Talk:ICC profile

"See Also:" about "VueScan" needs attention

Hello Everyone!

The truth is, by the time I added that edit my brain was fried; it was a last ditch effort to leave a placeholder that would remind me to deal with it later if someone else did not run with it first. The "See Also:" about "VueScan" would make a lot more sense if the preceding article included a topic such as "Why ICC Profiles Should Be Periodically Recreated", perhaps followed by a topic such as "Tools Required To Create/Recreate ICC Profiles".

My (fuzzy by then) thinking was that all users should be able to make their own ICC profiles if they want to, so pointing at a tool to do so seemed good. Perhaps the needed improvements might look something like this:

Why ICC Profiles Should Be Periodically Recreated

High end (professional-grade) peripherals typically do come with factory supplied ICC Profiles, sometimes customized by serial number, but often more or less generic by model number. Sometimes the profiles are even more generalized as being for any CRT or any LCD (or TFT LCD), or any ink-jet from a particular company (which assumes you are using their ink and their paper!). And entry level (mass-marketed consumer) devices might well be expected to not have a profile available at all. (Then there is the case of digital cameras, which can benefit from profiling using the same reflective IT8 calibration target routinely applied to flatbed scanners; does anyone know if any but the most expensive models are supplied with factory profiles? BTW, Wikipedia's own article about digital cameras does not mention ICC profiles.)

Trouble is, whether any existing profile is custom or generic, as devices age they deviate from the "reality" asserted by the profile. A prudent or meticulous user, seeking excellence, might routinely schedule, as a sort of preventive maintenance, the recreation of his or her system's profiles on a monthly, quarterly or annual basis. An alternative approach would be to redo the profiles at the beginning of each major revenue-producing project, and not worry about it for internal projects never to be seen by clients in particular or the public in general.

Finally, one must not believe a profile makes the device perfect. The successful profile makes the device better (hopefully dramatically so) than an unprofiled unit. Due to issues related to gamut, perfection is indeed elusive. However, modern technology is approaching the level of incredibly good.

Tools Required To Create/Recreate ICC Profiles

This would be an overview about software and supplies, including the digital colorimeter sensor for monitors, the reflective and transmissive IT8 calibration targets for various scanners and digital cameras, high quality papers and inks for the printing of temporary IT8 replicas, and comments explaining why the same paper and ink a project will be printed with must be used to print the IT8 replicas, which are in turn scanned by a recently profiled flatbed scanner (otherwise the result would be bogus) to create the profile of the printer.

IT8 calibration targets are to be stored in an archival manner between uses, just as you would (or should!) store your most precious family photos or those of your proudest achievements. IT8 replicas are never to be reused as a new tool to calibrate yet other devices! Their only purpose is to see how close a given printer can come to the goal. In fact there may be no particular reason to save them once the new printer profile is created.

For final good measure, how about a topic such as:

Brief Summary: How To Profile, by Device

This topic should include only the broad strokes that illuminate the concepts of how to profile: monitor (colorimeter), printer (inks and paper), scanners (reflective IT8 for flatbed, transmissive IT8 for film), digital cameras (reflective IT8, typically in the context of a studio with controllable lighting), digital front projectors (BTW, I don't know how one would go about profiling a projector... Perhaps someone who has worked with the new digital 3-chip DLP projectors in a movie theatre could enlighten us (ew, a pun!).

Only broad strokes are appropriate since the user should favor the advice given by the vendor(s) of both the devices being profiled, and the supplies and tools used to attain the goal.

BTW, I am not an expert on ICC profiles, but as an advanced amateur photographer, I have created them for my own system. Therefore I probably know enough to be dangerous. ;-O

Additional "See also:"s might include: Kodak, X-Rite, and others involved in either developing or supplying IT8 targets; Dye-sublimation printer; Color calibration;

If the above outline can be fleshed out, verified, and referenced, the "stub" status could be removed from this article.

Badly Bradley 11:10, 10 February 2007 (UTC)[reply]

Sounds like good ideas, but what sources would be used? We cannot write from personal knowledge alone. I don't have any books that cover these issues. Notinasnaid 11:31, 10 February 2007 (UTC)[reply]

T'is indeed a problem. Most of what I stated above is from personal experience. Most of the resources for my self-taught success were user's manuals that came with my system and the peripherals and apps I bought. What I couldn't find there, I found on the web. Visits to the library (formerly one of my favorite places) or bookstore are rare these days.

I suppose one could use the web to hunt down authoritative books on the subject, which would kill 2 birds with one stone: verification and citation. That "No Original Research" rule is a pain! Are citations from user's manuals allowed?

A worse problem for me: I never intended to sink so much time into this worthy project. Already I am causing trouble on the home front by neglecting my important duties. If someone else wants to flesh out my outline, that would be wonderful.

Badly Bradley 12:15, 10 February 2007 (UTC)[reply]

I don't mean to be an ass, but the Wikipedia is not a how-to, not an instruction manual -- but the Wikiversity is! You can create a page on the topic there and hopefully attract collaboration.--John Bessa (talk) 23:32, 31 December 2009 (UTC)[reply]

ISO 12234-4 does not exist?

The page refers to "ISO 12234-4: Photography - Electronic still-picture imaging – Part 4: Exchangeable image file format (Exif 2.2) (ISO TC42)". However, that standard does not appear to exist, according to the ISO web site. Should that reference be removed? 80.187.208.18 (talk) 09:15, 6 July 2008 (UTC)[reply]

It does seem peculiar that it's not to be found on iso.org, given that it's so widely referenced (here). Maybe they just don't have the pubication available yet? Dicklyon (talk) 03:23, 7 July 2008 (UTC)[reply]

There is no such standard. Not even in development (also Exif 2.2 is called JEITA CP-3451) https://standards.iteh.ai/catalog/tc/iso/ecb14394-ee54-4b32-87ac-b1d7d54d922e/iso-tc-42 109.252.94.59 (talk) 21:50, 30 December 2020 (UTC)[reply]

This just made me laugh

If necessary the PCS is converted between L*a*b* and CIE XYZ...

Try reading it aloud. rowley (talk) 19:12, 17 January 2010 (UTC)[reply]

Version differences?

Some information on version differences would be nice to have, since some color management programs offer different ICC profile versions for choice. -- 85.179.127.236 (talk) 20:40, 25 May 2010 (UTC)[reply]

Top listed reference.

As of 18th of September 202(4) when I'm writing this, any reference to a Finch, Arnaud does not seem to appear anywhere on the linked article provided. Either a archived version of the site may be needed or the reference may need to be removed. ~RCLZN (talk) 02:03, 19 September 2024 (UTC)[reply]

Here's an archived version of this in case it changes and I look weird later on. ~RCLZN (talk) 02:08, 19 September 2024 (UTC)[reply]