R help - Nov 2020 - RGB -> CYMK, with consistent colors

All,

I used the very useful colorspace package for the plots in my book
(pdf available here): http://knosof.co.uk/ESEUR/

The color makes the plots look great, on screen.
To get lots printed, the printer requires converting the images to use cmyk
(a common requirement for larger  printers, I'm told).  See page 11 here:
https://www.ingramspark.com/hubfs/downloads/file-creation-guide.pdf

No problem, the script below uses ghostscript to achieve this:

gs -o ESEUR-cmyk.pdf \
     -sDEVICE=pdfwrite \
     -sProcessColorModel=DeviceCMYK \
     -sColorConversionStrategy=CMYK \
     -sColorConversionStrategyForImages=CMYK \
      ESEUR.pdf

the problem is that the converted colors don't look nearly as
good.  For instance the cyan now looks blue, and prints as pure blue.

I can regenerate the images, and explicitly specify cmyk.  But using:

pdf.options(colormodel="cymk")

does not change anything.  The colors look remarkably similar to
those produced via the ghostview route.

I have been looking at color profiles and trying to find a
way of modifying an ICC profile (yes, it looks difficult).

Does anybody have any ideas for producing cmyk images that have
the same (or close enough) look as the RGB?

On Sun, 29 Nov 2020 at 14:26, Derek Jones <gmane at knosof.co.uk> wrote:

[...]
> I can regenerate the images, and explicitly specify cmyk.  But using:
>
> pdf.options(colormodel="cymk")
>
> does not change anything.  The colors look remarkably similar to
> those produced via the ghostview route.
[...]
> Does anybody have any ideas for producing cmyk images that have
> the same (or close enough) look as the RGB?
Have you tried printed a few pages in CMYK?

A monitor is based on mixing light using Red-Green-Blue. So it is not
possible for the monitor to show
CMYK which must be printed on paper to view correctly.

Regards
Martin

	[[alternative HTML version deleted]]

I had a long discussion on this topic with the Springer production
group when my book was in production.
Statistical Analysis and Data Display: An Intermediate Course with
Examples in R, second edition
Richard M. Heiberger and Burt Holland.
https://www.springer.com/gp/book/9781493921218

As I now understand it, the physical inks used in printing cannot
produce the same range of colors as the computer screens.
The issue is not the notation, but rather the underlying technology.

I chose to let the publisher make the conversion.  I looked at the set
of pdfs for your book, and to
anyone else but you, I think they would look fine in slightly different colors.

Rich

On Sun, Nov 29, 2020 at 8:26 AM Derek Jones <gmane at knosof.co.uk>
wrote:
>
> All,
>
> I used the very useful colorspace package for the plots in my book
> (pdf available here): http://knosof.co.uk/ESEUR/
>
> The color makes the plots look great, on screen.
> To get lots printed, the printer requires converting the images to use cmyk
> (a common requirement for larger  printers, I'm told).  See page 11
here:
> https://www.ingramspark.com/hubfs/downloads/file-creation-guide.pdf
>
> No problem, the script below uses ghostscript to achieve this:
>
> gs -o ESEUR-cmyk.pdf \
>      -sDEVICE=pdfwrite \
>      -sProcessColorModel=DeviceCMYK \
>      -sColorConversionStrategy=CMYK \
>      -sColorConversionStrategyForImages=CMYK \
>       ESEUR.pdf
>
> the problem is that the converted colors don't look nearly as
> good.  For instance the cyan now looks blue, and prints as pure blue.
>
> I can regenerate the images, and explicitly specify cmyk.  But using:
>
> pdf.options(colormodel="cymk")
>
> does not change anything.  The colors look remarkably similar to
> those produced via the ghostview route.
>
> I have been looking at color profiles and trying to find a
> way of modifying an ICC profile (yes, it looks difficult).
>
> Does anybody have any ideas for producing cmyk images that have
> the same (or close enough) look as the RGB?
>
> ______________________________________________
> R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.