Resizing an image is one of the most common tasks in digital work, and one of the most misunderstood. Resizing incorrectly can make images blurry, distorted, or far larger than they need to be. Here is how to do it properly for every common use case.
Pixels, Resolution, and What They Mean
A digital image is made up of pixels: tiny squares of colour arranged in a grid. The dimensions of that grid, for example 1920 by 1080, describe how many pixels wide and tall the image is. A 1920 by 1080 image contains about 2 million pixels total.
Resolution in pixels per inch (PPI or DPI) describes how densely those pixels are packed when printed. A 300 DPI image at 4 inches wide contains 1200 pixels across that 4-inch width. The same 1200 pixels printed at 72 DPI would be about 16 inches wide.
On a screen, DPI/PPI mostly does not matter. What matters is the pixel dimensions. A 1920 by 1080 image displayed on a 1920 by 1080 monitor will fill the screen exactly regardless of what DPI value is stored in the file metadata. For print, DPI is critical.
Making an Image Smaller (Downscaling)
Reducing an image's dimensions is called downscaling, and it is a lossless operation in terms of image quality: you are simply keeping fewer pixels from the original. The image will look identical to the original, just smaller.
Downscaling is used for:
- Reducing file size before uploading to a website or sending by email
- Fitting an image within a specific display dimension
- Creating thumbnails or preview images
- Reducing the resolution of a photo taken at too high a resolution for its intended use
Most image resizing tools use one of several resampling algorithms to decide which pixels to keep and how to average adjacent pixels. Bicubic and Lanczos resampling produce the sharpest results for downscaling. Nearest-neighbour resampling preserves hard edges, which is useful for pixel art but produces jagged results for photographs.
Making an Image Larger (Upscaling)
Increasing an image's dimensions is called upscaling, and it is where things get complicated. When you make an image larger, you are asking the software to invent pixels that did not exist in the original. There is no way to recover detail that was not captured, so upscaling always results in some quality loss.
Traditional upscaling algorithms (bicubic, bilinear) produce blurry results when scaling up significantly. AI-based upscaling tools use machine learning to predict what the additional pixels should look like based on patterns learned from thousands of images. The results are dramatically better than traditional methods, often producing usable results even when doubling or quadrupling an image's dimensions.
If you need to upscale an image significantly for print, consider using an AI upscaling tool rather than a standard resize tool. The difference in output quality is substantial.
Maintaining Aspect Ratio
The aspect ratio is the relationship between an image's width and height. A 1920 by 1080 image has a 16:9 aspect ratio. A 1000 by 1000 image has a 1:1 (square) aspect ratio.
When resizing, you should almost always maintain the aspect ratio unless you specifically need a different shape. Resizing without maintaining aspect ratio stretches or squashes the image, distorting everything in it. Faces become wider or narrower. Circles become ovals. Buildings lean.
Most image resizing tools lock the aspect ratio by default and adjust both dimensions proportionally when you change one. If you need to fit an image into a specific width and height that has a different aspect ratio, you have two choices: crop the image to the new ratio before resizing, or use "fit" mode which adds padding (usually white or transparent) to fill the remaining space.
How to Resize an Image in Your Browser
You do not need Photoshop or any installed software to resize an image. A browser-based resizer does everything on your device in a few seconds. Here is the process.
- Open the Image Resizer tool.
- Upload your image, or drag and drop it onto the tool.
- Enter the new width or height. With aspect ratio locked, the other dimension updates automatically.
- Choose whether to resize by exact pixels or by percentage.
- Download the resized image. It is never uploaded to a server, so it stays private.
If the goal is a smaller file rather than smaller dimensions, resizing is only half the job. Read on for the difference, because it is the single most common point of confusion.
Dimensions vs File Size: They Are Not the Same Thing
People often say they want to make an image smaller when they mean one of two very different things: fewer pixels, or fewer kilobytes. Resizing changes the pixel dimensions. Compression changes the file size for a given set of dimensions.
Reducing dimensions does reduce file size, because there are fewer pixels to store. But you can also shrink file size dramatically without touching the dimensions by compressing the image, which stores the same pixels more efficiently. For web use, the best results usually come from doing both: resize to the dimensions you actually need, then compress.
As a rule of thumb, if an image will be displayed at 800 pixels wide on a page, there is no benefit to serving a 4000-pixel-wide file. Resize it down to roughly the display size (or twice that for high-density screens), then compress. This is one of the biggest wins for page load speed.
Resizing Without Losing Quality
Downscaling is effectively lossless in perceived quality, so making an image smaller is safe. The quality traps are elsewhere.
- Do not upscale a compressed image repeatedly. Each round trip through a lossy format like JPEG throws away a little more detail. Work from the original whenever possible.
- Pick the right format for the content. Photographs suit JPEG or WebP; graphics with sharp edges and few colours suit PNG. Resizing a logo saved as JPEG will show blocky artefacts around the edges.
- Resize once, from the highest-quality source. Going from the original to your target size in a single step preserves more detail than resizing in stages.
Cropping vs Resizing
These are different operations that people often conflate. Resizing keeps the whole image but changes its dimensions. Cropping removes part of the image to change its shape or focus.
When a target size has a different aspect ratio from your image, resizing alone will distort it. Crop to the correct ratio first, then resize. For example, to fit a landscape photo into a square profile picture, crop it to a square, then resize that square down to the required pixels.
Common Size Requirements
| Platform or use case | Recommended dimensions | Notes |
|---|---|---|
| Website full-width hero image | 1920 x 1080px | Compress after resizing |
| Blog post image | 1200 x 630px | Also used for social sharing |
| Twitter / X header | 1500 x 500px | 3:1 ratio |
| Facebook profile picture | 400 x 400px minimum | Displayed as circle |
| LinkedIn profile picture | 400 x 400px minimum | Up to 7680 x 4320px |
| Instagram post (square) | 1080 x 1080px | 1:1 ratio |
| Email signature image | 200 to 300px wide | Keep file size under 30KB |
| Print at 4 x 6 inches | 1200 x 1800px at 300 DPI | Higher resolution for larger prints |
Frequently Asked Questions
How do I resize an image without losing quality? Make it smaller rather than larger, work from the original file, and resize in a single step. Downscaling does not visibly reduce quality. Only upscaling invents detail and softens the result.
Why is my image blurry after resizing? The most likely cause is upscaling, where the software has to invent pixels. Blur can also come from resizing an already-compressed image or from an inappropriate resampling method. Start from the original and avoid enlarging beyond the source resolution.
How do I resize an image to a specific file size, like under 1MB? Resize the dimensions to what you actually need, then compress. Compression, not resizing, is what targets a specific kilobyte budget for a given set of dimensions.
Does resizing change the aspect ratio? Only if you unlock it. Keep aspect ratio locked and both dimensions scale together, so nothing is stretched. Unlock it only when you deliberately want a different shape.
Resize an Image Free
Resize any image to exact pixel dimensions or by percentage, directly in your browser. Your image is never uploaded to a server.