Digital BW, The Print

Steve,

so have you - been awfully generous! I hope everyone out there 
appreciates the time and effort you're putting in here trying to 
explain the issue.

Thanks! I

Also, why is it so hard to understand - or accept - that the way 
photographers did it 10 - 20 years ago eye balling their prints was a 
totally different workflow and process? Wet darkroom and digital 
"darkroom" are two very different animals!

Come on folks - we're not doing wet darkroom processes here - spend the 
time learning and use the routines folks like Steve and Paul and others 
are teaching/explaining here - on their own time - and you all will be 
heaps better off.

The day I started doing calibrated workflow (Monaco EZ Color with 
Monaco Optix XR PRO - inexpensive - $350 now for the whole workflow - 
monitors, printers, papers) was the beginning of a new era for me - and 
my clients - for the first time I was/am able to get what you see is 
what you get from any one of my monitors/macs/paper to the printers I 
am using! No more spending hours trying to match the output to the 
monitor, wasting time, paper and ink - and when I FTP or send a client 
a CD/DVD with an image(s) I know they will be able to use it/them and 
not come back unhappily telling me the colors are way off.

Christer


			Christer, AKA Christer Rosewell
"It's the artist's job to accomplish two things-to stir the emotions of 
the viewer
	  and to lay bare the soul of his subject." Jousuf Karsh
      		  Member EP (Editorial Photographers)
			  http://www.ChristerArt.com
		  	   3.6 million visitors to date..


On Apr 9, 2005, at 8:27 AM, 
DigitalBlackandWhiteThePrint@yahoogroups.com wrote:

> They
> have been awfully generous with their time thus far - if we ask them 
> nicely
> perhaps they can help further.
>
> Steve

[Non-text portions of this message have been removed]

I think Roy's earlier post about how various spaces display may be the
most relevant to you. You state you are linearizing your processes to
LAB, but the shadows on your monitor in 2.2 are compressed compared to
your print.
I assume a step wedge printed with one of your processes linearized to
lab separates to your satisfaction.
Gamma 2.2 will not display those values the same at all.
It may be worth your time to download Roy's gray lab working space,
install it, open a step wedge, assign (don't convert) Roy's space to
it, and see if you get a better match.
It won't address the differing dmin/dmax dilemma, but the progression
of values may be much better, particularly in the areas you now have
trouble seeing on the monitor.
Tyler

Tyler,

> I think Roy's earlier post about how various spaces display may be the
> most relevant to you.

Yes, I noted it.

> You state you are linearizing your processes to
> LAB, but the shadows on your monitor in 2.2 are compressed compared to
> your print.

Yes.


> I assume a step wedge printed with one of your processes linearized to
> lab separates to your satisfaction.

Yes, but that is how I designed them.  I think there may be advantages to
that set of target densities, so I intend to continue that for the
foreseeable future.

> Gamma 2.2 will not display those values the same at all.

True.  A custom dot gain curve like you explained years ago is what I've
used for some time to get the monitor to match the print.  There appears to
be a large range of solutions that others have employed to get to what I
think most of us see as one of the top goals -- to get a reasonable match
between the monitor and print.


> It may be worth your time to download Roy's gray lab working space,
> install it, open a step wedge, assign (don't convert) Roy's space to
> it, and see if you get a better match.

Oh yes, I've done that, and it doesn't change the results.

Contrary to the majority of frequent posters here, I'm not sure B&W
photographers need to be burdened with much concern about color spaces, etc.
One attraction of B&W is the lack of concern for color and all its
complexities.  Matching print densities to the monitor's relative
"densities" does not, I believe, have to be very complex at all.  A simple
curve (like your custom dot gain curve) or, even better, a set of sliders,
may be all we B&W guys need and need to know about.  

In a sense, all I suggested with my original post was that rather than a
custom dot gain curve to softproof the monitor, the same thing might be done
at the printing end of the process instead.  In my workflow it could be part
of a curves set with the printing curve.  That way the user would just drag
a curves set from a small carrier file to the file to be printed.  That
seems like a fairly easy workflow.

I suspect that same curve could also do what the linearization overlays in
the rips do -- make profiles for one paper useable for another similar
paper.  That would be a major step is the flexibility of the workflow I use.


Paul
www.PaulRoark.com

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Paul Roark"
<paul.roark@v...> wrote:
...
> > I assume a step wedge printed with one of your processes linearized to
> > lab separates to your satisfaction.
> 
> Yes, but that is how I designed them.  I think there may be
advantages to
> that set of target densities, so I intend to continue that for the
> foreseeable future.

Yes, I agree. But here's the dilamma-
You designed them to print nicely and evenly spaced on paper, from a
2.2 gamma file, NOT to match what you see on the monitor or reproduce
a 2.2 gamma progression. Had they been designed to print with a gamma
2.2 progression of tones, you would have seen on paper a similar
compression of shadows. I'm not saying that would have been the way to
go at all, or that there is anything wrong with your appraoch, I'm
only trying to point out the disconnect.
For example, if the entire workflow was designed for a gamma 2.2
working space, and a print process developed to present that same
progression, then one would naturally edit one's tones accordingly
(probably openning shadows), and they would print accordinly.

So it seems to me that the simplest solution to the mismatch is the
working space, since the aproach to linearizing the proggression of
tones on paper makes common sense as is.
> 
> > Gamma 2.2 will not display those values the same at all.
> 
> True.  A custom dot gain curve like you explained years ago is what I've
> used for some time to get the monitor to match the print.

That remains viable, but as you no doubt have run into, more curve
points would be nice to nail it.

> > It may be worth your time to download Roy's gray lab working space,
> > install it, open a step wedge, assign (don't convert) Roy's space to
> > it, and see if you get a better match.
> 
> Oh yes, I've done that, and it doesn't change the results.

I'm very suprised actually, with a 21 step file open, if I assign 2.2
I can barely distinguish 95% from 100%, if I assign Roy's lab, it
opens right up. It's a dramatic difference.
I wonder what could be different there, and can see why you might have
discounted it if you don't see what I see here.

> Contrary to the majority of frequent posters here, I'm not sure B&W
> photographers need to be burdened with much concern about color
spaces, etc.

I just don't see any way around it. Those settings and how they are
used effect everything. Unfortunately color management at the OS
level, and in the apps, has forced us to deal with it. It can't be
turned off, even "do not color manage this bla bla" is misleading,
there are still profile and setting dependant conversions going on
behind the scenes. There is no "none" to select in the profile
selction list at the OS level, etc..
This evolution has made things more difficult for everyone, but also
much better than where we were before. I can do things now that were
impossible before these controls, both B&W and color.

I guess it depends on how far those photographers want to pursue the
technology behind their craft. That may sound condescending and I
don't mean that at all. I know several amazing photographer/printers
who never got on top of their craft and consistantly come out of the
darkroom with amazing results. I think they are exceptions though, I
can't think that way, I have to get to the bottom of it.

I was not really answering your original post, but was thinking back
more to when you posted about disappointing monitor calibration results.
If you come up with some easily implemented aproach to monitor/print
matching for B&W I think that's great. The main point of Carolyn's
earlier post was that there are many many ways to do things.

But I'm still wondering why you don't see a difference with that lab
space...
How about this.. use the soft proof method above as you have been, but
instead of a custom curve, load Roy's gray space. Keep "preserve"
checked as before.

Of course, if you are after a solution to use in other apps, possibly
even some not icc aware and/or without soft proof features, some
simpler solution is necessary, as you suggest. I believe you said you
got a good match with your curve method, and as implied by Carolyn, if
it works...

Tyler

> From: Tyler Boley <tyler@...>

> 
> I'm very suprised actually, with a 21 step file open, if I assign 2.2
> I can barely distinguish 95% from 100%, if I assign Roy's lab, it
> opens right up. It's a dramatic difference.
> I wonder what could be different there, and can see why you might have
> discounted it if you don't see what I see here.
> 


Look at the respective L figures of the steps.  Assign a different profile
and you are assigning a completely different luminance to the particular
step.

of course, that's what I was trying to get at without the tech stuff...
Should have worked.

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, Steve Kale
<stevekale@b...> wrote:
> 
> 
> 
> > From: Tyler Boley <tyler@t...>
> 
> > 
> > I'm very suprised actually, with a 21 step file open, if I assign 2.2
> > I can barely distinguish 95% from 100%, if I assign Roy's lab, it
> > opens right up. It's a dramatic difference.
> > I wonder what could be different there, and can see why you might have
> > discounted it if you don't see what I see here.
> > 
> 
> 
> Look at the respective L figures of the steps.  Assign a different
profile
> and you are assigning a completely different luminance to the particular
> step.

In other words, the difference between any two greyscale workspaces is just
their gamma....LAB simply has a gamma that leads to nice equal increments in
luminance.  Gray Gamma 2.2 doesn't.

> From: Tyler Boley <tyler@...>
 
> 
> 
> of course, that's what I was trying to get at without the tech stuff...
> Should have worked.
>

Actually that's a bit circular.  LAB is designed with a definition of L
(named luminance) and a gamma such that L increases uniformly as we step
from pure white to pure black.  When we look at L values, for say Gray Gamma
2.2, we simply find that they don't have the same characteristic.  But in
the end, its simply all about gamma...ahhh contrast!

> From: Steve Kale <stevekale@...>



> LAB simply has a gamma that leads to nice equal increments in
> luminance.

Within the context of the exchange Paul and I were having, I can't
place the relevance of this.
But hey, have at it!
A little wine with that trout?
Tyler

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, Steve Kale
<stevekale@b...> wrote:
> Actually that's a bit circular.  LAB is designed with a definition of L
> (named luminance) and a gamma such that L increases uniformly as we step
> from pure white to pure black.  When we look at L values, for say
Gray Gamma
> 2.2, we simply find that they don't have the same characteristic. 
But in

> the end, its simply all about gamma...ahhh contrast!
> 
> 
> > From: Steve Kale <stevekale@b...>
> 
> 
> 
> > LAB simply has a gamma that leads to nice equal increments in
> > luminance.

> From: Steve Kale [mailto:stevekale@...]
>
> In other words, the difference between any two greyscale
> workspaces is just
> their gamma....LAB simply has a gamma that leads to nice equal
> increments in luminance.  Gray Gamma 2.2 doesn't.

Technically, LAB doesn't have a gamma. Neither does sRGB. A gamma is an
exponent that defines a simple curve. Both Lab and sRGB have a more complex
shape that is linear at the dark end. The eye/brain doesn't respond either
linearly or exponentially throughout its range, so it's necessary to use the
more complex shape in order to provide something resembling uniform steps.

However, "gamma" is starting to come to mean nothing more than the midscale
gray level. Gamma 2.2 means that a midscale gray is 0.5 to the 2.2 power in
luminance. The higher the gamma, the darker the gray. I wish these terms
wouldn't wander in meaning, but unless someone has a better term, I guess
we'll loosely refer to it as "gamma".

--

Ciao,               Paul D. DeRocco
Paul                mailto:pderocco@...

Well it does have a gamma.  It's just not linear ie can't be expressed by a
single number.

> From: "Paul D. DeRocco" <pderocco@...>

> 
> Technically, LAB doesn't have a gamma. Neither does sRGB. A gamma is an
> exponent that defines a simple curve. Both Lab and sRGB have a more complex
> shape that is linear at the dark end. The eye/brain doesn't respond either
> linearly or exponentially throughout its range, so it's necessary to use the
> more complex shape in order to provide something resembling uniform steps.
> 
> However, "gamma" is starting to come to mean nothing more than the midscale
> gray level. Gamma 2.2 means that a midscale gray is 0.5 to the 2.2 power in
> luminance. The higher the gamma, the darker the gray. I wish these terms
> wouldn't wander in meaning, but unless someone has a better term, I guess
> we'll loosely refer to it as "gamma".
>

> From: Steve Kale [mailto:stevekale@...]
>
> Well it does have a gamma.  It's just not linear ie can't be
> expressed by a single number.

But gamma _is_ a single number. "Gamma" is the greek letter given to the
exponent in the formula that relates numbers or voltages to luminance: the
luminance is the numeric value or voltage, expressed as a fraction from 0 to
1, raised to the gamma power, multiplied by the maximum luminance. To say
that sRGB has a "gamma" of 2.2 is only an approximation.

--

Ciao,               Paul D. DeRocco
Paul                mailto:pderocco@...

Also why Clayton's methodology works nicely and isn't inappropriate.
He's simply using one of those already worked out, highly technical
results of industrial wizardry that someone else took the trouble and
$ to pursue first. Kinda like buying sliced bread and thinking it's
really ok to eat it instead of baking your own.

He also paid for it along with Pshop:)

Regards
Duane

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Tyler Boley"
<tyler@t...> wrote:
> 
> of course, that's what I was trying to get at without the tech stuff...
> Should have worked.
> 
> --- In DigitalBlackandWhiteThePrint@yahoogroups.com, Steve Kale
> <stevekale@b...> wrote:
> > 
> > 
> > 
> > > From: Tyler Boley <tyler@t...>
> > 
> > > 
> > > I'm very suprised actually, with a 21 step file open, if I
assign 2.2
> > > I can barely distinguish 95% from 100%, if I assign Roy's lab, it
> > > opens right up. It's a dramatic difference.
> > > I wonder what could be different there, and can see why you
might have
> > > discounted it if you don't see what I see here.
> > > 
> > 
> > 
> > Look at the respective L figures of the steps.  Assign a different
> profile
> > and you are assigning a completely different luminance to the
particular
> > step.