[Openicc] The OutputIntent and Color Managed Printing

[Chris Murphy]

On May 20, 2013, at 11:41 AM, Michael Vrhel <michael.vrhel at artifex.com> wrote:

> Personally to me, I don't see any difference in the solution to a retargeting problem and a proofing problem.

If the source PDF is ICCBased or CMYK only, it's pretty straightforward and I'd say there's no difference. But if the document calls for spot colors, embossing, varnishes, etc. then there could be a big difference. A particularly good proof would either extract that channel or layer to a separate page (with registration marks intact), or mark it with a pattern. Whereas retargeting to a composite device it may be appropriate to drop such effects entirely. So it's that less than 1% exception that instantly turns into a manual handling, arts and crafts type scenario.

> 
> Furthermore, more than one output intent (i.e. an arary of output intent dictionaries) can be present in a PDF document.

I'm pretty certain in all current PDF/X's that only one dictionary is allowed. The proposal for future PDF/X to be based on PDF spec 2.0+ is for multiple dictionaries (e.g. per page output intent), so that a cover or insert pages can be defined for printing on different media, while the whole document is still in a single PDF.

> 
>> * Ghostscript uses the OutputIntent as a proofing profile and the
>> sOutputICCProfile is used to specify the print profile of the output
>> device
> 
> The presence of (an) output intent(s) in a PDF document basically suggets that re-targeting is likely necessary, if none of the output intents matches the sOutputICCProfile color space. But re-targeting is IMO not necessarily the same as proofing.

Retargeting should not be expected, proofing should be expected. That's about the only difference I can think of. 

Retargeting implies the PDF/X was prematurely or incorrectly created, such that the Output Intent was set contrary to the actual printing condition.

Chris Murphy