Image Dimensions vs File Size: What's the Difference?

Published August 12, 2026

"Your image is too large." It's one of the most frustrating error messages on the internet, because large means two completely different things and the site rarely tells you which one it meant. Did it reject your photo because it's 4000 pixels wide, or because it's 6 megabytes? Those are separate problems with separate fixes, and trying the wrong one wastes your afternoon.

This is the plain-English explanation of the four numbers that describe every image — dimensions, file size, DPI and quality — what each one actually controls, and how to know which one to change.

Dimensions: how many pixels there are

Dimensions are the width and height of an image in pixels — the tiny coloured squares it's built from. They're always written width first: 1920×1080 means 1920 pixels across, 1080 down.

Dimensions determine how big the image can be displayed or printed before it starts to look soft. A 400×400 photo looks perfect as a profile picture and terrible as a full-screen banner, because there simply aren't enough pixels to fill the space — the ones that exist get stretched. Dimensions are a hard limit on detail: you can throw pixels away, but you can never truly get them back.

You change dimensions by resizing.

File size: how many bytes it takes up

File size is how much storage the image occupies — measured in kilobytes (KB) and megabytes (MB). It's what fills your phone, what an email attachment limit measures, and what a visitor has to download before they see your page.

File size is influenced by dimensions but not determined by them. Two 1920×1080 images can be 200 KB and 4 MB, because file size also depends on:

• Format — JPG and WebP compress photos aggressively; PNG stores them almost losslessly and ends up many times larger.
• Quality setting — a JPG saved at 60% is a fraction of the same image saved at 100%.
• Content — a busy forest photo has far more detail to encode than a plain blue sky, so it compresses worse at identical settings.

You change file size by compressing, by changing format, or — most effectively — by resizing.

How the two are connected

Here's the relationship that explains most of what confuses people. The number of pixels in an image is width × height, so shrinking both sides cuts the data squared:

• 4000×3000 = 12,000,000 pixels — maybe 6 MB
• 2000×1500 = 3,000,000 pixels — maybe 1.5 MB (half the width, a quarter of the weight)
• 1000×750 = 750,000 pixels — maybe 400 KB (a sixteenth of the original)

That's why resizing is such a blunt, effective instrument. Halving an image's width removes roughly 75% of its file size, and if the image was oversized for where it's going, it costs you nothing visible. Compression, by contrast, usually saves 40–60% and always trades away a little fidelity.

The professional order of operations: resize first, compress second. There's no point carefully compressing pixels you were about to delete.

DPI: the number that fools everyone

DPI ("dots per inch", sometimes PPI) is the most misunderstood setting in imaging. Here's the truth: DPI is just a note written in the file's metadata. It tells a printer how many pixels to squeeze into each inch of paper. It does not add or remove a single pixel, and it has zero effect on how an image appears on any screen.

Change a photo from 72 DPI to 300 DPI without touching its dimensions and the file size stays byte-for-byte identical. Nothing improved. All you changed was the printing instruction — the same photo would now print smaller and sharper on paper.

The practical rule: if someone asks for "300 DPI", what they really need is enough pixels. Multiply the printed size in inches by 300. An 8×10 inch print needs about 2400×3000 pixels. Get the pixel count right and the DPI tag sorts itself out. For anything shown on a screen — web, social, email, presentations — ignore DPI completely.

Quality: what compression throws away

The "quality" slider you see when saving a JPG or WebP controls how much detail the compressor is allowed to discard. At 100% it keeps nearly everything and the file is fat. At 50% it aggressively simplifies areas the eye is bad at noticing, and the file shrinks — but push too far and you see artefacts: blocky squares in smooth gradients, muddy halos around sharp edges, blotchy skies.

For photographs, 75–85% is the sweet spot: nearly invisible on screen, typically half the file size of maximum quality. Below 60% most people can spot the damage.

One warning worth remembering: JPG compression is lossy and cumulative. Every time you open a JPG and re-save it, more detail is destroyed — permanently. Edit from your original, save once, and keep the untouched master file somewhere.

Which one do you need to change?

Read the error message and match it to the fix:

• "Maximum 2000 pixels wide" → a dimensions limit. Resize.
• "File must be under 2 MB" → a file size limit. Resize first (usually enough on its own), then compress if needed.
• "Image must be exactly 1080×1080" → a dimensions requirement with a fixed shape. Crop to 1:1, then resize.
• "Image resolution too low" / looks blurry when printed → not enough pixels. Find a bigger original; enlarging won't rescue it.
• "Please upload at 300 DPI" → they want enough pixels for print. Printed inches × 300 = pixels needed.

Real example: a form wants a photo under 500 KB and yours is 5 MB straight from your phone at 4032×3024. Resizing it to fit inside 1500×1500 alone typically lands you around 300–400 KB, in one step, with no visible loss on any screen. Compression alone would have got you to maybe 2 MB while making the picture slightly worse. That's the whole lesson in one example.

A quick reference table in words

Dimensions — measured in pixels, changed by resizing, controls how large the image can be shown, and cannot be recovered once reduced. File size — measured in KB/MB, changed by resizing, compressing or changing format, controls upload and load times. DPI — metadata only, irrelevant on screen, matters solely for print. Quality — the compression setting, trades file size against visible artefacts, and its damage is permanent for JPG.

Is it private and free?

Yes. Toolyard's resizer and compressor both run entirely inside your browser using the HTML Canvas — your images are processed on your own device and are never uploaded to a server. No account, no watermark, no file limit, and it keeps working with your connection switched off, which is the easiest way to prove nothing is being sent anywhere.