Assign Color Profile vs Convert

There's an interesting history to sRGB too. It was intended to describe an average consumer-grade CRT monitor with the phosphors in common use at the time. As such, the primaries matched very well, and the sRGB tone response curve was also a very close match to an actual physical unit (that's why the curve has all these idiosyncrasies like varying gamma and the flat "toe" near black).

This had the implication that almost any monitor could be calibrated to reproduce sRGB with very high accuracy.

That's no longer the case. LCDs have different physical properties. The primaries don't match - they are close, but shifted sideways quite a bit. The tone response curve is also quite different. Most importantly, LCDs have a marked dip near black, which CRTs didn't have, and which is therefore not reflected in the sRGB tone curve. The net result is that you tend to see excessive black clipping without color management and a custom monitor profile.

But at the time, sRGB was a monumental achievement. You could say it was the start of color management.

The difference between assign and convert:

To assign an ICC profile simply changes the profile "tag", it switches the icc profile that's embedded.  The image appearance may change significantly - it's rare that assign profile is needed (unless an image arrives with no embedded profile).

To convert translates an images colour values to those needed in the new colourspace (AND then we ensure we embed the ICC profile when saving). In this case image appearance will only change if the destination profile is defining a smaller colourspace, think newsprint, for that the colours have to be desaturated. 

Think of a book:

Its cover tells us its German, if the content IS German it would be erroneous to change the cover to a French one.

However, if it has no cover, then adding one aids understanding. 

And in the rare case that it has the wrong cover, switching to the right one would be advantageous. That's assign.

If the book is German and needs to be read in French, then it doesn’t just need a new cover it needs actually translating and when that’s done it needs the right cover too. That's convert.

It would help us to help you if we knew in what circumstances you are making this decision please.

Is the eventual use for print?

In what type of media?

What icc profile is embedded in the images currently?

I hope this helps

neil barstow, colourmanagement.net

[please do not use the reply button on a message in the thread, only use the one at the top of the page, to maintain chronological order]

. 

"Adobe RGB 1998 is designed for offest press. sRGB is designed for digital output or digital viewing. (ex., iPad and digital press)." Katie_Houghton

Hi Katie,

I feel it's difficult to totally simplify ICC profile choice including working colour space choice.

Digital presses seem to be ever more closely moving towards the gamut of ISOcoatedV2, in some cases this will mean that sRGB is over restricted for digital print. They have a way to go as yet from what I've seen.

I suggest users think of  "sRGB" as a default for internet. However, some basic photo workflows e.g HP's home user inkjet printers and lab services using devices such as Fuji Frontier are basically tuned to sRGB also. If working and sending files to clients with little or no interest in colourmanagement sRGB can be safer also, irrespective of final use.

I remember a very high end advertising photographer client who received varying criticism of colour balance on critical files from client who (after a few revisiions), turned out to be viewing on a Blackberry!

For years sRGB was able (as intended) to represent "the average" monitor display (IMO, although that’s been OK for a lot of home users, 'average display'' is a concept that has little place in a colourmanaged workflow, though - I am sure we all agree that ideally displays really should be calibrated & profiled).

sRGB's definitely not a "bad" colourspace but, IMO, for critical use in a print workflow larger colourspaces can provide some advantages (and some pitfalls for users with no understanding of conversion. assigning etc.).

why was sRGB designed:

" sRGB is a device independent RGB space which was proposed and adopted by Microsoft and Hewlett Packard in 1996) as a standard color space for the average user.

It is supposed to represent the gamut of the average user's monitor. Problems associated with this include:

• There is no such thing as an average user - especially with the high-end displays used by graphics professionals
• There is no average monitor.

sRGB is not all bad. The basic idea of converting images to sRGB which are headed to the web (and when the audience is the general public) makes sense. - - - ."

from Steve Upton's colorwiki

I hope this helps

neil barstow, colourmanagement.net

[please do not use the reply button on a message in the thread, only use the one at the top of the page, to maintain chronological order]

D.Fosse

you're right about ECI-RGBv2 not being widely adopted in the UK, it's quite widely used in Europe though. 

In Europe there have been attempts to replace it with eciRGB, but that never caught on. eciRGB is mostly used by agencies as a matter of company policy. Most people I hear about say that they "have to" use it, rarely that they want to use it.

Bit of background:

There's a small portion of ISOcoatedV2(ECI).icc that isn't accommodated in Adobe RGB, I imagine it's rarely used in imaging, but it is there potentially in the print colourspace.

ECI wrote (about ECI_RGB v1: "- while there were and still are other reasonable offerings around in terms of RGB working space ICC profiles ECI wanted to see one that … has a gamut that covers all colors that can be printed on today's printing presses - whether sheet fed or web offset, gravure or newsprint - but not much beyond (in order to not to waste precision for bits that never really get used)"

neil barstow, colourmanagement.net

[please do not use the reply button on a message in the thread, only use the one at the top of the page, to maintain chronological order]

Adobe RGB 1998 is designed for offest press. sRGB is designed for digital output or digital viewing. (ex., iPad and digtial press).

If your image is taken in sRGB or you are using it in a digtal format then you should keep it there since the colors in CMYK will be far different than what you see on a device (ex web). If your image was taken in Adobe 1998 then you should go to CMYK for offset printing (CMYK is a very small space so never used for editing) since your colors will stay true. Don't assign RGB to another RGB as your colors will visually shift while being the same value but you can convert. If you are shooting in RAW you can go the whatever space is best for you. (That's what I prefer though my bought images are sRGB so they stay there)

Bottomline is if you are born in a space then stay in the space unless you have to convert (even converting it is not exactly the same but a controlled change). If you are going to an offset press then convert to CMYK. If you are going to a digital press then talk to the printer (some still want conversion but most will not) but some will request you stay in sRGB but never Adobe RGB 1998. Blurb use to have a great document on this: https://support.blurb.com/hc/en-us/articles/207795936-Convert-CMYK-image-to-sRGB

 Basically color management is about the control of color and not about matching color. Bottomline we take and edit images using RGB and sometimes print via CMYK (sometimes we use a bigger gamut then that hence we stay in RGB for digital). I personally mostly use sRGB today.

 In the picture below you can in the first image that sRGB is smaller than Adobe 1998. In the second image you can see for most part all colors in CMYK are in gamut for Adobe RGB 1998. In the thrid image you can see there are some colors in CMYK that are not in gamut for sRGB since I go to digtal output. 

Don't assign! Always convert.

Assign is what you use if the file doesn't have a profile at all - or if someone assigned the wrong profile previously.

When you convert into a different color space, you're not supposed to see any difference in general appearance. The numbers are remapped into the new color space, preserving appearance. That's the whole idea of color management.

To better understand the difference, open an image and put the histogram on top. When you convert, the histogram changes, but the image stays the same. If you assign, the histogram is unchanged because there's no remapping. But the image changes because the numbers get a new meaning.

The color space is what defines the numbers as specific colors.

The advantage of Adobe RGB is that it's larger, so it can contain more highly saturated colors. You won't see that directly on screen, because a standard monitor can't reproduce colors beyond sRGB. For that you need an (expensive) wide gamut monitor. But most printing processes can take advantage of it.

Note that Adobe RGB absolutely requires a fully color managed process to be represented correctly. If you're not sure, stick to sRGB.

Just converting an sRGB file into Adobe RGB won't accomplish anything in itself. But you will have a little more headroom for future edits.