File Size, Dimensions, Resolution and DPI: What Each One Actually Controls
Four words people use interchangeably that mean four different things. Which one a form is asking for, and which lever to pull when you need a smaller file.
Updated August 11, 20265 min read
"Make the image smaller" is genuinely ambiguous, and the ambiguity costs people hours. It can mean fewer pixels, fewer kilobytes, or a smaller printed rectangle — three different operations with different tools and different consequences. Add DPI, which most people believe controls quality and which usually controls nothing at all, and you have four terms that get used interchangeably and mean four separate things.
The four terms, separated
| Term | What it measures | Example | Changed by |
|---|---|---|---|
| Dimensions | How many pixels wide and tall | 4000 × 3000 | Resizing |
| File size | How many bytes on disk | 2.4 MB | Compression, format, dimensions |
| Resolution | Loosely, total pixel count | 12 megapixels | Resizing |
| DPI / PPI | Pixels per inch when printed | 300 DPI | A metadata tag; almost never changes anything on screen |
The first three are related but not interchangeable. The fourth is the one that causes real confusion, so it is worth taking on its own.
DPI does nothing on a screen
DPI is a number stored in the file that says how large the image should be *printed*. A 3000 × 2000 photo tagged at 300 DPI prints at 10 × 6.7 inches. The same file tagged at 150 DPI prints at 20 × 13.3 inches. The pixels are identical in both cases — only the printing instruction differs.
On a screen, DPI is almost always ignored. A browser, a phone gallery and an upload form care about the pixel dimensions — the one narrow exception being some browsers, which honour a resolution in a photo's EXIF data when it comes with matching pixel dimensions, so a 144 DPI screenshot from a high-density Mac can show at half size. So changing a file from 72 to 300 DPI without resampling does not improve it, does not make it larger, and does not make it acceptable to a printer that wanted more pixels.
What actually drives file size
Three things, in descending order of impact for photographic content:
- 1Pixel count. The dominant factor. Halving both width and height removes 75% of the pixels, and file size falls roughly in step.
- 2Format. For the same photo at comparable quality: PNG is largest, JPG is the baseline, WebP is about a third smaller, AVIF roughly half.
- 3Compression setting. Within a lossy format, how much detail the encoder is allowed to discard. Useful, but the lever that damages the picture most per byte saved.
Content matters too, and it surprises people. A photograph of a plain grey wall and a photograph of a forest can share dimensions, format and quality setting and differ several-fold in size, because compression works by finding repetition and one of those images has far more of it. It is also why a scanned page of text compresses to almost nothing while a portrait of the same dimensions does not.
Which lever for which problem
| What you were told | What they mean | What to do |
|---|---|---|
| "Max 2MB" | File size | Compress to a byte ceiling |
| "At least 1200px wide" | Dimensions | Resize — do not compress to fix this |
| "Between 20KB and 50KB" | File size range | Resize to KB inside the window |
| "High resolution" | Usually dimensions | Supply the largest original you have |
| "300 DPI" | Pixels for a print size | Multiply inches × 300 to get the pixels needed |
| "Too big to email" | File size | Resize dimensions first, then compress |
The single most common mistake in that table is compressing hard to satisfy a pixel-dimension requirement, or resizing to satisfy a byte limit. They are different axes; pulling the wrong one either fails the check or destroys the image on the way to passing it.
See the real format, exact dimensions and metadata — read from the file itself, on your device.
Analyze an imageMegapixels and what they are worth
Megapixels are width × height ÷ 1,000,000 — a 4000 × 3000 photo is 12MP. It is a single number for total pixel count and it is a poor proxy for quality, because sensor size, lens and light matter more than count. A 12MP photo from a good phone in daylight beats a 48MP one from a cheap sensor indoors, comfortably.
For practical purposes: 2MP is plenty for any screen use, 6MP covers a good-quality A4 print, and 12MP gives room to crop meaningfully and still print well. Beyond that you are mostly buying storage and slower uploads.
When you are unsure what a file actually is — and a .jpg extension is no guarantee of anything — the image analyzer reads the true format, dimensions, colour mode and DPI tag straight from the file's own header. That is usually a faster answer than working backwards from a rejected upload.
If you take one thing from all of this, make it the habit of asking which number a requirement is really about before you touch anything. "Smaller" from a colleague, a client and an upload validator can mean three different operations, and the cost of guessing wrong is a round of rework at best and a destroyed image at worst. Thirty seconds of clarification beats both.
Frequently asked questions
What is the difference between image size and resolution?
"Size" is ambiguous — it can mean pixel dimensions or bytes on disk. Resolution usually means total pixel count. A 4000 × 3000 image has fixed dimensions but its file size varies enormously with format and compression, so the two are related rather than the same thing.
Does changing DPI change image quality?
No, provided the pixels are not resampled. DPI is a tag saying how large to print the image; it has almost no effect on screen and adds no detail. Only changing the pixel dimensions changes how much information the file holds.
How many pixels do I need for a 300 DPI print?
Multiply the print size in inches by 300. A 6 × 4 inch print needs 1800 × 1200 pixels; an A4 page (8.3 × 11.7 in) needs about 2480 × 3508. If you do not have that many pixels, tagging the file 300 DPI does not create them.
Why are two images with the same dimensions different file sizes?
Compression works by exploiting repetition, and some pictures have far more of it. A plain background compresses to a fraction of a detailed forest scene at identical dimensions, format and quality setting. Format choice adds another several-fold difference.