Blaming the camera for blocky squares around text, when the damage was added by the export settings.
Compression
Compression artefact
Also called jpeg artifacts, compression artifacts or blocking. Here is what it means, when it changes what you export, and what people get wrong.
CompressionRuns on your deviceUpdated 10 August 2026
The short answer
A compression artefact is visible damage introduced by lossy encoding: blocky squares, halos around hard edges, muddy colour at boundaries, or stepped gradients. It is not noise from the camera, it is the encoder discarding detail. Artefacts show up worst on text, logos and flat backgrounds, and accumulate with every resave.
Updated
The short version
Compression artefact at a glance
| Quick fact | Detail |
|---|---|
| Cause | Quantisation in lossy encoding, not camera noise |
| Blocking grid | JPEG works in 8x8 pixel blocks |
| Worst subjects | Text, logos, flat backgrounds, gradients |
| Reversible | No, only masked or partly disguised |
| Gets worse with | Each lossy resave, especially after cropping |
Slide the table sideways to see every column.
Which artefacts should you be able to name?
- Blocking: the 8x8 pixel grid JPEG works in becomes visible as square tiles, most obvious in smooth areas such as skin, sky or a plain backdrop.
- Ringing: bright and dark ripples hugging a hard edge, caused by cutting off the high-frequency coefficients that described that edge. Black text on white is the classic victim.
- Mosquito noise: a shimmering cloud of speckles around text and sharp objects, which is ringing spread across the blocks that touch the edge.
- Colour bleed: colour smearing past its boundary, produced by chroma subsampling storing colour at half resolution. Saturated reds against white bleed the most.
- Banding and posterisation: a gradient collapses into visible steps once quantisation merges neighbouring tones into a single value.
- Texture loss: fine detail such as fabric weave or hair simply flattens into a smear, which is less obvious than blocking but often more damaging to a product shot.
Why do product photos suffer more than holiday snaps?
Lossy encoders hide their errors inside busy detail, which is why a photograph of a crowded street can take heavy compression and still look acceptable. Commercial images are the opposite kind of picture.
A pure white background gives the eye a perfectly uniform reference, so even faint blocking or a slight grey cast is instantly visible. A crisp product silhouette against that background is a hard edge, which is exactly what generates ringing. Overlaid price text, a brand logo and fine printed labelling all combine hard edges with saturated colour, which triggers ringing and colour bleed together. Add a soft studio gradient behind the product and you get banding as well.
The practical consequence is that a quality setting which looks fine on your camera roll can look cheap on a listing image, so judge compression on the actual asset at the actual display size.
Can you remove artefacts after the fact?
Not properly. The information that would have described the missing detail was rounded away at encode time and there is no record of it. Deblocking filters, denoising and AI upscaling can all make artefacts less noticeable, but they do it by inventing plausible replacement detail, not by recovering the original. Applied heavily, they trade a blocky picture for a waxy one.
The reliable fix is to go back to the highest quality version you still have and export again, once, at a sensible quality. If you have to work from a damaged file, a careful upscale can soften the worst of it, which is one reason a studio like Photomop pairs its compression tool with an upscaler. Whatever you do, never fix artefacts and then resave over the same lossy file, because that adds a fresh generation of them.
What people get wrong about compression artefact
Each one is a real failure mode, not a style preference.
Sharpening an artefact-heavy image, which amplifies ringing and mosquito noise rather than hiding them.
Checking for artefacts at 25 percent zoom, where blocking and colour bleed are invisible until the image is published at full size.
Leaving chroma subsampling on for images with fine coloured text, then blaming the quality slider for the smeared edges.
Questions people ask
Compression artefact, answered
The follow-up questions people search for once they have the definition.
Why does my logo look fuzzy after saving as JPEG?
Hard edges and flat colour are the worst case for JPEG. The encoder discards the high-frequency data that defined the edge, leaving ringing halos and colour bleed. Save logos and graphics as PNG or lossless WebP instead, where edges stay exact.
Do artefacts get worse every time I open the file?
No, only when you save it again through a lossy encoder. Viewing is harmless. Saving repeatedly is what stacks damage, and it is much worse if you crop or resize between saves, because the block grid shifts and new errors land in new places.
Can AI upscaling remove JPEG artefacts?
It can reduce how visible they are. An upscaler recognises blocking and ringing as damage and reconstructs cleaner edges, but it is generating detail rather than restoring what was lost. Results are good on mild damage and unreliable on heavily compressed sources.
Which quality setting stops artefacts appearing?
For photographs, most people stop seeing artefacts somewhere in the low to mid eighties on a typical JPEG scale. Images with text, logos or flat colour need higher, or a lossless format. The number is encoder-specific, so verify by eye at full size.
Same cluster
More from the glossary
Neighbouring entries, so every term in the set is one hop from every other.
Compression ratio
Compression ratio is the size of the original image divided by the size of the compressed file, written as something like 10:1.
Colour tag
A colour tag is the small piece of metadata in an image file that declares which colour space its pixel values belong to.
Contrast
Contrast is the tonal separation between the light and dark parts of an image.
From the studio
Tools that do this
Each one runs on your own device, so the photo stays where it is.
Sources
Where the facts on this page come from. Every link opens the specification, standard or documentation the claim was read at.
Keep reading
Elsewhere on Photomop
Platform image rules
Image size, file type and background rules for 22 selling platforms, each one sourced.
Read itConvert image
JPG, PNG, WebP and AVIF, with an honest answer about HEIC and alpha.
Read itAll the tools
The whole toolbox in three families, with what each one needs before it can run.
Read itOn-device privacy
The pipeline in five steps, what leaves your device, and how to check it yourself.
Read it
Now do it to a photo
Photomop is a photo studio that runs on your own device. Resize, crop, compress, convert, batch edit, remove a background and change one all work in the browser tab you are reading this in, at full resolution and with no watermark. The on-device page shows you how to check that the photo stays put.