Digital BW, The Print

I'm at the point of entering densities from the 21 step wedge and am
using a scanner and PS CS to measure.  I've set the second info
palette to LAB but when moving over the each density it only shows
whole numbers and no decimal point in "L".  All the other QTR curves
are out to three decimal places.  Is there a way to get PS CS to read
that way?  I must be missing something.

Thanks,

Tom

Tom,

Photoshop rounds all L* values to integers and I don't know of any 
easy, painless way around that.  Since matte papers have L* values 
that usually go from the mid teens for shadows to 96-98 for 
highlights, and glossy papers go from single digits in shadows to 96-
98 in the highlights, you should be okay using integers.  Dividing 
the overall dynamic range by 21 still gives you about 4 L* units 
difference from step to step depending on the paper you are 
profiling.  You will still get a pretty darn good linearization.  
Spectros and densitometers can give you finer precision, but you may 
not see all that much of a difference as long as your scanner is 
reading raw data that hasn't been adjusted or converted during 
scanning.  (Auto features should be disabled when scanning).

Give it a try and see how it comes out.  

Lou

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Tom Husband" 
<thusband@s...> wrote:
> 
> I'm at the point of entering densities from the 21 step wedge and am
> using a scanner and PS CS to measure.  I've set the second info
> palette to LAB but when moving over the each density it only shows
> whole numbers and no decimal point in "L".  All the other QTR curves
> are out to three decimal places.  Is there a way to get PS CS to 
read

> that way?  I must be missing something.
> 
> Thanks,
> 
> Tom

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Louis Dina"
<lbdina@c...> wrote:
> 
> Tom,
> 
> Photoshop rounds all L* values to integers and I don't know of any 
> easy, painless way around that.  Since matte papers have L* values 
> that usually go from the mid teens for shadows to 96-98 for 
> highlights, and glossy papers go from single digits in shadows to 96-
> 98 in the highlights, you should be okay using integers.  Dividing 
> the overall dynamic range by 21 still gives you about 4 L* units 
> difference from step to step depending on the paper you are 
> profiling.  You will still get a pretty darn good linearization.  
> Spectros and densitometers can give you finer precision, but you may 
> not see all that much of a difference as long as your scanner is 
> reading raw data that hasn't been adjusted or converted during 
> scanning.  (Auto features should be disabled when scanning).
> 
> Give it a try and see how it comes out.  
> 
> Lou

Thanks Lou.  I'm pretty sure I have a raw scan so I'll just proceed
and see what it looks like.

Thanks again, your posts are always very informative.

Tom

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Tom Husband"
<thusband@s...> wrote:
> 
> I'm at the point of entering densities from the 21 step wedge and am
> using a scanner and PS CS to measure.  I've set the second info
> palette to LAB but when moving over the each density it only shows
> whole numbers and no decimal point in "L".  All the other QTR curves
> are out to three decimal places.  Is there a way to get PS CS to
read
> that way?  I must be missing something.

If you're using a Mac there's an Apple utility (included in the OS X 
installation) called "DigitalColor Meter". It can be used to read
color values in any area of the screen (including "CIE Lab" to three
decimal places). It also allows the adjustable aperture size to be
made larger than Photoshop's eyedropper maximum of 5x5.

Phil

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Tom Husband" 
<thusband@s...> wrote:
> 
> I'm at the point of entering densities from the 21 step wedge and am
> using a scanner and PS CS to measure.  I've set the second info
> palette to LAB but when moving over the each density it only shows
> whole numbers and no decimal point in "L".  All the other QTR curves
> are out to three decimal places.  Is there a way to get PS CS to read
> that way?  I must be missing something.
> 
> Thanks,
> 
> Tom

As Louis said, integer values are probably good enough.  But there are
a couple ways to coax a little more precision out of PS.   One is that
you can use RGB values -- at least they go 0-255 instead of 0-100.
Also if you have scanned 16-bit the Info palette has a 16 bit mode 
where the L values show 0-32768.

Roy

> As Louis said, integer values are probably good enough.  But there are
> a couple ways to coax a little more precision out of PS.   One is that
> you can use RGB values -- at least they go 0-255 instead of 0-100.
> Also if you have scanned 16-bit the Info palette has a 16 bit mode 
> where the L values show 0-32768.
> 
> Roy

Thanks Roy.  I seem to have something else going on when I try to
measure.  Everything is all blocked up beyond 70.  I'm going to start
over.  I can see different densities in the print out all the way to
100 but not in photoshop.  Must be my scanner.  I'm using Vuescan and
it's a raw scan with nothing added but that has to be it.  Shouldn't
be the monitor as it's calibrated.  I'll play around some more tomorrow.

Tom

> If you're using a Mac there's an Apple utility (included in the OS X 
> installation) called "DigitalColor Meter". It can be used to read
> color values in any area of the screen (including "CIE Lab" to three
> decimal places). It also allows the adjustable aperture size to be
> made larger than Photoshop's eyedropper maximum of 5x5.
> 
> Phil

That would be nice but I'm on a PC.  I need a decent scan before I can
do anything.

Thanks anyway,

Tom

Hi Roy.

I suspected QTR could accept RGB data for linearization, but wasn't 
sure.  I didn't even think about 16 bit.  Thanks for the 
clarification.

BTW, I love both your proofing techniques - ie, Gray Lab profiles, 
and creating soft proofing profiles with a 21 step wedge.  Great 
contributions that work very well.  Thanks again.

Lou

> .....integer values are probably good enough.  But there are a
> couple ways to coax a little more precision out of PS.   One is that
> you can use RGB values -- at least they go 0-255 instead of 0-100.
> Also if you have scanned 16-bit the Info palette has a 16 bit mode 
> where the L values show 0-32768.
> 
> Roy

> As Louis said, integer values are probably good enough.  But there are
> a couple ways to coax a little more precision out of PS.   One is that
> you can use RGB values -- at least they go 0-255 instead of 0-100.
> Also if you have scanned 16-bit the Info palette has a 16 bit mode 
> where the L values show 0-32768.
> 
> Roy

Dumb question alert.

OK got a decent scan showing 21 steps.  Now I'm inputting the values.
 The numbers I got from Photoshop were 12 through 100 in L value.  Do
I enter those whole numbers?  That didn't work.  Nothing was created
when I hit the create curve button. Do I somehow use the Table of
Ideal Densities from the tutorial by Tom Moore and Roy?  Somehow I
think my density numbers are suspect or I have to convert them to
something else.  The values on the UT2-EEnhMatte-cool curve are 0.047
through 1.590.  Sorry but I'm really lost here.

Tom

Hi Tom,

You have to enter them in descending order, from high to low (left to
right), whole numbers are ok.

--nick

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Tom Husband"
<thusband@s...> wrote:

> 
> OK got a decent scan showing 21 steps.  Now I'm inputting the values.
>  The numbers I got from Photoshop were 12 through 100 in L value.  Do
> I enter those whole numbers?  That didn't work.  Nothing was created
> when I hit the create curve button. Do I somehow use the Table of
> Ideal Densities from the tutorial by Tom Moore and Roy?  Somehow I
> think my density numbers are suspect or I have to convert them to
> something else.  The values on the UT2-EEnhMatte-cool curve are 0.047
> through 1.590.  Sorry but I'm really lost here.
> 
> Tom

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Nick H. Nugent"
<nghin@p...> wrote:
> 
> Hi Tom,
> 
> You have to enter them in descending order, from high to low (left to
> right), whole numbers are ok.
> 
> --nick

I've done that but when I hit the create button the graphs that come
up are empty.  Both the linearization and the Ink curves.

Tom