People say “can you compress this?” when they mean “make the file smaller,” and then get frustrated when compression alone barely moves a giant image. The reason is that they skipped the step that matters most. Let's separate the three ideas.
Resize — change the dimensions
Resizing changes how many pixels the image contains: 4000×3000 down to 1600×1200, say. This matters enormously because file size is roughly proportional to the number of pixels. Halving both the width and the height leaves you with a quarter of the pixels — around a 75% size cut before you compress anything.
The catch people fear — “won't it look worse?” — only applies to upscaling. Making an image smaller just discards pixels you weren't showing anyway, so it stays crisp. Making it bigger invents pixels that were never there, so it goes soft. The rule: resize down freely, never up.
Compress — change the bytes
Compression keeps the dimensions but stores the same pixels more efficiently. Lossy compression (JPEG, WebP, AVIF) discards visual detail you can't perceive; lossless (PNG) reorganises data with zero loss. Compression is what turns a correctly-sized image from “fine” into “tiny,” and it's covered in depth in lossy vs lossless compression.
Optimize — the whole recipe
“Optimize” isn't a fourth thing; it's the umbrella term for doing all of the above in the right order to reach the smallest file with no visible loss:
- Resize to the dimensions you'll actually display (×2 for retina sharpness).
- Pick the right format — WebP/AVIF for photos, PNG for graphics.
- Compress to a sensible quality (~80).
- Strip metadata (EXIF, thumbnails, colour profiles you don't need).
Why the order multiplies your savings
Say you start with a 6 MB, 6000×4000 photo for an 800px column:
| Step | Action | Result |
|---|---|---|
| Start | 6000×4000 JPEG q100 | ~6 MB |
| Resize | → 1600×1067 | ~900 KB |
| Compress | → quality 80 | ~250 KB |
| Format | → WebP/AVIF | ~150 KB |
That's a ~97% reduction with no visible difference in the column — and compression alone (skipping resize) would have left you nearer 2 MB. Order matters.
Do it all in one pass
You don't need three tools. The compressor resizes, converts, compresses and strips metadata together — pick a preset or set a max width and quality, and it applies the whole recipe to every image in your batch. If you're aiming this at a website, pair it with image compression for web speed.
Frequently asked questions
What is the difference between resizing and compressing an image? +
Resizing changes the pixel dimensions (e.g. 4000×3000 → 1600×1200). Compressing keeps the dimensions but reduces the file size by encoding the pixels more efficiently. Resizing removes pixels; compressing removes bytes. Both reduce file size — resizing usually more dramatically.
Which reduces file size more, resizing or compressing? +
Resizing, when the image is much larger than needed — because file size scales with the number of pixels. Halving both width and height removes ~75% of the pixels. Once dimensions are right, compression finishes the job. Used together they multiply.
What does “optimize” actually mean? +
Optimizing is the umbrella term for making an image as small as possible without visible quality loss — typically resizing to the display size, choosing the best format, compressing to a sensible quality, and stripping metadata. It is the whole recipe, not a single action.
Will resizing an image lower its quality? +
Making an image smaller (downscaling) keeps it sharp — you are only removing pixels you were not displaying. Making it larger (upscaling) does reduce quality, because there is no real detail to add, so the result looks soft.