There are two main ways to make an image smaller. One is resizing — reducing the number of pixels, i.e. the resolution. The other is compression — keeping the pixel count but packing the data more tightly. They produce different results, so using each in the right place lets you cut file size effectively while losing as little quality as possible.
Once you understand the mechanics, you can fix the two frustrations everyone hits: 'why is it still blurry after I shrank it' and 'why won't it get small enough.'
Lowering resolution (resizing)
Resizing reduces the width and height in pixels. Shrink a 4000×3000 photo to 1600×1200 and the pixel count drops sharply, so the file size falls fast. On screen and on the web, the original is usually far larger than the size it is actually displayed at, so resizing to the display size keeps the visible quality the same while cutting the file size.
Just remember that pixels you remove cannot be brought back. If you might print large later, keep the original separately.
Increasing compression
Compression keeps the pixel count but lowers a JPG's quality value. The lower the quality, the smaller the file — but go too far and colors smear or you get blotchy edges (artifacts). Photos tolerate a fair amount of compression, whereas images full of text or crisp lines are fragile and turn messy quickly.
Measured: how much does halving the width save?
People often say resizing beats compression, but rarely with numbers attached. We held JPG quality at 80 and reduced only the width of the same photo.
| Width | Pixel size | Pixel count | File size |
|---|---|---|---|
| 100% | 1600x1067 | 1,707,200 | 269 KB |
| 75% | 1200x800 | 960,000 | 105 KB |
| 50% | 800x534 | 427,200 | 33 KB |
| 25% | 400x267 | 106,800 | 5 KB |
Look at the halfway row. The pixel count falls to a quarter, but the file drops from 269 KB to 33 KB, roughly 8.1 times smaller. Fewer pixels means neighboring pixels resemble each other more closely, so compression efficiency rises alongside. That is why the size comes before the quality slider.
When quality loss is fine, and when it isn't
Before shrinking, ask whether the loss is acceptable for the purpose.
- Usually fine — social media uploads, blog body images, preview thumbnails, email attachments
- Be careful — printed materials, product photos, text-heavy documents, archival originals
- Don't delete the original — keep it separate from the compressed or resized copy so you can rework it later
Web and print follow different rules
On the web, screen pixels are what matter, so matching the display size is enough — and you can shrink aggressively for faster page loads. Print is measured by dot density (usually around 300 dpi), so an image that looks fine on screen can come out blurry on paper. If printing is the goal, don't drop the resolution too far and don't over-compress.
Start from a target size
In practice, the question is often 'how small does it need to be?' When there is a fixed limit — a 5 MB attachment cap, a 2 MB forum limit — it is easier to compress toward that target. Switching's compress tool lets you set a target size and fits the file under it, and the resize tool lets you adjust the resolution too. Everything happens inside your browser, so files are never uploaded to a server.