Resizing images in bulk without uploading them
· 5 min read
Bulk resizing is the most common image job there is, and the one where the defaults quietly cost you the most quality. Doing it in the browser removes the upload, and understanding two ideas removes most of the disappointment.
Aspect ratio is the decision, not the pixel count
If you set both a width and a height that do not match the source proportions, something has to give: the image is stretched, squashed, or cropped. Setting only one dimension and letting the other follow keeps the proportions and is almost always what you meant.
Upscaling does not recover detail
Enlarging a small image invents pixels by interpolating between the ones that exist. It can look smoother, but no detail that was not captured comes back. If the source is a 400 pixel thumbnail, a 4000 pixel version of it is a larger thumbnail. Start from the largest original you have rather than from something already reduced.
Resizing repeatedly compounds the loss. Keep the originals and resize from them each time, rather than resizing the resized copy.
Picking a target size
- For the web, width is what matters and most layouts never exceed about 1600 pixels wide.
- For printing, the physical size and the printer's resolution set the requirement, and it is much higher than a screen needs.
- For an upload limit, reducing dimensions usually saves more than raising compression, and it costs less visible quality.
Why local processing helps here specifically
Bulk means many files, and uploading many files is slow, fails halfway, and puts every one of them on someone else's disk. Processing in the tab removes the transfer entirely: the slowest part becomes your own machine, and nothing leaves it.
Where it runs out
Memory is the only real ceiling. Hundreds of full-resolution phone photos in one pass is comfortable on a laptop and can exhaust a low-end phone. If a batch fails, split it in two. That is a limit of the device rather than a quota we impose, and it is the trade for not uploading anything.
How to batch upscale images at 2x or 4x: enlarge a whole folder at 2x or 4x in your browser.
How to batch resize images to exact sizes, in your browser: presets for Instagram, YouTube and the web, and why a photo may not come out at exactly the preset size.
iLoveIMG alternative: batch edit images without uploading: what differs from an upload-based tool, and where each fits.
Frequently asked questions
Will resizing reduce quality?
Reducing dimensions discards detail by definition, but done once from the original it is usually invisible. Repeated resizing of already-resized copies is what causes visible degradation.
Can I resize without distorting the image?
Yes. Set one dimension and let the other follow the original proportions.
Does upscaling add detail?
No. It interpolates between existing pixels. Detail that was never captured cannot be recovered.