BriskFile

Guides ·

How to reduce image file size without losing quality

Three methods, ranked by how many bytes they save and what they cost you in quality. The one that saves the most is the one most people skip.

Resize first, compress second, change format third. Almost everyone reaches for compression first, which is the weakest of the three and the one that costs visible quality soonest.

The three methods, ranked

MethodTypical savingQuality costWhen to use it
Reduce the dimensions75–95%None, if the image is still larger than it will be displayedThe image is bigger than anywhere it will be shown
Re-encode at lower quality50–80%None up to about quality 80; visible below 70Dimensions are already right
Change format25–50%None — it is a different encoder, not lower qualityYou control where the file is going
Strip metadata1–5%NoneEvery time; it is free

Why resizing wins

File size scales with the number of pixels, and pixel count scales with the square of the dimensions. Halving the width and height leaves a quarter of the pixels — before any compression at all.

A 12-megapixel phone photo is 4000 pixels wide. Emailed to someone who views it on a laptop, it is displayed at perhaps 1000 pixels. Ninety-four percent of the data was never going to be seen. No amount of clever compression competes with simply not sending it.

The rule: find out how large the image will actually be displayed, and resize to that. For email, 1600px on the long edge is generous. For a website, match the layout width. For a document, 1000px is usually more than enough.

Why “without losing quality” is mostly achievable

Below about quality 80, JPEG artefacts start to be visible on close inspection. Above about 90, you are roughly doubling the file for a difference nobody can see without pixel-peeping.

Quality 82 is where most photographs stop looking different from the original. That single setting typically removes 60–80% of a phone photo with no visible change — because phones save at a conservative default rather than an efficient one.

The exception is images that have already been compressed. Re-encoding a JPEG stacks artefacts on top of the ones already there. If you have the original, start from that.

Format changes, honestly

Changing format is the only one of the three that reduces size without discarding anything the previous format kept — you are re-encoding the same picture with a better encoder.

FromToTypical savingCatch
JPEGWebP25–35%Fine for web and modern software; some older desktop apps still refuse it
PNG (photo)JPEG80–90%Lossy, and permanent
PNG (graphic)WebP lossless20–30%None worth mentioning
WebPAVIF20–30% againSlower to encode; check your target browsers
HEICJPEGLarger, not smallerYou do this for compatibility, not size

If a form is rejecting your file

Do not guess at quality settings. Use a size target: state the limit and let the tool find the highest quality that fits under it. Guessing means either overshooting the limit or landing far below it and throwing away quality you did not need to lose.

If the file will not fit even at low quality, the dimensions are the problem, not the compression. Reduce those and try again.

Common questions

Does compressing an image reduce its resolution? No. Compression changes how the pixels are stored, not how many there are. Resizing is what changes resolution — they are different operations, often confused because both make the file smaller.

Can I get quality back by compressing less later? No. Lossy compression is one-way. Keep the original if you might need it.

Does stripping EXIF data help? Barely for size — usually a few kilobytes. It is worth doing for privacy, since EXIF often carries GPS coordinates and a device serial number.

Why is my compressed file bigger than the original? Usually re-encoding something already efficiently compressed, or saving a photograph as PNG. If the tool cannot beat the original it should hand it back unchanged.