Digital BW, The Print

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, Steve Kale <stevekale@=
b...> 
wrote:
> Roy ­ thanks a lot.  There is a lot for me to think about here but I am
> afraid it is now going to be the end of next week before I can get back i=
nto
> it fully.  I must admit I have not yet made it into the Sepia world and a=
m
> thinking about this purely from the perspective of my 2100 loaded with
> conventional UC inks and trying to get a B&W image that matches what I se=
e
> on the screen.  In particular, I was thinking of the warm curves where on=
ly
> two inks are used K and LK.   Your points re a Sepia toner have made me t=
ake
> a step back for further thought.   What of this hypothetical case: I load=
 my
> printer with two shades of brown ink ­ dark brown in the K position and
> light brown in the LK position. Presumably QTR will faithfully print my g=
rey
> scale image in shades of brown brown (with a nicely linearized curve whic=
h I
> forgot is linearized only with density values rather than RGB
> measurements....).  Now if you I were to do an ICC profile according to y=
our
> instructions I could preview the image in its printed state.  But if I di=
d
> not want a brown image but rather a grey image then everything would be o=
ut
> of gamut because there is no combination of brown that can make black (fo=
r
> example).   A conventional use of ICC profiles (a la the colour world) wo=
uld
> show this.  I guess this is where the gap in the circle is with my earlie=
r
> thinking and I understand your points ­ it¹s not that I started with a br=
own
> image and wanted to check how it would rendered in the printer¹s brown
> space.  Presumably, though, it has some interesting eventual practicality=

> for a workflow which stays in RGB.  If one could profile the full colour
> space capable of a particular printer and ink setup ­ be it just capable =
of
> pure dead neutral shades of grey, warm shades of grey, or one much more
> complex that introduces another element(s) of colour tint via a toner(s) =
­
> then one could work in RBG, introduce a tint or not, and manipulate the
> image in RBG to increase or decrease the tint with the knowledge that
> WYSIWYG.  
> 
> So two last questions for now if you will permit me.    In the current se=
tup
> described by Carl and yourself, if I wish to produce a sepia-toned image,=

> can I alter and preview real time the amount of sepia that will be produc=
ed?

If you already have a way to produce a sepia image -- then and only then
it's possible to have a ICC proof to preview what you'll get.  But that ICC=

profile plays no role and can't help you actually print that tone.
I guess this is the main difference with color.  With color you also use th=
e
same ICC profile to help you get the print.   The reason it works for color=

is that both the display and the print driver take RGB values and those RGB=

values do influence the color of the print.

> I suspect that all I can see is a version of the grey image sepia-toned
> given one ink/curve setup and that if I want to see a more-sepia version =
I
> would have to construct and apply another ICC profile for a different
> toner/ink mix.  In a simpler case, if I can preview it  with the warm-EEM=

> curve or the cool-EEM curve but if I want to see it with 50:50 then
> presumably another ICC profile must be produced.....The last question:  i=
n
> order to see the sepia version doesn¹t my workspace have to be a colour o=
ne
> or is this taken care of automatically by PS when I proof?

PS takes care of it all.   The 50/50 blend is a feature of QTR that has no
counterpart in ICC profiles, so yes you'd have to print out a 50/50 wedge a=
nd
create a proof profile for it.

Actually the blend issue is an interesting topic and this may actually help=

in understanding profile usage.

It's a little work but you can make two layers -- one from each profile --
and vary the opacity of the top layer to simulate QTR blending.   You 
can't have two profiles in one file so you need to do the effect of each
profile and get back to a common profile.

1) you should have settings: gray gamma 2.2 and Adobe RGB
2) take Grayscale and Duplicate it, conver to RGB and duplicate again.
    -- now you have two Adobe RGB files with gray image.
3) take one RGB file and Assign to Profile:  ICC Profile 1 (warm) and
then Convert to Profile: Adobe RGB  -- Relative Colorimetric and Black Poin=
t set.
4) take the other and Assign ICC Profile 2 (cold) and Convert to Adobe RGB =
too
5) now Select All, Copy one image, and Paste onto second.

You now have two layers and vary the opacity.  Also look at the RGB
values of each layer separately and combined.

If you follow all the changes -- number wise (RGB values) and color
wise on the display, it may help see what goes on behind the scene.

Roy


> 
> Cheers 
> 
> Steve
>