Re: [Gimp-developer] suggestion for new versions of GIMP

["gespertino gmail com" <gespertino gmail com>]

2011/11/27 Bogdan Szczurek <thebodzio gmail com>:
> In my opinion, it's not only a habit. I think better thing to say would be:
> print industry is more often _tolerating_ RGB. I think it's because of quite
> reasonable conversion profiles to CMYK and because more, and more often,
> different content creators (clients) provide materials in RGB. I think they
> do so partly because Adobe is trying to convince people that they can create
> a document in e.g. InDy and publish the same document as an ebook in epub,
> website, flash movie and whatnot by one click. All of the same quality. I
> say: BS :). Why? Because of some lowest common denominator, about which
> later :). Document prepared for print will be most probably impaired when
> forcibly cast to e.g. web space :).

Last year I attended to the most important trade show of the
argentinian print industry invited by the most renowned institution of
pre-press training in Argentina, and they didn't seem to "tolerate"
Late Binding as the dumb little brother.
Late binding is an alternative workflow that is gaining acceptance, as
valid as early binding (of course, with its advantages and
disadvantages, just like the other)

> Try to create webpage layout in
> Photoshop, use it's facilities to export it to HTML + images and ask a good
> webdesigner what he thinks about this code :). Granted it'll look right in
> browser, but it won't change a fact that the code is a total trash. But I
> digress :)

Why would I do that? I use GIMP :-p
Seriously: any decent web designer wouldn't use PS and slices. That's
indeed a set of tools for the lowest common denominator.

>
> I agree, most of the time, relying on good RGB profile for conversion to
> CMYK is sufficient. But there are times when it's better to modify CMYK
> values manually (of course in the same workspace as printhouse's). Example:
> in dark areas of photos one could wish to add a bit of warm tint to places
> already printed with a 100% black. You _can_ do it by creating your own
> profile but you'd have to test it with your printer. I also had to cope, a
> couple of times, with some blues or reds in photos converted from RGB to
> CMYK. In RGB – nice, in CMYK – awful. Who's going to guarantee my photos
> won't be destroyed that way during conversion? Nobody :). It works _most_ of
> the time, _not_always_. It can crash _unexpectedly_ on some more or less
> significant subtleties.

Who's going to guarantee that the CMYK combination you chose in your
software will look like you expect on paper? Soft-proofing, which is
RGB.
Papers, inks and press settings vary from provider to provider and
that's why serious printshops create their own profiles.
There's no way you can know for sure the printed appearance of a
particular CMYK combination without a printed proof.
Additionally, tweaking CMYK separatios manually can result in
combinations that go beyond the profiles TAC values.
I wonder how many designers know this and understand what can go wrong
with CMYK.
in my personal experience most of them don't know, and just choose
CMYK because "it's for print", and end up stripping profiles because
the file is smaller and mixing 100% CMYK because it looks truly black.
Don't underestimate people who work with late binding and don't
overestimate people who choose CMYK ;-)

> That's why I stick with CMYK for print, not because of a habit :).

I'm not saying it's an invalid choice. If it's an informed choice and
it works for you, great.
I chose working in RGB and I get what I want. I get the amount of
control I need and my clients are pleased.
Works for me!™

>> Working with early binding (CMYK from the beginning) isn't exactly a
>> good idea in this age of cross-media publishing.
>
> IMHO cross-media publishing have as many good as bad sides. Each media has
> it's own specificity. Cross-media means that one have to find a _lowest_
> common denominator of multiple (often greatly differing) presentation
> technologies. This also means that some potential sacrifices have to made in
> the area of color reproduction.

gamut-wise, choosing RGB doesn't necessarily mean "the lowest common
denominator". Some generic CMYK profiles have smaller gamuts than
sRGB!
I'd say that going to a generic CMYK at the beginning sacrifices
color, hence you're working with the lowest common denominator.
Keeping RGB using colors that fit inside the gamuts of all the devices
involved for the essential colors (for instance branding colors) give
you a good balance of color preservation and color reliability.
I don't think it's necessary to sacrifice the color richness of a
photography to the lowest common denominator for every intended
output, and afaik that's what you do when switch your assets to CMYK
at the beginning of the pipe.

>> I switched to intermediate/late binding two or three years ago, and if
>> I see a difference, it favors RGB.
>
> If color management process would be observed you wouldn't see any
> difference ;). Yes… I know… theorethically ;)
>
>> Mostly in CTPs (my print provider recently added a hybrid AM/FM CTP
>> and RGB performs much better than CMYK there).
>
> Then it's something wrong with color management here :).

A little clarification: Both were almost identical color-wise, but the
definition of FM screening was better in RGB.
This was the test file (interesting test btw)
http://ubuntuone.com/p/XX2/

When I said "it favors RGB" I meant when using the same file with
different providers.
I did some tests sendind the same files to different providers both in
RGB and CMYK to see what happened. Comparing the samples I find RGB
more even accross providers.

>> Even the decompose command will be useful :-p
>
> Even better :) – obsolete.

Ha! :D