Digital BW, The Print

Hi John,

The embedded profiles and working spaces can be confusing.  Lots of things
happen behind the scenes so I'll try to clear up some of it.

The idea with all the stepwedges is to have the "raw" numbers sent to the
printer drivers -- i.e. no profile conversions should be performed.
The trouble is that CM (color management) happens just about everywhere.

All the step wedges have the correct "raw" numbers in them.  The intension
is to not "use" the embedded profile in any of them -- (they were all GG 2.2
or untagged but it looks like I managed to get the 21-random assigned to Gray Lab).
The important thing when you read a stepwedge into PS is to NOT do a profile
conversion.  Profile assignment is OK, in fact it will always have an explicit
or implicit profile no matter what.  An untagged file is always treated as the 
working grayspace.    There are only two ways you can tell what data is really
there -- the Info palette using K or the histogram. I think we all have a
tendency to look at RGB or Lab values to see the full 8 bits (ie 0-255 instead
of just 0-100).  However the RGB values are the profile CONVERSION from the
grayspace of the data to the default RGB working space.  If these two spaces
have a different gamma or transfer curve the numbers are not the same.
So the K values of a stepwedge ought always to be 0,5,10,..50...95,100. and
the histogram will also look evenly spaced.  Looking at RGB or Lab is dependent
on the working spaces.

The second important part is during the Print with Preview dialog page.  Here
the Print Space must be always be Same As Source when making stepwedges
for linearization or icc profiles.  This Same as Source ensures that the CM
does not do a conversion -- note that it really doesn't matter what the
source profile is because the idea is to just pass the raw numbers through.

If you want to get a RGB version of the stepwedge you have to be careful how
you do it.  The important thing is that the gray source profile and the RGB
destination profile (working space) have the same gamma or xfer function.
Some possible legal pairs are:
Gray Gamma 2.2 -- Adobe RGB
Gray Gamma 1.8 -- Apple RGB
QTR Gray Lab      -- QTR RGB Lab

If you tend to go back and forth Gray <--> RGB and like to look at both K
and RGB in the Info Palette, its probably best to use pairs like above.

> Side questions
> A gray pixel value of K=50% and L*=50 should be 118 when converted to a RGB
> gamma 2.2 space, correct?
Yes that is the conversion:
     K=50 in gray lab is the same as 
     L=50 is same as 
     RGB=118 in GG 2.2 or AdobeRGB

> QTR – (RGB and Gray) Lab space gamma is 2.45, but only at L*=50, correct?
The Lab spaces don't have a real gamma -- its a different mathematical function.


It takes a while to see all this going on underneath but hopefully
this is a start.

Roy


--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "John Moody" 
<moodymz3@y...> wrote:
> Nobody has any comment on this?
> 
> John
> 
> -----Original Message-----
> From: DigitalBlackandWhiteThePrint@yahoogroups.com
> [mailto:DigitalBlackandWhiteThePrint@yahoogroups.com]On Behalf Of John Moody
> Sent: Tuesday, July 26, 2005 5:05 PM
> To: DigitalBlackandWhiteThePrint
> Subject: [Digital BW] Step-21 files and embedded profile space
> 
> All of the step wedges in the 2.3 version of QTR has a gray gamma 2.2
> profile embedded, _except_ the step-21-random file, which has QTR - Gray Lab
> profile embedded.
> 
> It seems to me they should have no profile, or it should be QTR – Gray Lab
> if you want to use them with the generate icc tool.  Is that right?
> Furthermore, if you want to convert them to RGB to profile something other
> than QTR, such as BO printing through the Epson driver, they should not be
> converted to your RGB color space, but converted by "convert to profile,
> QTR – RGB Gray Lab".  Do I have that right?  This would give pixel values
> for the 50 and 95 patch of 128 and 13, rather than 118 and 24 that a gamma
> 2.2 RGB space would have.
> 
> Here is the question:  What values should an _untagged_ 21-step wedge have,
> when printed for use with the QTR-generate icc tool?
> Should it go up linearly by 255/20?
> 
> Side questions
> A gray pixel value of K=50% and L*=50 should be 118 when converted to a RGB
> gamma 2.2 space, correct?
> QTR – (RGB and Gray) Lab space gamma is 2.45, but only at L*=50, correct?
> 
> Best regards,
> John Moody
>