Image Quality Checker — What's Wrong With This File?
Upload a single image. DiffALL measures focus, noise, JPEG block structure and clipping, tells you which one is the weakest, and says what to do about it.
Free, no sign-up — your files never leave the comparison.
Drop an image here or click to browse
JPG · PNG · WebP · BMP · TIFF — up to 20 MB
Free, no sign-up — your image is deleted as soon as the check finishes.
Reading the image…
What DiffALL measures
Focus
Edge energy measured per tile and reported from the sharpest region — so a portrait with a deliberately blurred background is judged on the face, not the bokeh.
JPEG Compression
Block structure measured across the 8-pixel grid against the detail between blocks. Catches a file saved at low quality, or re-saved several times.
Noise
Noise level in grey levels, estimated on the flattest parts of the frame so texture is never counted as grain.
Clipping
How much of the image is blown to pure white or crushed to pure black — detail that is gone and cannot be brought back.
No overall score
You get a band per dimension and the weakest one named. A single number would average away the one thing that is actually wrong.
It says when it cannot tell
A plain gradient gets no focus verdict; a flat-fill screenshot gets no compression verdict. Silence beats a confident wrong answer.
Supported formats
Use cases
Before you publish
Check an export for blown highlights or heavy compression before it goes on a site or into print.
Is this the original?
Heavy block structure is the signature of a file that has been saved and re-saved rather than exported once from the source.
Triage a batch
Find out quickly which shots are soft or noisy enough to be worth re-taking.
Settle an argument
"It looks fine to me" becomes a measurement: this much clipping, this much noise, this much blocking.
Frequently asked questions
How do you check the quality of an image?
Upload one image. DiffALL measures four things — how sharply focused the sharpest part of the frame is, how much noise sits in the flat areas, how much JPEG block structure is present, and how many pixels are clipped to pure white or black. Each gets a band from Excellent to Poor, and the weakest one is named at the top.
Why is there no overall score out of 100?
Because it would hide the thing you need to know. An image that is perfectly exposed, noise-free and completely out of focus is not 75% good — it is unusable, and an average would say 75%. You get the weakest dimension named instead, plus every dimension listed.
Can it tell me if my image was upscaled?
No, and it will not guess. Two methods were built and measured for this, and both confused film grain and night scenes with real detail. Rather than report a confident wrong answer, the check is not offered. It is the honest limit of measuring one file with no original to compare it against.
Does it work on screenshots and graphics?
Partly, and it says so. The compression check relies on photographic detail between the block edges, which a flat-fill graphic does not have, so on a screenshot that dimension is left out rather than reporting a clean bill of health it cannot support.
Is my image stored on your servers?
No. The file is deleted as soon as the check finishes. Nothing is kept and nothing is inspected.
What is the difference between this and comparing two images?
This needs one file and asks "what is wrong with it". The comparison tool needs two and asks "what changed between them" — SSIM, PSNR and a pixel heatmap. If you have the original as well, the comparison tells you exactly what an export or an edit cost you.
Ready to compare images?
Free, no install, no sign-up required for your first comparison.
Start comparing now ↑