ALL IMAGE TOOLS56
Image Pixelator
Pixelate an Image, or Just One Region of It
Cover a face, a plate or a code in blocks — with the region measured in the image's own pixels.
About Image Pixelator
This pixelator replaces the detail in an image with flat blocks. Drag a rectangle on the preview to cover one part of the picture — a face, a number plate, a code in a screenshot — or leave the region alone and the whole image is pixelated. The rectangle is held in the image's own pixels rather than the screen's, so the X, Y, Width and Height fields always read what the engine will actually use, however the preview happens to be scaled to fit. Block size is yours to set, and the output is PNG, JPG or WebP. One thing has to be said plainly: this is not redaction. The blocks are baked into the downloaded file, so the detail underneath really is gone — but when what was covered comes from a small, known set, such as a six-digit code or a date, someone can re-render every candidate at the same block size and compare it against your blocks. If something must stay secret, crop it out rather than pixelate it. Everything runs in this tab, and the image is never uploaded.
Questions
Can pixelation be reversed?
Not from the pixels alone — the original detail is not stored anywhere in the file. That is not the same as being safe. When the covered content comes from a small, known set — a six-digit code, a number plate, a date — someone can re-render every candidate at the same block size and compare it against your blocks, and pixelation has been undone that way. For anything that must stay secret, crop it out rather than pixelate it.
Does this pixelate the whole image or just part of it?
Either. Drag a rectangle on the preview to cover one region; with no rectangle set, the whole image is pixelated. The dashed outline marks the region on screen only — it is not drawn into the file you download.
What does block size actually mean?
It is the width of one flat square, measured in the image's own pixels. A 16 px block across a 4000 px wide photo is far less obvious than the same 16 px on a 400 px thumbnail, which is why the readout stays in the image's units instead of the preview's.
Why do the region numbers not match where I dragged on screen?
They do — they are the same rectangle expressed in the image's pixels. The preview is scaled down to fit the stage, so a 150-pixel drag across a 1200 px image shown 600 px wide is 300 image pixels. The X, Y, Width and Height fields always show the numbers the engine is given.
Which output format should I choose?
PNG keeps every block exactly as drawn and keeps transparency. WebP does both in a smaller file. JPG is smaller still but cannot carry an alpha channel, so a transparent source is painted white first, and its own compression softens the hard edges between blocks in a way PNG and WebP do not.
Is my file uploaded to a server?
No. Transmute processes everything locally in your browser using JavaScript and WebAssembly. Your files never leave your device — there is no server, no upload, no cloud processing.