Skip to main content
VVertex Solutions
All toolsPDF ToolsImage ToolsText ToolsCalculatorsDeveloperBlog
VVertex Solutions

Browse tools for merging and compressing PDFs, resizing and converting images, formatting text and JSON, running calculators, and working with code.

Popular Tools

  • Mortgage Calculator
  • GPA Calculator
  • Tip Calculator
  • BMI Calculator
  • Percentage Calculator
  • EMI Calculator
  • Merge PDF
  • JSON Formatter

Categories

  • All tools
  • PDF Tools
  • Image Tools
  • Text Tools
  • Calculators
  • Developer Tools

Company

  • About
  • About the Author
  • Editorial Standards
  • How We Verify Tools
  • How Data Is Processed
  • Calculator Comparisons
  • Disclaimer
  • Privacy Policy
  • Cookie Policy
  • FAQ
  • Accessibility
  • Changelog
  • Terms of Service
  • Contact
  • Blog
  • RSS Feed

© 2026 Vertex Solutions. All rights reserved.

Free tools. No signup. Privacy first.

  1. Home
  2. Blog
  3. JPEG Artifacts Explained: What Those Blocks and Halos Actually Are
Imageinformational6 min readPublished 2026-04-07 · Last updated 2026-10-23

JPEG Artifacts Explained: What Those Blocks and Halos Actually Are

JPEG artifacts are predictable side effects of lossy compression. Learn to spot blocking, banding, halos, and mosquito noise, and how to avoid stacking another save before you publish.

By Rehan Baqir

Quick answer

JPEG artifacts are the visible cost of lossy compression: 8 by 8 blocks, banded skies, halos on edges, and mosquito noise. A 6000 px hero can still look pixelated if it was saved around quality 55, cropped in a CMS, and saved again around quality 60. You can soften the damage slightly. You cannot restore the discarded detail.

JPEG artifacts are block edges, banded gradients, halos, and speckles left by lossy compression. Compress Image can export a web JPEG once from a master. It cannot put back detail a previous save already discarded. A hero can look pixelated on a phone at 6000 px wide if it was saved around quality 55, cropped in a CMS, and re-exported by an "optimize images" plugin around quality 60. Three lossy passes can turn the sky into a chessboard. The limit is stacked recompression, not a lack of pixels.

Those marks are the fingerprints of the quality setting and of every later save. Once you know the shapes, you stop blaming the screen.

How JPEG compression creates artifacts

JPEG is lossy. The encoder:

  1. Converts color to frequency data (DCT) in 8 by 8 blocks
  2. Quantizes, which throws away high-frequency detail it judges unimportant
  3. Encodes what is left

When quantization is aggressive (a low quality setting), blocks do not blend smoothly. Edges ring. Gradients staircase. Fine texture turns to mush.

That is how the format works, not a bug in the export dialog.

The kinds of JPEG artifacts

Blocking (macroblocking)

Visible squares appear in areas that should be smooth: blue sky, a beige wall, a skin shadow. They are most obvious below about quality 70 on a large display.

Raise quality, switch to WebP at a similar byte size, or (a rare photographer's trick) add a little real noise in the source before export.

Color banding

Skies shift in discrete stripes. The gradient is no longer continuous. JPEG shares the blame with 8-bit color depth. Compression makes it worse.

Raise quality, apply a slight gaussian blur on the sky only before export (used in production, and argued about), or switch to WebP.

Ringing (halos)

Bright or dark outlines sit along sharp edges: a white shirt against a black jacket, or text on a screenshot saved as JPEG.

Do not save screenshots that contain text as JPEG. Use PNG or lossless WebP. For photos, a moderate quality increase is the usual fix.

Mosquito noise

Fuzzy speckles swarm high-contrast edges: hair against sky, tree branches against clouds. The name comes from the insect-like look.

Use less aggressive compression. Avoid sharpening after a heavy JPEG save.

Color bleeding

Chroma subsampling merges color channels aggressively. The common mode is 4:2:0. Red text on white picks up pink fringes, and fine colored patterns smear.

Export with 4:4:4 chroma if the tool allows it. Use PNG for graphics that include colored text.

Why artifacts stack

Every Save As JPEG re-quantizes data that was already quantized. Social platforms, CMS plugins, and email clients often recompress on upload.

Camera JPEG (Q92) → Canva export (Q80) → CMS "optimize" (Q75) → Instagram re-encode

Each hop adds damage. The workflow rule is one lossy hop from the master.

Keep a TIFF or a high-quality master. Export the web JPEG once via Compress Image. Do not re-save the web file.

Step-by-step artifact audit

  1. Open the suspect image at 100% zoom, not fit-to-window.
  2. Pan across sky, shadow, skin, and fine pattern. Note blocks and banding.
  3. Check edges for halos and mosquito noise near contrast boundaries.
  4. Trace provenance. How many times was this JPEG saved?
  5. Re-export from the best master at quality 82 to 88, then adjust after a visual check.
  6. Compare bytes. If the new file is only about 10% larger and the artifacts are gone, ship it.
  7. Disable CMS double compression if a plugin re-encodes uploads blindly.

If the master is lost and the artifacts are severe, reshoot or find another source. Denoise tools help only at the margin.

Format comparison when artifacts matter

| Content | Safer format | Why | |---------|--------------|-----| | Photo hero | WebP or JPEG Q85+ | Size vs quality balance | | Screenshot | PNG / WebP lossless | Text edges | | Logo on photo | PNG overlay, don't bake JPEG | Halos on logo edge | | Thumbnail | JPEG Q70 to 75 | Small display hides sins | | Archival | TIFF / RAW | No generation loss |

See PNG vs JPG Guide for graphics versus photos, and WebP vs JPG for modern photo delivery.

Real-world examples

A band poster with fluorescent gradients banded at Q65. Q82 fixed the banding and the file grew 40 KB. That increase was worth it.

A LinkedIn headshot gets recompressed by LinkedIn. Starting from Q95 instead of Q75 kept the skin acceptable after upload.

A blog image at 1200 px wide and Q80 showed no visible artifacts on a retina phone. The same image at Q60 showed mosquito noise in the hair.

The numbers shift per image. Your own zoom test is the spec.

Common mistakes

Saving screenshots as JPEG turns text into soup. Use PNG.

Editing in JPEG layers adds a generation every time you flatten and save. Edit a lossless working file.

A high pixel count does not hide a low quality setting. A 6000 by 4000 image at Q50 still blocks when you zoom.

Sharpening after compression amplifies ringing halos.

Ignoring resize order does the same damage in a different order. Resize from the master, then compress once. Compressing first and then resizing the damaged file keeps the damage.

Privacy note

Artifact inspection is local work. A random "JPEG fixer" website uploads the image. Use browser tools that process on the device: Compress Image, and JPG to PNG when you need a lossless hop.

Alternatives when JPEG will not behave

  • WebP lossy is often cleaner at the same file size (WebP Converter)
  • PNG fits UI, diagrams, and anything with text
  • AVIF is newer and efficient. Check the fallback plan.
  • One clean file at 280 KB beats two garbage passes at 140 KB each

Read how image compression affects quality for how to think about the quality slider.

Limitations

Once mosquito noise and 8 by 8 block boundaries are baked into a low-quality JPEG, converting the file to PNG or re-saving at 100% quality only inflates the byte size. The lost detail stays lost.

Conclusion

Spot the blocks, bands, halos, and mosquito noise at 100% zoom. Re-export once from a real master, pick PNG or lossless WebP when the image contains text or UI, and stop plugins from compressing a file you already compressed. A third-generation JPEG cannot return to the RAW. Another save only adds a fourth generation of the same damage.

Related Tools

Free browser-based tools referenced in this article.

Featured
Compress Image
Compress image with a quality and max-dimension setting, and compare the file size before download.
JPG to PNG
JPG to PNG wraps the photo in a PNG. It does not restore detail the JPEG already lost.
New
WebP Converter
WebP converter among WebP, JPG, and PNG, with output quality and detected input type.

Key takeaways

  • Common signs are 8 by 8 pixel blocks in flat areas, color banding in skies, ringing halos around sharp edges, and mosquito noise near high-contrast boundaries.
  • Denoise tools can soften artifacts slightly. They cannot recover discarded data.
  • JPEG divides the image into 8 by 8 pixel blocks and compresses each block separately.
  • Lossy WebP can show similar issues when pushed too hard. It often keeps more detail than JPEG at the same file size. Lossless WebP and PNG avoid lossy artifacts.

Frequently Asked Questions

Common questions answered to help you get the most from this tool.

Common signs include 8 by 8 pixel blocks in flat areas, color banding in skies, ringing halos around sharp edges, and mosquito noise, the speckled pattern near high-contrast boundaries. They get worse each time you re-save a JPEG.

Denoise tools can soften them slightly. They cannot recover discarded data. A higher quality on the first export, and no second compression pass, beats repair.

JPEG divides the image into 8 by 8 pixel blocks and compresses each block separately. At low quality, the block boundaries become visible, especially in smooth regions like walls and skies.

Lossy WebP can show similar issues when you push it too hard, though it often keeps more detail than JPEG at the same file size. Lossless WebP and PNG avoid lossy artifacts entirely.

There is no universal number. Start at 80 to 85 for photos, inspect skies and edges at 100% zoom, and raise quality until the artifacts disappear. A complex image may need 90 or higher. A thumbnail may tolerate 70.

About the author

Rehan Baqir is the founder & editor of Vertex Solutions. He builds browser-based tools and writes guides about PDF workflows, calculators, image tools, and developer utilities. Article pages show their publication and update dates when available. Read the editorial standards and tool-testing method for the site's stated approach to content and verification.

More about Rehan Baqir →

imagejpegjpgartifactscompressionquality
Back to all articles

On this page

  • How JPEG compression creates artifacts
  • The kinds of JPEG artifacts
  • Blocking (macroblocking)
  • Color banding
  • Ringing (halos)
  • Mosquito noise
  • Color bleeding
  • Why artifacts stack
  • Step-by-step artifact audit
  • Format comparison when artifacts matter
  • Real-world examples
  • Common mistakes
  • Privacy note
  • Alternatives when JPEG will not behave
  • Limitations
  • Conclusion

Related Articles

  • HEIC vs JPEG: What Your iPhone Photos Actually Are and How to Convert Them
  • WebP vs JPG: Which Image Format Should You Use?
  • PNG vs JPG: When to Use Each Format