Image Resizer

How it works

Choose your images, decide whether you are working in percent, width or height, and enter the number. The aspect ratio is always preserved: you give one edge and the other is calculated from it, so nothing is squashed and nothing is cut off. In percent mode, 50 halves the picture and 200 doubles it. In width or height mode you give a pixel target instead — typing 1200 into width brings a 4000-pixel camera photo down to something a web page can actually use. Enlarging is where the limit sits: scaling up stretches the pixels you already have and cannot invent the detail that was never captured, so a small image blown up simply looks soft. Every result appears in the list with its new dimensions and is downloaded separately.

This tool is also known as image resizer, resize image, change image dimensions, resize photo to a specific width, make a picture smaller in pixels.

What is Pixel?

A pixel is a single dot of colour, and an image is a grid of them: 4000×3000 means four thousand dots across and three thousand down, twelve million in total. Resizing is nothing more than changing how many dots describe the same picture, which is why halving both edges leaves a quarter of the pixel count rather than half. Those dimensions also set the ceiling on how large the picture can be displayed before the individual dots start to show.

What is Aspect ratio?

The aspect ratio is the relationship between width and height — 16:9, 4:3, 1:1 and so on. Resizing leaves it untouched as long as both edges change by the same factor; force one edge on its own and faces widen, circles turn into ovals and text leans over. This tool will not let that happen: you supply one measurement and the second is computed from the original ratio. Nothing is trimmed either, so the entire frame arrives at the new scale.

Which size for which job?

For a picture that spans the full width of a web page, 1600 to 2000 pixels across is already more than most screens will ever ask for. Inside an article, 1000 to 1200 pixels is the point where the image still looks sharp but no longer dominates the page weight.

Email is the strictest case, because mail servers commonly reject a message once its attachments pass 20 to 25 MB in total. Bringing the long edge down to 800 or 1000 pixels usually solves that in one step.

• Full-width page image → 1600-2000 px wide
• Picture inside an article → 1000-1200 px wide
• Email attachment → 800-1000 px wide

Dimensions and file size are two different levers

Cutting the pixel dimensions does cut the file size, and it does so faster than people expect: halving both edges leaves a quarter of the pixels. But the two are not the same measurement, and a picture can be heavy at modest dimensions if it was saved at a high quality setting.

So if the pixel size is fine and only the weight is a problem, resizing is the indirect route. Compressing the same file addresses the weight directly and leaves the dimensions alone.

Frequently asked questions

Will the aspect ratio change?

No. You set one edge and the other is derived from it, so circles stay round and faces keep their proportions. The tool never stretches a picture to reach a target.

Does enlarging a small image improve it?

No. Scaling up spreads the pixels you already have across a larger grid; the missing detail cannot be reconstructed, so the result looks blurred. If you need more detail, go back to the original file at full size.

What format do I get back?

JPG sources are saved as JPG, and everything else is saved as PNG so transparent areas survive. The new dimensions are added to the file name, which keeps the resized copy easy to tell apart from the original.

Can I resize several images at once?

Yes. The same setting is applied to every file you choose, and each one appears in the list with its own new dimensions so you can download them one by one.

Does shrinking to less than half blur the image?

Below half size the tool changes method on its own: it switches to detail-preserving downscaling (the Lanczos method), where every pixel of the result is computed from all the neighbouring pixels of the source instead of only two of them. Thin lines survive and jagged, stepped edges do not appear. The result line tells you when this happened. Above half size the difference is invisible, so the faster path stays in place.