[Openicc] meta data in test chart

[Chris Murphy]

On Jan 22, 2011, at 12:03 AM, Richard Hughes wrote:

> On 22 January 2011 02:25, Chris Murphy <lists at colorremedies.com> wrote:
>> But the manufacture of ink and paper is always subject to variation. So at any time we can get a change in behavior, and it's usually not predictable. You put the paper in, make a reference print - oh how about that! Time to reprofile! That's about the best case scenario.
> 
> Maybe I'm looking at this from a different perspective, specifically
> these people: http://www.packagekit.org/pk-profiles.html
> 
> Most of these people won't ever change the defaults. Most of these
> people won't profile anything. But what I want to allow is Bevan to
> buy a ColorMunki, plug it in, print out a reference, scan it in, and
> then the printer is "fixed". These kind of people will just re-profile
> if they are going to print 400 photos on glossy paper, and probably
> just for special occasions.
> 
> I know it's a very different use case to the ones you guys and girls
> are discussing, but still an important one in my opinion. We just need
> to make the common case work out of the box, and the special-case
> possible.

No that's pretty much the usage case. The instant someone is going to do ICC - whether canned profiles or custom measured profiles - it's a totally different workflow right now on Mac OS and Windows.

I wish it worked like you describe. I'd much rather have the workflow stay the same, and simply "update" the driver for your printer and media that you're using. But that's not how it's designed. The profile is treated very secondary, rather than primary. It isn't even the default workflow on Mac OS and Windows. Proprietary is the default. If you want to use ICC then you have to change at least the driver settings, and possibly application settings too.


Chris Murphy