How to Compress Images Without Visible Quality Loss

You can usually cut an image's file size by a large amount without anyone noticing, by doing three things in order: reduce the pixel dimensions to what you really need, choose a format that suits the content, and then lower the quality setting only as far as it looks identical. Strictly speaking, lossy compression always discards something, so "no quality loss" really means "no loss you can see". This guide shows how to get there.

Why image files are big in the first place

An image is a grid of pixels, and each pixel stores colour. A 4000 x 3000 photo has 12 million pixels. Stored raw at three bytes per pixel, that is about 36 MB. Compression is what turns that into the few megabytes your camera or phone actually saves.

Two things drive the final size: how many pixels there are, and how much work the compressor is allowed to do on each one. You can influence both, and the first is usually the bigger win.

Step 1: Resize to the size you will actually use

Most oversized images are oversized because of pixel count, not because of poor compression. If a picture will appear 800 pixels wide on a web page, uploading a 4000 pixel wide original wastes about 25 times the pixels (4000 / 800 = 5 in each direction, and 5 x 5 = 25).

Work out your target before anything else:

• A profile photo for a form might need only 400 x 400 pixels.
• A blog image rarely needs more than the width of the content column, often around 1200 to 1600 pixels to look sharp on high-density screens.
• A print at a given size needs enough pixels for its quality target, which is covered in our guide to resolution and DPI.

Use the image resizer for this, keeping the aspect ratio locked so nothing looks stretched. Shrinking pixels this way throws away detail nobody would see at the display size anyway.

Step 2: Choose the right format

Format choice can matter as much as quality settings. As a rule of thumb, photographs compress best with a lossy format such as JPG or lossy WebP, while flat-colour graphics, screenshots and text compress best with a lossless format such as PNG.

Saving a photograph as PNG is a common cause of heavy files. Saving a screenshot as JPG is a common cause of fuzzy text. Matching the format to the content fixes both problems before you touch any quality slider. If you are unsure which to pick, our comparison of PNG, JPG and WebP walks through it.

Step 3: Tune the quality setting by eye

For lossy formats, the quality setting trades size against fidelity. The relationship is not a straight line. Dropping from the very highest setting to a slightly lower one often shrinks the file dramatically with no visible change, while dropping from a moderate setting to a low one produces smaller savings and obvious damage.

A sensible method:

1. Start with a high setting and note the file size.
2. Lower it in steps, comparing the result with the original at 100 percent zoom.
3. Stop at the lowest setting where you cannot spot a difference in smooth areas such as skies and skin, and around text or thin lines.
4. Keep that setting for similar images.

The right number is different for every image, because detailed, noisy pictures hide compression better than smooth gradients do. Do not rely on a single "magic" value.

An image compressor lets you try this quickly and compare sizes before and after.

What to look for when judging quality

Zoomed-out previews hide problems. Check these areas at full size:

• Gradients: look for visible bands or steps in a sky or a studio backdrop.
• Edges: look for fuzzy halos or specks around lettering and sharp outlines.
• Flat colour areas: look for blocky patches, especially in dark shadows.
• Faces: skin should still look smooth, not waxy or blotchy.

If you see any of these, raise the quality a notch or switch format.

Metadata and other hidden weight

A photo file carries more than pixels. It can include camera details, a thumbnail, colour profile data and sometimes location information. Removing unneeded metadata trims the file a little, and it has a privacy benefit if you are about to share a photo publicly. The saving is small compared with resizing, so treat it as a bonus, not the main method.

A worked example

Suppose you have a 6000 x 4000 pixel photo that is 8 MB, and you want it on a web page at 1200 pixels wide.

1. Resize to 1200 x 800. The pixel count drops from 24 million to 960,000, which is 25 times fewer, so the file will fall to a small fraction of its old size before any quality change.
2. Keep it as JPG because it is a photo.
3. Lower the quality in steps until you first notice a difference, then go back one step.

The exact final size depends on the picture, but the order matters: resizing first usually delivers most of the saving, and the quality tweak refines it.

Common mistakes

• Compressing a file that is already compressed many times. Each lossy save adds more damage, so always work from the original.
• Resizing up. Enlarging a small image does not add detail and often increases file size.
• Chasing the smallest possible number. A file that is 5 percent smaller but visibly worse is a bad trade.
• Using JPG for logos and screenshots with text.
• Forgetting upload limits. Check the maximum size and accepted formats before you start.

Quick checklist

• Decide the display or print size first.
• Resize to that size, with the aspect ratio locked.
• Pick JPG or WebP for photos and PNG for graphics.
• Lower quality step by step and inspect at 100 percent.
• Keep the original file untouched.

Put it into practice

Start with the image resizer to get the pixel dimensions right, then run the result through the image compressor. If a site only accepts one format, the image converter will switch it for you.