Yahoo Groups archive

Digital BW, The Print

Index last updated: 2026-04-28 22:56 UTC

Thread

Re: [Digital BW] Converting Photoshop Curves into ICC profiles - Can This be Done?

Re: [Digital BW] Converting Photoshop Curves into ICC profiles - Can This be Done?

2003-03-08 by Editor P.O.V. Image Service

davidspielman2003 wrote:

>Originally posted this question on the Color Theory group - got no
>answers! Will try here.
>~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>Hello Group,
>
>Can anyone tell me if there is any way (3ed party PC software) that
>one can use to convert the curve functions in Photoshop (Transfer
>Function - *.ATF, Curves -*.ACV, Map Settings - *.AMP) into an ICC
>profile (.ICM, or *.ICC) ? 
>
>
>  
>
Theoretically you can use ColorVision's Doctor Pro to accomplish this...

You'd apply the the curve to the correct profile and generate a new 
profile..

The caveat is that your images will look strange in softproof mode..  

However, the new profile, should, in theory, properly convert images for 
the printer..

The only real advantage in thisapproach would be that you could do all 
your adjustments in the plain ol generic Pshop working profile and  then 
outpput to the new profile..

Hmmm.. I'll try and find time to try this out over the weekend with a 
small print and my existing VM-S curves...
Keith

 

"Just some guy," and caretaker of the Multiverse's largest EPSON printer 
User Community (highly recommended by Vogon Poets and MegaDodo 
Publications), at:

http://groups.yahoo.com/group/EPSONx7x_Printers/
 
"For the rest of you out there, the secret is to bang the rocks together 
guys"

 




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

Re: [Digital BW] Converting Photoshop Curves into ICC profiles - Can This be Done?

2003-03-08 by Ernst Dinkla

----- Original Message -----
Show quoted textHide quoted text
From: "Editor P.O.V. Image Service" <editor@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Saturday, March 08, 2003 10:05 AM
Subject: Re: [Digital BW] Converting Photoshop Curves into ICC
profiles - Can This be Done?


>
>
> davidspielman2003 wrote:
>
> >Originally posted this question on the Color Theory group -
got no
> >answers! Will try here.
>
>~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
~~~~~
> >Hello Group,
> >
> >Can anyone tell me if there is any way (3ed party PC software)
that
> >one can use to convert the curve functions in Photoshop
(Transfer
> >Function - *.ATF, Curves -*.ACV, Map Settings - *.AMP) into an
ICC
> >profile (.ICM, or *.ICC) ?
> >
> >
> >
> >
> Theoretically you can use ColorVision's Doctor Pro to
accomplish this...
>
> You'd apply the the curve to the correct profile and generate a
new
> profile..
>
> The caveat is that your images will look strange in softproof
mode..
>
> However, the new profile, should, in theory, properly convert
images for
> the printer..
>
> The only real advantage in thisapproach would be that you could
do all
> your adjustments in the plain ol generic Pshop working profile
and  then
> outpput to the new profile..
>
> Hmmm.. I'll try and find time to try this out over the weekend
with a
> small print and my existing VM-S curves...
> Keith

The other advantage would be the use of "curves" in
drivers/programs that are ICM compatible but don't have curve
adjustments.

It is nice to have profiles based on curves but it would be even
better to include tone corrections of the actual output in the
profile and get rid of any colour info.

Ernst

Re: [Digital BW] Converting Photoshop Curves into ICC profiles - Can This be Done?

2003-03-12 by Ernst Dinkla

Keith, you wrote:

> Theoretically you can use ColorVision's Doctor Pro to
accomplish this...
>
> You'd apply the the curve to the correct profile and generate a
new
> profile..
>
> The caveat is that your images will look strange in softproof
mode..
>
> However, the new profile, should, in theory, properly convert
images for
> the printer..
>
> The only real advantage in thisapproach would be that you could
do all
> your adjustments in the plain ol generic Pshop working profile
and  then
> outpput to the new profile..
>
> Hmmm.. I'll try and find time to try this out over the weekend
with a
> small print and my existing VM-S curves...


Did you try it and did it work ?

Ernst

Re: [Digital BW] Converting Photoshop Curves into ICC profiles - Can This be Done?

2003-03-13 by Tyler Boley

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Ernst Dinkla" <
E.Dinkla@c...> wrote:
 
> Did you try it and did it work ?

Ernst, this is all old Dan Culbertson stuff. You can create profiles 
for quad inks. You can edit curves into those or any other printer 
profile with Doctor Pro, other software as well I'm sure.
However, sep curves created to work on files in a working space (as 
most used here are) will not work. The idea is to be able to convert to 
the profile for printing, converting to a quad (or any other) printer 
profile will significantly alter the RGB values in the file, therefore 
those curves will not result in the desired final RGB values for 
printing as they did when applied to the working space they were 
created in. They were created to work on AdobeRGB (or whatever).
So you'd want to edit those curves into a profile that does not 
originally alter the RGB values, like a working space. Doctor Pro will 
not edit working spaces. Alternatively, you can create a null printer 
profile, and edit your curves into it. Unfortunately since it was not 
created from the inkset/paper, it will not have the advantage of 
preview. It's only usefulness will be to print through, no more 
advantage than running a curve anyway. A null printer profile is not 
that easy to come by either.
Another option is to start from scratch. Create the profile from the 
inkset, attempt to create sep curves that print well on a file already 
converted to that profile, then edit those in. Unfortunately the new 
RGB content after conversion is not very useful for creating optimal 
quad sep curves.
However, you can do all this on CMYK profiles using old Culbertson 
tricks.
Creating really good RGB quad printer profiles, that also preview well, 
will be much more complex than it may first appear, and editing in sep 
curves made by trial and error does not take good advantage of all the 
data provided by the initial profile anyway.
Tyler

Re: [Digital BW] Converting Photoshop Curves into ICC profiles - Can This be Done?

2003-03-13 by Ernst Dinkla

Tyler, you wrote:

> Ernst, this is all old Dan Culbertson stuff. You can create
profiles
> for quad inks. You can edit curves into those or any other
printer
> profile with Doctor Pro, other software as well I'm sure.
> However, sep curves created to work on files in a working space
(as
> most used here are) will not work. The idea is to be able to
convert to
> the profile for printing, converting to a quad (or any other)
printer
> profile will significantly alter the RGB values in the file,
therefore
> those curves will not result in the desired final RGB values
for
> printing as they did when applied to the working space they
were
> created in. They were created to work on AdobeRGB (or
whatever).
> So you'd want to edit those curves into a profile that does not
> originally alter the RGB values, like a working space. Doctor
Pro will
> not edit working spaces. Alternatively, you can create a null
printer
> profile, and edit your curves into it. Unfortunately since it
was not
> created from the inkset/paper, it will not have the advantage
of
> preview. It's only usefulness will be to print through, no more
> advantage than running a curve anyway. A null printer profile
is not
> that easy to come by either.
> Another option is to start from scratch. Create the profile
from the
> inkset, attempt to create sep curves that print well on a file
already
> converted to that profile, then edit those in. Unfortunately
the new
> RGB content after conversion is not very useful for creating
optimal
> quad sep curves.
> However, you can do all this on CMYK profiles using old
Culbertson
> tricks.
> Creating really good RGB quad printer profiles, that also
preview well,
> will be much more complex than it may first appear, and editing
in sep
> curves made by trial and error does not take good advantage of
all the
> data provided by the initial profile anyway.

Yesterday I experienced some of the obstacles you describe here.
I still think there are possibilities to overcome them. I am not
after a good screen representation but a good quad printing
profile to be used in other (driver) applications than Photoshop.

In the past I have asked how a null printer profile is created
but never got an answer. The way I think it is possible to create
a good hexatone profile is for example by using Doctor Pro to
convert initial separation curves into a profile based on a null
profile. Use that as the printer profile to print a grey
stepwedge and measuring that wedge to create a general curve
again that will be implanted in the first profile with Doctor Pro
again. That could take care of the RGB shift you describe but
with the advantage that this second step can also be used to
compensate any printer quality drift in time as well. Like
linearisation but in this case at the end of the process.

So the two questions are: how do you create a null printer
profile and is it possible to write a tool that translates
spectrometer or densitometer greyscale data into a PS curve ? If
it doesn't already exists.

Any flaws in this concept ?

I have checked some IP profiles and more than in any other
profiles I came across curves in the profiles that are used to
correct the output. That's an RGB driver as well so it must be
possible to use the curves in the profiles for this purpose.

Related to this:
There's profile creation software that can change a normal colour
profile into a so called "Rich Grey". It is used to get more
neutral prints from colour printers by taking out the colour data
of the colour profile (if I did understand it correctly). I got
some samples of profiles and menu screenpics probably from
Richard Millott on the Colorsync list. www.ICCtools.com is the
company, the software is ICCtoolbox pro. The profiles have an
extension pf but they are cmyk printer profiles, I can't place
them however. Anyone familiar with that application and the
profiles ?

Ernst

Re: [Digital BW] Converting Photoshop Curves into ICC profiles - Can This be Don

2003-03-13 by Tyler Boley

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Ernst Dinkla"
<E.Dinkla@c...> wrote:

> In the past I have asked how a null printer profile is created
> but never got an answer.
I suspect my knowledge falls just short of the easy solution. Editing
the header in a working space profile from mntr to prtr doesn't seem
to work.
Look at it this way, if the profiling software sees exactly the same
measurement numbers as the color data in the target file, it should
create a null profile. I have yet to take this very far, but you get
the point.

Another possibility is that the canned Epson profiles seem to have
some sort of shell profile called "standard", then setting specific
sub profiles. The 3000 or example installs SC3000 Standard, then
several related to media and resolution such as SC3000 PQP_1440. This
"Standard" seems to be equivalent to ColormatchRGB, but is a printer
profile. I haven't tested this thouroughly but it seems to be true. If
you can deal with your null profile having a gamma of 1.8, or can then
edit it to an more useful overall gamma, you can use it as a base null
to edit curves into. It works.

>...is it possible to write a tool that translates
> spectrometer or densitometer greyscale data into a PS curve ? If
> it doesn't already exists.
It certainly seems possible, RIPs that have linearization features
built in, that support measurement devices directly, seem to do
something very similar, just perhaps not  a PS curve specifically.
Will they export a .acv curve?

> Any flaws in this concept ?
Only if it doesn't work <G>

> There's profile creation software that can change a normal colour
> profile into a so called "Rich Grey". It is used to get more
> neutral prints from colour printers by taking out the colour data
> of the colour profile (if I did understand it correctly). I got
> some samples of profiles and menu screenpics probably from
> Richard Millott on the Colorsync list. www.ICCtools.com is the
> company, the software is ICCtoolbox pro. The profiles have an
> extension pf but they are cmyk printer profiles, I can't place
> them however. Anyone familiar with that application and the
> profiles ?
I've downloded it and tried it briefly in demo mode. I tried making a
gray profile from a CMYK profile of a quad setup. It did work, and
will load and preview, very cool. Just haven't found a way to put it
to good use yet.
Way too many irons in the fire.
Tyler

Re: [Digital BW] Converting Photoshop Curves into ICC profiles - Can This be Don

2003-03-14 by Ernst Dinkla

Tyler, you wrote:

> --- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Ernst
Dinkla"
> <E.Dinkla@c...> wrote:

> Look at it this way, if the profiling software sees exactly the
same
> measurement numbers as the color data in the target file, it
should
> create a null profile. I have yet to take this very far, but
you get
> the point.

That was my approach too some time ago but I couldn't figure it
out.

> Another possibility is that the canned Epson profiles seem to
have
> some sort of shell profile called "standard", then setting
specific
> sub profiles. The 3000 or example installs SC3000 Standard,
then
> several related to media and resolution such as SC3000
PQP_1440. This
> "Standard" seems to be equivalent to ColormatchRGB, but is a
printer
> profile. I haven't tested this thouroughly but it seems to be
true. If
> you can deal with your null profile having a gamma of 1.8, or
can then
> edit it to an more useful overall gamma, you can use it as a
base null
> to edit curves into. It works.

This is interesting, there may be other ways to convert space
profiles to "null" printer profiles. Better ask that question of
null profiles again but this time in the colorsync list.

> >...is it possible to write a tool that translates
> > spectrometer or densitometer greyscale data into a PS curve ?
If
> > it doesn't already exists.
> It certainly seems possible, RIPs that have linearization
features
> built in, that support measurement devices directly, seem to do
> something very similar, just perhaps not  a PS curve
specifically.
> Will they export a .acv curve?

The Wasatch SoftRip has the Spectrocam driver integrated in that
linearisation so it is hard to get the info on the data
conversion. I have asked Wasatch to take a look at the different
workflows of quad printing (through the MIS site) and work out a
universal scheme that allows the integration of B&W printing in
the Rip. The Rip has curves for adjustment and linearisation per
Cc Mm Y K channel.
I have suggested that an overall linearisation after those steps
would be nice to even out any flaws in the curves and at the
cross over of the channels. I have no idea whether they actually
bother about a request from one Dutchman. It seems that
Ergosoft's Posterprint now has Quad support so I am not without
hope.

I will ask a befriended programmer to take a look at the ACV file
and how a Spectrocam reading can be used to make one.
Might even be easier to go to a grey profile directly if any
curve to profile conversion is creating RGB flaws in the process.

> > There's profile creation software that can change a normal
colour
> > profile into a so called "Rich Grey". It is used to get more
> > neutral prints from colour printers by taking out the colour
data
> > of the colour profile (if I did understand it correctly). I
got
> > some samples of profiles and menu screenpics probably from
> > Richard Millott on the Colorsync list. www.ICCtools.com is
the
> > company, the software is ICCtoolbox pro. The profiles have an
> > extension pf but they are cmyk printer profiles, I can't
place
> > them however. Anyone familiar with that application and the
> > profiles ?
> I've downloded it and tried it briefly in demo mode. I tried
making a
> gray profile from a CMYK profile of a quad setup. It did work,
and
> will load and preview, very cool. Just haven't found a way to
put it
> to good use yet.

Nice to know that it can work. No Macs here.

Ernst

Re: Converting Photoshop Curves into ICC profiles - Can This be Don

2003-03-14 by sjhenry

<E.Dinkla@c...> wrote:
> I will ask a befriended programmer to take a look at the ACV file
> and how a Spectrocam reading can be used to make one.
> Might even be easier to go to a grey profile directly if any
> curve to profile conversion is creating RGB flaws in the process.
> Ernst

ACV file format for gray scale curve is as follows

Bytes    Description
0-1      Magic Number??? always 1
2-3      # of curves (for gray it is alwasy 1) 1
4-5      # of points in the curve

6-7      Old value for first point on the curve
8-9      New value for first point on the curve

10-11    Old value for second point on the curve
12-13    New value for second point on the curve

and so on for each point on the curve.

"Old Value" means the original RGB value for that gray ranging from 0 
to 255. "New Value" means the RGB value obtained after printing the 
target ranging from 0 to 255. Also I assume in gray scale r = g = b.

RGB curve essentially the same but with 3 individual curve for r,g 
and b. These are from my observation. So verify it in a hex editor 
what I told is correct or not.

Jones.

Re: Converting Photoshop Curves into ICC profiles - Can This be Don

2003-03-14 by Roy Harrington

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, 
"sjhenry" <sjhenry@y...> wrote:
> <E.Dinkla@c...> wrote:
> > I will ask a befriended programmer to take a look at the ACV 
file
> > and how a Spectrocam reading can be used to make one.
> > Might even be easier to go to a grey profile directly if any
> > curve to profile conversion is creating RGB flaws in the 
process.
> > Ernst
> 
> ACV file format for gray scale curve is as follows
> 
> Bytes    Description
> 0-1      Magic Number??? always 1
> 2-3      # of curves (for gray it is alwasy 1) 1
> 4-5      # of points in the curve
> 
> 6-7      Old value for first point on the curve
> 8-9      New value for first point on the curve
> 
> 10-11    Old value for second point on the curve
> 12-13    New value for second point on the curve
> 
> and so on for each point on the curve.
> 
> "Old Value" means the original RGB value for that gray ranging 
from 0 
> to 255. "New Value" means the RGB value obtained after 
printing the 
> target ranging from 0 to 255. Also I assume in gray scale r = g 
= b.
> 
> RGB curve essentially the same but with 3 individual curve for 
r,g 
> and b. These are from my observation. So verify it in a hex 
editor 
> what I told is correct or not.
> 
> Jones.

I think the old and new values are reversed.  In Photoshop 
lingo it looks to me to be Output then Input. 

Also with RGB and CMYK I see 5 curves and 6 curves
respectively.  The first one is the composite RGB or CMYK,
followed by the individual RGB or CMYK, and then at the
end there always seems to be an extra curve that is just
(0,0) to (255,255).  I have no idea what this last thing is but
all the acv's I've seen have it.  Maybe there's some hidden
feature to PS.

Roy

Move to quarantaine

This moves the raw source file on disk only. The archive index is not changed automatically, so you still need to run a manual refresh afterward.