Low Quality Image Maker
Degrade an image the way repeated saving and reposting actually does it.
Your image is processed in this page and is never uploaded. Because the file is re-encoded here, the saved image also carries no EXIF metadata - the camera model, the timestamp and any GPS coordinates in the original are gone from the copy you download.
What this tool does
Choose an image and set how many times it should be saved, reopened and saved again. Each pass compounds the previous pass's compression artefacts, which is what produces the genuinely crunchy look. A single low-quality export only gives you smooth blocking.
Runs an image through repeated save-and-reopen cycles at low quality, so artefacts compound the way they do when a picture is reposted many times.
Loses a little resolution each pass and scales back up at the end, reproducing softness as well as blockiness.
How to use it
- Choose an image, or drag one onto the drop panel.
- Tap a preset - slightly rough through to destroyed - which sets the generations, the quality, the resolution loss and the colour boost together.
- Or set them yourself: generations is how many save-and-reopen cycles to run, and resolution loss is what makes a re-uploaded image soft as well as blocky.
- Flip between Degraded and Original to judge how far it has gone, then download or copy the result.
Why use this maker
Memes and effects rely on the look of a heavily re-shared image, and a single export does not create it. Repeated saving at low quality builds the artefacts naturally.
Why we built it
We built the low quality image maker to reproduce generation loss from repeated saving and reposting, not just one bad save. Compare results with the image comparison slider and restore size with the image compressor.
Related tools worth a look: the Normal Map Generator and the Signature Extractor.
Who this tool is for
- Meme makers and social media users
- Designers wanting a worn, retro look
- Video editors making a degraded effect
When to use it
- Making a deep-fried or "reposted" meme look.
- Creating retro low-fidelity effects.
- Testing how an image holds up under compression.
- Demonstrating generation loss.
What generation loss actually is
The reason one low-quality save never looks right.
JPEG is lossy: it discards information the eye is least likely to miss, and reconstructs an approximation when the file is opened. Save once at low quality and you get an approximation of the original - blocky, but smoothly so.
Save that approximation again and the compressor is now working on an image that already contains blocking artefacts. It treats those artefacts as real detail and produces artefacts of its own on top of them. Repeat that a dozen times and the errors reinforce each other into ringing round every edge, colour bleeding well outside its boundaries, and blocks with hard visible borders.
That compounding is what people are looking at when an image is described as having been reposted to death. It is not a stronger version of one bad save - it is a qualitatively different result, which is why a single quality slider cannot reach it however far you drag it.
This is also why archives care about it. Every time a file is transcoded on the way through a system, a little more is lost, and none of it comes back.
Why the artefacts look the way they do
JPEG works on eight-by-eight blocks of pixels. Each block is converted into a set of frequency coefficients, the high-frequency ones are thrown away first because fine detail is least noticeable, and what remains is stored coarsely. At low quality almost nothing but the block average survives, and the block edges stop lining up with their neighbours - which is the visible grid.
Sharp edges suffer worst. An edge is made of high frequencies, so discarding them leaves a ripple alongside every hard line. That is the halo around text in a badly compressed screenshot, and it is why text degrades far faster than a photograph of a landscape.
Colour goes before detail. JPEG typically stores colour information at half the resolution of brightness, because the eye is much less sensitive to it. Each generation halves it again in effect, which is why colours smear across boundaries long before the shapes become unreadable.
| Control | Effect |
|---|---|
| Generations | How many times artefacts compound. The main control. |
| Quality per save | How much is discarded each pass. Low values reach the effect in fewer passes. |
| Resolution loss | Softness. Reproduces screenshot-and-reupload cycles. |
| Saturation boost | The over-cooked colour of an image through many editors. |
It only goes one way
Nothing here can be undone. Once a compressor has discarded a frequency band, the information is gone - there is no record of what was there, only of what replaced it. Software claiming to restore a heavily compressed image is inventing plausible detail, not recovering the original.
Keep your original. This tool always works from the file you chose rather than from its own previous output, so adjusting the sliders never compounds a result you have already produced - but the file you downloaded is permanently degraded.
One counter-intuitive effect: at very low quality with dithering-like noise present, a file can occasionally grow rather than shrink, because noise is expensive to encode. The tool reports which way it went.
What you give it, and what you get back
| Item | Details |
|---|---|
| Input | One image. |
| Controls | Number of passes and quality. |
| Output | A degraded JPEG. |
Issues you may run into
| What you see | Why it happens | What to do |
|---|---|---|
| Result is too mild | Few passes or high quality. | Add passes and lower quality. |
| Image is smaller than the original | It is reduced to 1600 pixels. | This is a processing limit. |
| Other formats are expected | It produces JPEG. | Convert afterwards if needed. |
Assumptions and limitations
- Images are reduced to 1600 pixels on the longest edge before processing, because the encode-and-decode cycle runs up to 25 times.
- JPEG only. Other codecs degrade in their own characteristic ways, which are not reproduced here.
- The browser's own JPEG encoder is used, so the exact artefacts differ slightly between browsers.
- No control over chroma subsampling or quantisation tables directly.
- Transparency is lost, since JPEG has none.
- One image at a time.
Frequently asked questions
What is generation loss?
The compounding degradation from repeatedly re-encoding an already-compressed file. Each pass treats the previous pass's artefacts as real detail and adds its own on top, so errors reinforce rather than simply accumulate.
Why does one low-quality save not look right?
Because a single save gives smooth, even blocking. The ringing, colour bleed and hard block borders people associate with a heavily reposted image only appear once artefacts have compounded across many passes.
What causes the blocky squares in a JPEG?
JPEG compresses in eight-by-eight pixel blocks. At low quality little more than each block's average survives, and neighbouring blocks stop matching at their shared edges - which shows up as a visible grid.
Why do the colours smear before the detail does?
JPEG usually stores colour at lower resolution than brightness, because the eye is far less sensitive to colour detail. Each generation effectively coarsens it further, so colour spreads past its boundaries while shapes are still readable.
Can the original quality be recovered?
No. Discarded information is not recorded anywhere. Tools that claim to restore a heavily compressed image are generating plausible detail, which is not the same as recovering what was there.
Why does the file sometimes get bigger?
Noise is expensive to compress. At very low quality the artefacts themselves become high-frequency noise, and encoding that can take more bytes than encoding the smoother original did.
Is my image uploaded anywhere?
No. Every pass runs in this page on your own device. Nothing is transmitted at any point.
What does 'Detail left' mean, and can I compare the result to my original?
Detail left is the share of the original pixel count still present after the resolution-loss passes shrink the image down and stretch it back up - a low number means most of the actual resolution was thrown away before the JPEG re-encoding even started, which is a large part of the degraded, re-uploaded look. A separate Before/After toggle swaps the preview between the degraded output and the untouched source image, so you can see exactly how much the passes changed before you download anything.