DPI explained: why 300 DPI does not make your photo sharper

We build browser-based file tools and test every one of them against the forms and specs people actually have to satisfy.

A form says “scan at 300 DPI”. A printer says the file is “only 72 DPI”. You open the image, change a box from 72 to 300, save, and send it off — and somebody tells you it is still wrong.

They are right, and the reason is the single most useful thing to understand about image files: an image file has no physical size at all.

What an image file actually contains

A grid of pixels. That is it. A photo might be 1200 pixels wide and 1800 tall, which is 2.16 million coloured dots in a rectangle. Nowhere in that data is there a width in centimetres, because the file does not know whether it will end up on a billboard or a stamp.

DPI — dots per inch, sometimes PPI for pixels per inch — is a note attached to the file saying how densely those pixels should be packed when something prints them. It is a suggestion about the future, stored alongside the image rather than inside it.

The arithmetic, which is the whole concept

Printed size = pixels ÷ DPI. Our 1200 × 1800 photo therefore prints as:

Look at the third column. It never changes. Raising the DPI tag did not add detail — it made the same detail print smaller and therefore tighter. That is the entire effect.

So why does changing it to 300 do nothing?

Because if you only change the tag, you have told the printer to produce a smaller print. The image is no sharper; it is denser on a smaller piece of paper. People expect the first and get the second, and the confusion survives decades of explanation.

There is a second reason the change often appears to do nothing at all: screens ignore the DPI tag entirely. A browser, a phone gallery and a messaging app all lay the image out by its pixel dimensions. You can set an image to 1 DPI or 9000 DPI and it will look identical on every screen on earth. Only printing — and software preparing something for print — reads the number.

You can see what your own file declares with the DPI checker: it reads the tag, the pixel size, and the physical size those two imply.

When DPI genuinely matters

• Printing anything you will hold. About 300 DPI is the usual target for photographs at reading distance — that is roughly where most eyes stop resolving individual dots.
• Large-format work goes the other way. A banner viewed from ten metres is fine at 50 DPI, and insisting on 300 means a gigantic file for no visible gain.
• Scanning documents. Here the DPI you scan at is real, because it decides how many pixels get captured in the first place. 300 DPI on a scanner means something; 300 DPI typed into an existing file does not.
• Software that lays out by physical size — design tools, word processors placing an image at “5 cm wide” — uses the tag to decide the default.

What a form asking for “300 DPI” usually means

Nearly always: give us enough pixels. The person who wrote the requirement was thinking of a scanner, where DPI and pixel count are the same decision. Applied to a file you already have, the instruction is not quite meaningful.

So translate it. If the form wants a photograph 35 × 45 mm at 300 DPI, it wants roughly 413 × 531 pixels — and if your image has those pixels, it will be accepted whatever the tag says. If it has 200 × 250 pixels, no amount of re-tagging will rescue it, because the detail was never captured.

Resizing by print size does this translation for you: enter centimetres and a DPI, and it works out the pixel dimensions you actually need.

Re-tag, or resample? Pick deliberately

These are different operations and most tools blur them together.

• Re-tagging changes the stored number and leaves every pixel untouched. The file is byte-for-byte as detailed as it was. Use this when something demands a particular DPI value and your pixel count is already sufficient — which is the common case for application forms.
• Resampling genuinely changes the pixel count, interpolating new ones on the way up or discarding them on the way down. Going up invents detail that was never photographed and usually looks soft; going down is lossy but safe.

Change image DPI lets you do either and shows the printed size as you type, so you can see which one you are actually asking for. Convert to 300 DPI is the same thing with the most-requested number filled in.

The short version

• An image is pixels. DPI is a note about how to print them.
• Printed size = pixels ÷ DPI. Raising DPI alone shrinks the print; it adds nothing.
• Screens ignore DPI completely.
• “300 DPI” on a form almost always means “enough pixels”. Work out the pixel count and ignore the rest.
• You cannot add detail that was never captured — and any tool claiming otherwise is interpolating, not recovering.