Resizing and compression are different
A 4000 × 3000 image has twelve million pixels. If a page only displays it at 800 pixels wide, sending the original may waste bandwidth. Resizing can reduce the number of pixels before the image is saved. Compression then reduces file size by representing those pixels more efficiently.
The best settings depend on where the image will be used. A small thumbnail does not need the same dimensions as a print master. Keep an untouched original until you have checked the compressed or resized result.
Original: 4000 × 3000 pixels, 4.8 MB. Target use: an 800-pixel-wide blog card. Resize to around 800 pixels wide, preserve the original aspect ratio, then choose a suitable format and inspect the result. The final file size varies with image content and settings.
Choose the right format
JPEG is widely compatible and often useful for photographs. WebP can produce compact photos and graphics in modern web workflows, though you should check the needs of older software or recipient systems. PNG is useful when lossless detail or transparency matters, but it can be much larger for photographs. SVG is suited to vector artwork such as simple icons, not ordinary camera photos.
Do not select a format only because it is fashionable. Think about transparency, gradients, sharp edges, editing needs and the devices that will open the file. If another person must edit it later, keep an editable original alongside the compressed version.
Preserve the aspect ratio
The aspect ratio is the relationship between width and height. A 16:9 image is wide; a 9:16 image is vertical; a 1:1 image is square. Resizing while preserving the ratio avoids stretching faces, logos or objects. If a target platform requires a different ratio, crop intentionally and check that the important subject remains visible.
For a profile picture, central framing may work well. For a thumbnail, the subject and any text need enough space at small sizes. Do not stretch a square image into a banner just to fill a box; choose a crop or a background that suits the target dimensions.
Compression quality and visible artefacts
Lossy compression discards some information to reduce size. At high compression, you may see blocky edges, blurred textures, colour banding or halos around text. Lower compression usually preserves more detail but creates a larger file. The right setting is a balance rather than the smallest possible number.
Compare the result at the size people will actually view it. Zooming to 400% can reveal flaws that no one sees in a normal card, while judging only from a tiny preview can hide damage. Pay special attention to text, skin, thin lines and gradients.
Resolution, print and privacy
Pixel dimensions usually matter more for screen display than the DPI value stored in a file. Print workflows may need a target physical size and pixel density, so follow the printer’s instructions. A DPI label alone does not create extra detail if the source image has too few pixels.
Before uploading a photo, consider whether it contains a home address, school badge, document number, location metadata or another person’s face. Use local browser tools when possible and understand when an upload is sent to an external AI provider. Never upload confidential images just to test a compressor.
A five-point image check
First, make a copy of the original. Second, set target dimensions and preserve the ratio unless you intentionally crop. Third, choose a format appropriate for the content. Fourth, inspect the file at its real display size and compare it with the original. Fifth, reopen the saved output to confirm that it is not corrupted and has the expected dimensions.
Keep filenames meaningful and avoid overwriting the original accidentally. For a website, use descriptive alt text that explains the image’s purpose rather than repeating keywords. A compressed file that loads quickly and remains clear is more useful than a tiny file that makes the content hard to understand.