Reducing photo file size without visible quality loss
Start from the largest source, choose JPG for photos, and re-encode once. These careful steps shrink files while keeping quality viewers will not notice.
What actually makes an image file large
Four things decide the size of an image file: the number of pixels, the format, the compression quality, and any metadata stored alongside the image. Understanding them in this order is what makes shrinking a file predictable instead of guesswork.
Pixel count matters most. Doubling the width and height of an image quadruples the number of pixels, and the file typically grows by roughly the same factor. Compression quality comes second, and format comes third — but format can change the size by several times on the wrong kind of image.
Start from the best source
File size starts at the source. If you have a high-resolution original, convert from that master instead of repeatedly re-saving a copy that has already been compressed.
Every lossy re-save discards a little more detail and adds a little more noise, so a file that has been compressed several times can look worse and be no smaller than a clean conversion from the original.
Right-size the pixel dimensions first
This is the single most effective step, and it is often overlooked. A 3000-pixel-wide photo uploaded to a site that shows it at 800 pixels is mostly wasted data. Resizing to the display size typically cuts the file to a fraction of its original size with no visible loss at all.
Work out the largest width the image will actually be displayed at, then resize to that width. For a photo shared by email, 1600 pixels wide is usually generous; for a full-width website hero, 1920 pixels is a common target.
JPG for photos, PNG for graphics
For photographs and images with smooth color, JPG gives the smallest file at a given quality. PNG keeps every pixel but costs several times the size, with no visible benefit for most photos because the camera already introduced far more detail than the eye can resolve.
For screenshots, logos, or images with text and flat color, PNG keeps the edges sharp and is often the right trade-off. Converting those to JPG saves space but wraps text in visible noise.
Choosing a JPG quality setting
JPG quality is a slider, and the useful range is narrower than most people expect. Around 80 to 85 percent is the sweet spot for photographs: file size drops sharply compared to 100 percent, while the difference is essentially invisible in normal viewing.
Below about 60 percent, artifacts become noticeable as blocky patches in smooth areas such as skies and skin, and as ringing around edges. Above about 90 percent, the file grows quickly for improvements almost no one can see.
One careful conversion beats many
Convert once to the format you need and check the result at full size. If the file is still too large, reduce the pixel dimensions again rather than compressing the same file harder and harder.
Repeatedly converting between lossy formats compounds the damage. Each pass adds another layer of artifacts on top of the previous one, so quality falls faster than size drops.
Strip metadata you do not need
Photographs often carry embedded metadata: camera settings, timestamps, and sometimes GPS coordinates. This data can add a surprising amount to the file size, and location data is a privacy concern when sharing images publicly.
Removing metadata reduces the file a little and removes personal information at the same time, which makes it a good habit for anything shared online.
Consider WEBP for web use
If the image is headed for a website you control, WEBP usually beats JPG outright: at matching visual quality it is typically 25 to 35 percent smaller, and it can also carry transparency. That means faster pages and less bandwidth for the same appearance.
Keep JPG when the destination needs maximum compatibility, such as email attachments or uploads to systems that may not accept newer formats.
Verify the result before deleting the original
Open the converted file at 100 percent zoom and compare it with the original side by side. Check smooth areas for blockiness, edges for halos, and text for softness. Zoom out to the real display size as well, because that is how the image will normally be seen.
Once you are satisfied, archive the original and use the smaller file as your working copy. Keeping the master means you can always produce a different size or format later without starting over.