[prev in list] [next in list] [prev in thread] [next in thread] 

List:       freedesktop-openicc
Subject:    Re: [Openicc] colord Printing Plans
From:       Peter Skarpetis <peters () skarpetis ! com>
Date:       2011-02-28 8:32:21
Message-ID: 2F8FEB53-E556-4322-9473-6169F5DEC9B2 () skarpetis ! com
[Download RAW message or body]


On 28/02/2011, at 12:35 PM, Graeme Gill wrote:

> peters@skarpetis.com wrote:
> > DeviceRGB should be left alone for the cases where the output is RGB. For other \
> > output colourspaces a default, such as sRGB, should be assigned.
> 
> While that may make sense in some cases, it typically leads to user
> dissatisfaction in many others.
> 
> Typically "dumb" software takes screen numbers (numbers it sends to the
> non-color managed display) and dumps them into DeviceRGB values in
> PS or PDF or the equivalent in GDI and sends them to the printer.
> Typically the printer native RGB space is not the same as the display
> RGB space that the application and user expects the colors to be in.
> 
> So the correct thing for a color managed RIP to do to meet users (and application
> writers) expectations is to interpret the DeviceRGB values in the
> users display colorspace and convert to the actual printer devices RGB space.
> If the display space is unknown, then sRGB would typically be assumed.
> 
> An additional expectation when the RIP is a print server, is that
> it will make all the printer devices behave in a similar fashion,
> so that a document will look the same, irrespective of the device
> it is sent to.
> 
> For both these reasons the default should be to define DeviceXXX to be a
> particular colorspace. (This is why PS has the "UseCIEColor" flags, and
> why PDF has the equivalent default colour spaces option).
> 
> Naturally there should be ways of passing DeviceXXX through without
> color management for the purposes of calibration, profiling and
> application level color management.
> 

Grahame I agree with everything you said, however the problem we are trying to solve \
is ill behaved and has many, conflicting, solutions.  I doubt that we will find a \
simple, user friendly solution that would seamlessly work for everybody. Somewhere a \
more advanced mode will have to be introduced so that users with some knowledge can \
easily make the system behave as expected. This will avoid that whole Mac OS X / \
Adobe / Epson / printer driver disaster we all experienced.

Peter Skarpetis

_______________________________________________
openicc mailing list
openicc@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/openicc


[prev in list] [next in list] [prev in thread] [next in thread]