Re: CSS overprinting

On Tue, 2013-06-18 at 17:39 -0400, Lea Verou wrote:

> I’m not sure how overprint could be controlled, since it could be for
> the entire element, or just the text etc. It looks more like a
> blending mode.

+1 to thinking about this.

Related, there's also undercolour removal --

Typically an offset lithography press (the sort used for most commercial
bulk printing) has a limit on the amount of ink at any one place. Often
a given CMYK value with cyan, magenta and yellow all non-zero but black
zero can be represented (or approximated) by adding black and reducing
the other inks in a process known as undercolour removal.

There's also trapping, or making narrow corners wider to avoid getting
blobs of ink caught in them. The "N" of Franklin Gothic has a clear
example of allowing for this, giving it a very different look on
displays than on printed paper.

These things are usually done by specialist "preflight" software, or by
e.g. photoshop plugins. Other things not handled currently by CSS
include
. specifying bleed and crop
. placing colour registration marks outside the crop area
. imposition (folding printed sheets into pages before the cropper uses
the crop marks to cut the edges off so you can open the book)
. binding (long edge or short edge? hardback or paperback?...)
. spot colours - it's common to print with e.g. CMYK + one or two others
for highlights, e.g. a particular orange.
. duotones and other non-CMYK non-RGB models, often with different
colour gamuts
. fold-outs, changing paper size / orientation part-way through a job
(even multi-tray office printers can do this)
. folding, envelope stuffing (e.g. for mass mailers, bank statement etc)

Job Definition Format (JDF) does some of this, and you can do manual
overprint in SVG of course, with more or less effort.

>  However, if we add a blending mode for it, what will it do for RGB?
> I'm not sure if overprinting is even a thing in RGB.
I don't think so. Light-based displays don't have the same blending
characteristics :)

This sort of issue - the role of CSS in printing - is definitely a topic
for the Workshop we're holding in Paris this September.

I'm eager to see solutions for overprinting, as well as for all the
other issues and I don't want to suggest "don't do one unless they are
all done."  I still think the CSS WG needs to have some separate task
forces, or maybe just interest groups and/or community groups, to split
off some of this necessary work.

In the meantime, note that because of out-of-line rendering,
overprinting is asynchronous to the element tree. I agree it's like
compositing, but of the whole rendered page.

Liam

-- 
Liam Quin - XML Activity Lead, W3C, http://www.w3.org/People/Quin/
Pictures from old books: http://fromoldbooks.org/
Ankh: irc.sorcery.net irc.gnome.org freenode/#xml

Received on Tuesday, 18 June 2013 22:09:11 UTC