Do We Need To Embed Profiles Into Image Files?

Do We Need To Embed Profiles Into Image Files?

If correct colour management and maintaining colour accuracy as much as possible as image files are moved from camera, to editing software, screens or printers, then the answer is a definite yes....

Hopefully this explanation and analogy will make it clear as to why.

First, we need to be aware that rgb colour values for a given true colour are device dependent and not absolute. For example, say an image pixel is set to rgb 0,128,0. Yes that will display as a shade of green. But the shade of green a screen will display will be relative to pure green (0,255,0) and every screen will display 0,255,0 slightly differently because of design parameters, manufacturing tolerances, screen settings and other factors. To standardise the colours a screen displays is where calibration comes in. But calibration is not the subject of this post. We need to embed profiles into image files whether we have a calibrated screen or not. The point I am making so far is that rgb values for a given true colour are device dependent.

So how do we get around the fact that all screens will display a given colour slightly differently? This is where embedding the colour space that was used to create the image plays a vital role.

Consider this analogy that hopefully explains.

Lets say we have some text in Spanish on a sheet of paper that we need to translate to Greek on another piece of paper. The source and target pieces of paper are analogous to the colour spaces used by 2 different screens. The text on each piece of paper is analogous to the rgb values of an image in the respective colour space.

Now, if we had a Spanish to Greek dictionary then the translation would be simple but lets say such a dictionary does not exist and so we then have to do the translation from Spanish to Greek via an intermediary language, say English, using a Spanish -> English and English -> Greek dictionary. This is exactly what happens in colour management when an image is transfered from one screen to another. The Spanish -> English dictionary is analogous to the colour profile of the source screen and the English -> Greek dictionary is analogous to the colour profile of the target screen.

So when an image is opened on a different screen the colour management system uses the profile embedded in the image to convert the image's colours to a true colour (most likely Lab colour since Lab is device independent) using the embedded colour profile. This step is analogous to the source text being translated from Spanish to English.

The colour management system then uses the target screen's profile to convert the Lab colours to the target screens colour space. This is analogous to translating the English text to Greek.

So now, just like you have the original Spanish text translated to Greek on another piece of paper, you now have the colours in the image "translated" to the color space of the target screen.

How closely the colours match on the 2 screens is mainly dependent on whether the source and target colour spaces/profiles are calibrated colour spaces or not. But that is a whole different discussion.

So hopefully if you have read down to this line, you can see why it is vital to embed the colour space/profile that was used to create an image into image files to be used on other devices (screens, applications, printers)

HTH someone