sRGB vs Display P3: Why Your Photo Looks Different on Every Screen

Images Nov 1, 2026 5 min read

You export a photo, it looks right. You upload it, and on the website the colours are flat. Or the reverse: a screenshot from your phone appears garishly oversaturated on someone else’s laptop. Nothing was edited in between. The pixels are unchanged.

What changed is the interpretation, and understanding why takes one idea.

Numbers Are Not Colours

A pixel stores something like 255, 0, 0. It is tempting to read that as “red”, but it is not. It means the most red this system can produce — and how red that actually is depends entirely on which system you mean.

A colour space is the definition that turns those numbers into actual colours. It specifies exactly which physical red, green and blue the maximum values correspond to. Change the colour space and 255, 0, 0 becomes a different red, even though not one bit in the file has moved.

This is the whole problem. The same numbers, read against two different definitions, produce two different pictures.

The Two Spaces You Will Meet

sRGB was standardised in 1996, built around what monitors of that era could do. It is the smallest common denominator and remains the default assumption of the web. Its range of reproducible colours — its gamut — is modest, which is precisely why it is safe: almost every display made since can show all of it.

Display P3 is the wider space now used by most phones, laptops and televisions. It covers roughly 25% more colour than sRGB, and the extra sits mostly in saturated reds and greens. A vivid sunset or a deep green leaf has colours that exist in P3 and simply cannot be represented in sRGB — they get clipped back to the nearest sRGB colour, which is duller.

So P3 is better, in the sense of being able to describe more. That is not the same as being the right choice for a given file.

How It Goes Wrong

Every image can carry a small tag called an ICC profile that says which colour space its numbers belong to. Colour management is simply the practice of reading that tag and adjusting the display accordingly.

The failure has one shape, with two directions:

P3 data, no profile, shown as sRGB. The software sees saturated numbers and interprets them against the narrower definition. Everything comes out too vivid — skin looks sunburnt, reds glow, greens turn radioactive. This is by far the most common version, and it is what people usually mean when they say an image “looks wrong on the web”.

sRGB data misread as P3. The opposite: everything looks washed out and slightly grey.

In both cases the picture is intact. Only the label is missing or wrong.

The Trap: Stripping Metadata

This deserves its own warning because it is advice people follow for good reasons and then cannot explain the result.

Optimising images for the web often includes removing metadata — camera settings, GPS, timestamps — to save bytes. It is sound advice. But the ICC profile is metadata too, and a tool set to strip everything will take it with the rest.

The file gets a few kilobytes smaller and the colours break, in a way that is invisible on the machine you are working on if that machine’s display happens to match the assumption. Keep the colour profile when stripping metadata. It is small, and it is the only part that changes how the image looks.

Convert, Don’t Assign

Image software offers two operations that sound alike and do opposite things. Getting them confused is how colours get destroyed permanently.

Convert to a colour space: the software recalculates every pixel so that the colour you see stays as close as possible to what it was. The numbers change; the appearance is preserved. This is almost always what you want.

Assign a colour space: the numbers stay exactly as they are and the label changes. The appearance shifts, because the same numbers now mean something else.

Assigning is useful in one situation: when a file lost its profile and you know what it should have been, assigning the correct one restores the truth. Used by accident — assigning sRGB to a P3 image instead of converting it — it bakes the wrong colours in.

What to Actually Do

Where P3 Files Come From Without You Asking

You may be producing P3 images without having chosen to. iPhone photos and screenshots are P3 by default. Modern cameras shooting wide-gamut profiles are too. Screen recordings and design tools on recent hardware often are as well.

This is why the problem appears out of nowhere for people who never touched a colour setting: the capture side went wide while the delivery side stayed sRGB, and the only thing holding it together is a tag that some tool in the middle may quietly discard.

Quick Reference

Solve your file problem: Image Convert Try it free — Image Convert →