[Scribus] RGB/CMYK/digital printers / wiki woes / Was: PDF color option
Mon Nov 12 14:42:15 CET 2007
2007/11/12, Dr. Werner Popken <Werner at stuerenburg.com>:
> > If you print on your Inkjet, you should use "monitor" since the printer
> > expects RGB input. For offset printing, "printer" is the right option.
> Just for your information (in case you run into the same troubles one
> As it turned out, the printer had aquired the latest and best laser
> printing equipment there is. They used the "printer"-pdf-version and
> the output was identical to the pdf, but very different to the
> (reference) Lulu-book. In particular, reds looked very much
> brownish-orange, and they could not make it any more reddish although
> they tried hard.
This sure sounds like bad RGB to CMYK conversion.
In consequence, the printer argued that Lulu had used offset for the
> art images as they were not able to achieve identical results and
> thought Lulu must have had the same problem. Having printed just one
> copy, this seemed ridiculous for cost reasons alone, but they held
> that Lulu would do this for marketing reasons. Magnifying the print,
> it was not easy for me as a layman to tell the difference from images,
> but the text revealed without doubt that Lulu does indeed print
> digitally: the letters are pixeled in contrast to offset, where they
> are sharp.
It is of course up to anyone to use an offset press even for only one copy
but I'd leave that as the very last field of investigation in that
particular case! Let's assume Lulu went digital.
Please note that sharpness depends far more on resolution and screening
technology than on the printing process itself. For instance, first
phototypesetters would output text at 635 dpi until Linotronic and Agfa (ans
possibly others) doubled that with 1270 in the late 1980's. In the early
90's this resolution became what we still have today, 2400-2540 and up to
3600 dpi. To add some fun, these reolutions also allowed various screening
technologies to emerge, including stochastic screening, etc.
These machines where outputting films, and on offset presses this poor
resolution showed the jagged edges on all curves. It took a good eye to
notice but was noticeable notheless. Digital printers can print at various
resolutions, sometimes not the same horizontally and vertically. This
affects the final rendering. But at 1200 or 2400 dpi, it takes a magnifier
to tell the difference. The real difference between digital (toner based)
and offset lies more in the rendering of colors. CMYK-toner-based digital
printing show a larger gamut and more contrast and density overall than
CMYK-ink-based offset printing.
Lulu would most probably not use secret knowledge, but still we
> couldn't figure out what the problem was. The obvious similarity
> between monitor appearance and the printer output produced from that
> same pdf file gave the crucial clue to the puzzle. After sending the
> pdf used for Lulu (RGB), they printed just as fine. So obviously their
> digital equipment expects and needs RGB.
> As we did not know, I sent them both RGB and CMYK versions, but
> somehow they didn't even give the RGB-version a try resulting in much
> confusion. The RGB version has some 240 MB, the CMYK version way more
> than 4 GB!
Plus, with such figures, I suspect both PDFs were not made with the only
difference being settings for RGB images and the other CMYK images. There is
one color channel less in RGB so we do expect a smaller file. But such a
difference is way beyond what one would expect. So, compression methods and
settings must be different too. As I am only a user, I have no deep
knowledge of what's under the hood, hence of what else this could be but the
difference of weight between the 2 files is nothing but suspiciously huge,
imho. Does anyone else has a clue?
I guess that printing people just have to learn something new when
> digital printing arrives. Why shouldn't professional laser printers
> have drivers doing the necessary conversion just as office printers do
> without saying? Some digital printing shops do require CMYK, though.
> Most probably they have old equipment with dumb drivers.
Actualy, "dumb" RIPs. :) But then, the equipment must be pretty old. It
would be interesting to know what equipment is this and what rip is attached
As you point out, the issue might be that the rip was not set to receive
CMYK data. Or that the wrong profile was used. It is often the case where an
equipment has so many features that has to be set manually (or different
queues one has to pick for each job) that it may well be a human error and
not an equipment limitation that lead to such a bad result. But I guess this
is going to be difficult to figure out...
On top of knowing what the equipment was, it would be interesting to know
what they have actually done when you say "they tried hard" to output your
By the way: pdfsam (pdfsam.org) does a great job - I wouldn't have
> been able to do this without - thanks a lot for the hint from the list
> and for Andrea doing a great job! Last but not least: I couldn't have
> done anything without Scribus! Thank you all very much for such a
> great tool!
> Also: I made 2 flyers for 2 books with Scribus as intended, but used
> other methods for the books themselves, producing sla-files via PHP
> and doing finetuning in Scribus afterwards. Actually, for the last
> run, I produced the catalog pdf part of the big book directly from
> fpdf instead of using a sla-file interstage, making 660 pages of 796
> total, joining all with pdfsam. I guess that everybody is apt to do
> the same from scratch, so there is no need to elaborate about that.
> But maybe an entry in the "Success stories"-section of the wiki would
> be adequate. How would I do that? The create account link is
> and gives a login screen and no option for a create-account-form.
> Next: http://wiki.scribus.net/index.php/Scribus_Friendly_Print_Shops
> shows outdated information with respect to Lulu (i.e. Hard cover is
> currently limited to 6"x9") - I'd like to correct that.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the scribus