Image Optimizer
What Image Optimizer actually does
Image optimization is a tradeoff, not a magic score. This workbench applies one explicit recipe to a bounded batch, then keeps the source and re-encoded image side by side with exact dimensions, bytes, and the decision to keep or skip a larger result. The full batch is preflighted before outputs are published, and browser format fallback is rejected instead of being mislabeled.
How to use it
- Choose up to 20 images within the displayed byte and pixel limits, then set output format, quality, maximum dimensions, and JPEG background.
- Preflight the complete batch, re-encode locally, verify the actual output MIME and decoded dimensions, and hold every download if a file fails.
- Review source and output pixels and byte changes, then export verified images with a versioned JSON receipt.
Useful for
- Prepare a group of website images with one reproducible size and format recipe.
- Inspect whether a WebP or JPEG quality setting produces acceptable visible artifacts before launch.
- Record which files became larger, failed, or were intentionally skipped instead of hiding batch exceptions.
Limits worth knowing
- Quality values are encoder inputs, not comparable visual-quality scores across formats or browsers.
- Every kept output is re-encoded; EXIF and ancillary metadata are excluded, while color conversion is managed by the browser.
- The browser workbench is capped at 20 files, 50 MB input, 32 megapixels per file, 120 megapixels per batch, and 100 MB output.
Questions people ask
Are images uploaded?
No. Decode, resize, encode, review, ZIP packaging, and receipt generation happen in this browser tab.
Why can an optimized file be larger?
Format, image content, dimensions, and metadata policy all affect byte size. You can keep larger re-encoded files or record them as explicit skips.
Does the tool preserve metadata?
No. Kept browser outputs are re-encoded and exclude EXIF and ancillary metadata. Use the local CLI when you need its explicit ICC-preservation behavior.