Resizing an image is simple, but understanding what happens behind the scenes helps you get better results. Type in the dimensions, click resize, done. And for a lot of use cases, that's exactly right. But there are a few situations where understanding what's happening under the hood saves you from a result that looks worse than it should or causes a submission to fail silently.
When you resize a raster image (a JPEG, PNG, or WebP photo), the software has to either add or remove pixels. Reducing the size is straightforward: pixels are dropped or merged. The image gets smaller and generally looks fine. Increasing the size is harder, because the algorithm has to invent new pixels that weren't in the original. It does this by sampling surrounding pixel values and interpolating. The result is a larger image, but a slightly softer one. At modest upscaling amounts, say 20 to 30%, you probably won't notice. At 2x or 3x, you will.
Pixels vs file size: what you're actually changing
Resizing and compressing are different things, but they're often confused. Resizing changes the pixel dimensions of an image. A 4000 by 3000 pixel photo resized to 1200 by 900 will have fewer pixels and generally a smaller file size, but the exact file size depends on the format and compression settings you choose when saving. Compressing an image reduces the file size by encoding the existing pixels more efficiently, without necessarily changing the dimensions at all.
If you need an image under a specific file size limit (like 100KB for a portal upload), resizing alone might not get you there. You may also need to compress the image separately, or use the main compressor on the homepage, which lets you control both dimensions and quality in one step.
When JPEG is the right output and when PNG is
If you're resizing a photo to share online, send by email, or upload to a portal, output as JPEG, or convert PNG to JPG first if that's your starting format. It produces smaller files and is accepted everywhere. The quality loss from JPEG compression at 80 to 90% is invisible at normal viewing sizes.
Use PNG output when your original image contains transparency or when you want to preserve transparent regions of the image. Note that in Fit mode, any added padding is currently white rather than transparent. PNG is also the right choice for screenshots, diagrams, or images with sharp text, because JPEG compression creates visible artifacts around high-contrast edges like text on a solid background. For those, PNG's lossless encoding keeps the edges clean.
Once the dimensions are right, a few other adjustments often come next. Crop image if you only need part of the frame, change DPI if the image is headed to print rather than a screen, or convert to WebP for a smaller, web-friendly file once you're happy with the size.