[Openicc] colord 0.1.0 released!

[Leonard Rosenthol]

On Mon, Jan 17, 2011 at 7:41 PM, Graeme Gill <graeme at argyllcms.com> wrote:

> Alastair M. Robinson wrote:
>
>> In that case the rendering pipeline needs to detect when source and
>> destination profiles are identical, and bypass the transform completely.
>> Allowing the CM engine to transform the data between two CMYK profiles
>> would be disastrous for profiling - even if they're identical - because
>> it would mess up the GCR.
>>
>
> But this is exactly the root cause of the Apple profiling problems !
>
> They treat matching profiles as a null transform, and as a side
> effect use it to pass unmodified numbers to the device.
>
>
the other issue is that you'd have to potentially do this test on an object
by object basis, since each object in PDF can be profiled separately.  There
is also the possibility for profiles as part of the transparency blending
spaces, etc.

This complexity is significantly(!!) reduced with PDF/X and PDF/A!  If there
were some way to ensure that all PDFs that came through CUPS were PDF/X-4
compliant, you'd be able to better optimize the entire process.  But I don't
think you have that level of control :(.

 Leonard
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.freedesktop.org/archives/openicc/attachments/20110117/21e16abb/attachment.htm>