often coincide with the optimal resize anyway.

These are the slip-ups I see most often, in order of frequency:

• Compressing the same JPG multiple times. Every lossy save adds loss: after three or four passes the artifacts pile up. Always start from the original file and compress once.
• Throwing away the original. The compressed version is for sharing; the high-resolution master should be kept, because there's no way back from the lightweight copy.
• Dropping to quality 50-60 "to be safe". Below 70 the degradation is visible and the savings marginal: if the file is still too heavy, the problem is the pixels, not the quality.
• Upscaling a small photo. Resizing only works downwards: stretching 800 px to 1920 px invents pixels and produces blur.
• Compressing before editing. Crops, text and filters belong on the full-quality file; compression is always the last step, as explained in the guide on how to edit photos online.
• Using PNG on photos "to avoid losing quality". All you get is a huge file: for photographs, a JPG or WebP at quality 85-90 is the rational choice.