Why not just set every encoder to quality 80 and compare?
Because the number means something different in each one. It is an index into that encoder's own internal tables, not a measurement of anything, and equal values produce visibly unequal pictures. This bench searches each encoder's parameter until the file it produces measures within a set distance of your original, then reports the setting it landed on as a result rather than as an input.
What does the measured figure mean?
It is multi-scale structural similarity between the produced file, decoded again, and your source — 1.0000 means the two are identical by that measure and anything lower is further away. It is computed on a 512-pixel reference so the cost does not grow with your file, and it is a statistic, not an eye.
Why is my picture sometimes larger in every format?
Because it was already encoded well. A JPEG that came out of a good encoder is close to what the format can do with those pixels, and putting it through a second encoder adds loss without buying space. When that happens the recommendation says to keep the file you already have, which is the only honest answer available.
Does anything leave this device?
No. The decode, every encode, the scoring and the download all happen in this tab. There is no server behind this site to send a picture to — it is a set of static files and a worker thread.
Why is AVIF greyed out?
Because the browser in front of you has no AVIF encoder of its own, and this site will not send a picture anywhere to borrow one. The option is tested on every visit — a browser whose canvas can write AVIF gets the column and the button. A WebAssembly encoder was measured instead and was too slow to present honestly: 3.3 seconds for one 12-megapixel pass on a fast laptop, and the quality search needs up to seven passes. The reference page has the numbers.
Can I convert a whole folder, and different files to different formats?
Yes. Drop any mix and each file gets a row in the matrix, with the size every format produced for it. Leave the target on Best per picture and each file takes the format its own numbers earn, or choose one format for all of them — and pin a different one on any single file by clicking its cell. The ZIP carries a manifest recording every decision.
What happens to transparency, metadata and animation?
Transparency is kept by PNG and WebP; JPEG cannot hold it, so a transparent file waits for you to pick a matte before a JPEG of it is written. A browser encoder writes pixels only, so by default nothing else goes in; camera details, artist, copyright and pixels per inch are added only when you switch them on, and GPS and XMP never are. Colours are converted to sRGB, and orientation is applied to the pixels first. An animated GIF gives its first, middle or last frame, as you choose; animated WebP and PNG give their first; or animated files can be left out.
My file is a screenshot. Why is JPEG marked as the wrong tool?
Flat colour and hard edges are exactly what a lossy photographic encoder handles worst: it softens the edges and rings around them, and it spends bytes doing it. The bench detects few distinct colours and a high proportion of hard edges, marks JPEG accordingly, and judges the lossless candidates instead.
What is an indexed PNG, and when should I choose one?
A PNG that stores a palette of at most 256 colours and one small number per pixel, instead of three full colour values. Logos, icons, charts and most screenshots use few colours, so they often come out at half the size or less with no visible change. Photographs have far more colours than any palette, and the measured score drops accordingly — the converter shows that number rather than hiding it. Dithering trades a little size for smoother gradients.
Can I set different formats for different kinds of file in one batch?
Yes. Batch rules give each input format its own output — HEIC to JPEG, PNG to WebP, JPEGs left exactly as they are — and a pin on a single file still outranks its rule. Another rule keeps the original whenever the converted file would not be smaller. The manifest in the ZIP lists every rule that was on.
Can it write my name, a copyright line or a print resolution into the files?
Only if you ask. Artist and copyright are written as EXIF text in JPEG, PNG and WebP; pixels per inch goes into the JPEG and PNG density fields and the EXIF resolution tags. Camera details (date taken, camera, exposure) can be copied from a JPEG source. GPS is never written. Every byte added is included in the size the page shows.
What happens to an animated GIF?
Every output format here is a still picture, so one frame is written: the first by default, or the middle or last frame, which this site reads from the GIF itself. You can also leave animated files out of the batch altogether. Animated WebP and PNG are read by the browser, which only hands over their first frame.
What are routines?
Named sets of conversion settings — format, quality bar, size, transparency, what is written into the file and the batch rules — applied in one press. Six come built in (web upload, email batch, photo archive, screenshots and logos, print hand-off, phone photos to JPEG), and you can save your own in this browser. No picture or file name is stored with them.