[Openicc] meta data in test chart

[Chris Murphy]

On Jan 21, 2011, at 11:26 AM, edmund ronald wrote:

> Ann, 
> 
>  I would really, really, like the color settings problem to be addressed by the ICC, with procedure guidelines as to what an inkjet printer should support, as to how it can be ensured that the print system can be used for profiling, how it can be ensured that the profile matches with the printer queue/ media employed. 
> 
> But the total mess that is inkjet printing on the platforms supported by the ICC shows no sign of abating, I see a minor member of the ICC (Xrite) having its clients unable to profile, when they buy the tool in order to, another minor member (Adobe) plagued by printing bugs from version to version of its least selling  product, another hardware vendor (Epson) distributing free mini-RIPS in the hope that users will be able to print by bypassing the print system of the base OS. 
> 
>  All we seem to be discussing here is the minor issue of the format for settings; well let me turn this round: If XML, Json or whatever is employed, I am sure we can embed the whole mess in a profile, or embed the profile in the settings or whatever, and make the system work regardless of syntax. 
> 
>  It would help if you chose to sit down *here* and made helpful suggestions to us chickens about how to design a system that works, as all the talking over there sure ain't produced results.

The problem on Mac OS X is not related to incorrect color settings. A way to reliably transport/recreate settings across computers is not the way to resolve messy inkjet printing.

On Windows it is an opt-in system of color management, so ICM/WCS do not become involved in conversions unless applications and/or print drivers explicitly ask it to. While the user interface is clunky and shows options that are mutually exclusive, we don't get a behind the scene override of abundantly clear user intent on Windows.

This is unlike ColorSync on Mac OS which is opt-out, meaning that it by default ColorSync conversions do occur unless applications explicitly uninvite it. And this is done with private, undocumented, service programming interfaces (SPIs) that only one company is really using at the moment, with two products. On Mac OS X we have a unique problem. The SPI is just not robust enough to act as a clear indicator of application intent.

I don't see a role for the ICC in this. They didn't cause these issues in the first place, they were the business choices made by Apple and Microsoft. And the manufacturers of the printers have tepid ICC support, and only with their proprietary Windows and Mac OS drivers. So I'm a little hard pressed to imagine what the ICC's role could be, other than a political one.


Chris Murphy