You edit a photograph until the sky is a particular blue. Send it to a friend and their phone makes it almost turquoise. Open it on an older laptop and it seems grey. No one necessarily changed the file. The file and the screen have different jobs, and the translation between them is where the discrepancy appears.

Pixels are instructions, not little pots of paint

A digital image records numbers for colour channels, usually red, green and blue. Those numbers do not describe an absolute colour by themselves. They need an agreed interpretation, called a colour space or profile, that says what the channel values are supposed to mean. A colour-managed system reads the image profile and translates its values for the actual display.¹,²

A familiar baseline is sRGB. Many cameras, phones and displays also work with wider-gamut spaces such as Display P3, which can represent some more saturated colours. If an app ignores the image profile or assumes the wrong space, the same numbers can produce unexpectedly vivid or subdued results.³

This is why an image that looks fine inside an editing app may change when uploaded or opened somewhere else. The file may have been exported with a different profile, the profile may have been omitted, or the viewing software may handle colour differently. The point is not that one set of pixel numbers is secretly the true one. The point is that numbers need a shared language.

Screens have physical limits

Even with correct profiles, displays are not identical. Their panels and light sources can reproduce different ranges of colour. A screen cannot show a colour outside its capabilities exactly; it must approximate or compress it. Two screens may also have different white points, contrast and peak brightness. One makes a shadow look deep and cool; another lifts it and gives it a warmer cast.

Calibration measures and adjusts a display so its output is more predictable. For precise work, a measuring device can characterise the screen and create a profile for colour-managed software to use.¹ That reduces avoidable differences, but it does not turn every panel into the same physical object. Nor will a profile rescue an image viewed through a blue-light setting or a device with an aggressively altered picture mode.

Brightness deserves its own mention. A bright screen in a dark room can make a photograph feel punchy, while the same image on a dim display in daylight may look washed out. Our visual system adapts to its surroundings; a white page seen beside a window is not experienced exactly like a glowing screen at night. This makes “same appearance” harder to achieve than “same data”.

What happens when you share a picture

A sensible workflow is to edit on a reasonably calibrated display, keep the image’s colour profile attached, and export in a broadly supported space when the destination is unknown. For general web sharing, sRGB remains a practical common denominator, while wider-gamut workflows can be useful when the whole chain supports them.³

The chain matters. Camera capture, editing software, export settings, website processing, viewing app, operating system and display each have a chance to interpret or alter the image. A site may resize or recompress a file; a phone may change its white balance automatically; a browser may be more or less careful with embedded metadata. When something looks wrong, inspect that chain before assuming the photograph itself is damaged.

There is also an artistic question. A photographer can aim for a controlled reference appearance, but cannot control every viewer’s display or room. That is true of print as well: ink and paper have their own colour range, and gallery lighting changes the result. Digital delivery makes the variation more obvious because the same file is opened on so many devices.

A small test for a confusing mismatch

If a shared photograph looks wrong, compare it in two colour-managed viewers on the same device. If it changes between apps, interpretation or profile handling is a likely suspect. If it looks consistent within one device but different across devices, inspect display settings: brightness, automatic colour temperature, picture modes and calibration.¹

You can also look at neutral areas. A white wall that shifts visibly warm or cool suggests a white-point difference. A vivid flower that changes more than the wall suggests differences in gamut or saturation handling. Shadow detail that disappears on one screen points towards contrast or brightness. These observations do not replace measurement, but they help identify which part of the chain deserves attention.

Colour profiles are especially important when moving between photography, design and print. A display makes colour with light; a print reflects light from paper and ink. A proofing workflow uses profiles to predict that translation, but a luminous blue on a screen may be impossible to reproduce with a particular printer and paper. The same principle explains why “send the same RGB file” cannot promise identical physical colour everywhere.²

For everyday sharing, perfection is not a realistic target. Consistency is. A predictable export, a sensible common colour space and a known editing display prevent many avoidable surprises, while leaving room for the honest fact that viewers bring their own screens and surroundings.

The practical lesson is reassuring. A colour mismatch does not mean your eyes are unreliable or that the image has mysteriously changed. A photograph is a set of instructions for producing light. Different screens, settings and surroundings carry out those instructions with different tools.