sRGB, Display P3 and colour profiles, explained

Numbers need a colour space

A pixel stores three numbers for red, green and blue. On their own, those numbers don't say exactly which red or which green. A colour space supplies that meaning. MDN describes a colour space as a defined range of colours, its gamut, with rules that let devices and software reproduce colour consistently.

So the same three numbers can mean different colours in different colour spaces. Full red in a wide colour space is a deeper red than full red in a narrow one.

Three colour spaces you'll meet

A wider gamut can show more saturated colours, but only on a screen or printer able to reproduce them.

• sRGB: the standard colour space, defined in IEC 61966-2-1:1999 according to the International Color Consortium's registry. MDN notes it was created for monitors, printers and the web, and is the most widely used colour space.
• Display P3: defined by Apple, a wide-gamut space with more vivid reds and greens than sRGB, as MDN puts it. Apple's device reference lists Display P3 as the native colour space of displays such as iPhone 7, iPhone 8 and iPhone X.
• Adobe RGB (1998): a wider space that MDN describes as designed to represent all the CMYK colours as RGB, which is why it is used in print-oriented work.

What a colour profile does

A colour profile, usually an ICC profile, is data that defines a colour space. As the WebKit blog explains, it can be embedded directly into an image so that software interprets the colour values correctly. A photo tagged as Display P3 says, in effect: read these numbers as P3.

An image without a profile is called untagged. MDN notes that sRGB is now usually the assumed colour space for content without an embedded profile, and WebKit states that it treats untagged images as sRGB.

Why colours can look duller after conversion

Now the cause of the flat photo is clear. Take a picture whose numbers were meant as Display P3 or Adobe RGB. If a tool keeps the numbers but drops the profile, software reads them as sRGB. Each number now points to a smaller, less saturated colour than intended, so the most vivid areas tend to look flatter and colours can shift slightly.

A related effect happens even with the profile kept. On an sRGB screen, the WebKit blog explains, colours outside sRGB have to be squeezed into colours the screen can show, and its example photo looks less vivid there. That is a limit of the screen, not of the file.

What KeepLocalHub does with profiles

This is a real limitation. The converters, compressor, resizer and cropper save the new file without an embedded colour profile. The pixel values are kept as they were decoded, and most apps then treat the file as sRGB. For sRGB images that is fine. For a photo with a wide-gamut profile, the result can look duller than the original.

Remove EXIF data works differently. For JPG, PNG and WebP it removes the metadata blocks without re-saving the pixels, and it keeps the colour profile and the orientation tag. HEIC and AVIF files are saved as a new JPG instead.

When to care, and what to do

For most everyday photos the difference is small. It matters when colour itself is the point:

• Sharing everyday photos online: usually no visible difference; convert as normal.
• Photos with Display P3 or Adobe RGB profiles where colour matters, such as product shots or artwork: compare the result with the original side by side before deleting anything.
• Only want to remove details before sharing: use Remove EXIF data on JPG, PNG or WebP, which keeps the profile.
• Print work in Adobe RGB: keep the original file and let your editing software handle colour conversion.