Digital BW, The Print

So let me ask a few dumb questions here about Paul Roark's embedding
of PS curves in ICCs.  This method could be used for the 2400 and K3
ink right?  Are "canned" curves available for the 2400 as a starting
point?  How would I go about creating a PS curve from scratch?  Would
it work with Epson ABW or just QTR?

Thanks,

Tom Husband

Saving a curve as an ICC is easy.  Go to Edit-> Color Settings-> Spot ->
Custom Dot Gain and then create the curve you want then OK.  Click on Custom
Dot Gain again and Save Spot.  This of course creates an ICC profile but
doesn¹t help you with identifying the coordinates of the appropriate
transfer curve.  I prefer to use QTR Create ICC. I print a step wedge with
my preferred Epson Adv B&W settings and then create an ICC profile from that
with QTR Create ICC.  A transfer curve that accounts for the vagaries of the
output, white point compensation (media relativity) and black point
compensation is calculated and embedded in an ICC profile.  In addition,
full colour information from my spectro readings is recorded in the ICC tag
used for soft proofing.  Not only do I get good and easy luminance
management at printing but I get full colour soft proofing.  Quite cool.

From: Tom Husband <tom.husband@...>
Reply-To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Date: Sat, 13 Jan 2007 22:11:58 -0000
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Subject: [Digital BW] Emedding Photoshop Curves in ICCs

 
 
 

So let me ask a few dumb questions here about Paul Roark's embedding
of PS curves in ICCs.  This method could be used for the 2400 and K3
ink right?  Are "canned" curves available for the 2400 as a starting
point?  How would I go about creating a PS curve from scratch?  Would
it work with Epson ABW or just QTR?

Thanks,

Tom Husband

 
      
 


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

Thanks Steve.  I'm understanding some of this but most is over my head.
I'll have to dig into this more.

On 1/13/07, Steve Kale <stevekale@btinternet.com> wrote:
>
> Saving a curve as an ICC is easy.  Go to Edit-> Color Settings-> Spot ->
> Custom Dot Gain and then create the curve you want then OK.  Click on
> Custom
> Dot Gain again and Save Spot.  This of course creates an ICC profile but
> doesn�t help you with identifying the coordinates of the appropriate
> transfer curve.  I prefer to use QTR Create ICC. I print a step wedge with
> my preferred Epson Adv B&W settings and then create an ICC profile from
> that
> with QTR Create ICC.


Now I create ICC profiles using QTR Create ICC-RGB and an ABW created step
wedge and then print through Qimage.

  A transfer curve that accounts for the vagaries of the
> output, white point compensation (media relativity) and black point
> compensation is calculated and embedded in an ICC profile.


I don't understand this bit.   A transfer curve that accounts for the
vagaries of what?  The ICC profile you just created?  Then do you create
another ICC profile using the numbers from the first one and the transfer
curve?

  In addition,
> full colour information from my spectro readings is recorded in the ICC
> tag
> used for soft proofing.  Not only do I get good and easy luminance
> management at printing but I get full colour soft proofing.  Quite cool.


I don't have a spectro but do have a densitometer.  Am I still OK?

Thanks again,

Tom


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

Tom, this issue is not easily grasped at first and i think you are
considering a lot of info that may not be relevant.
Paul's use of applying curves to icc profiles is very different than
simply "editing" the overall curve of a profile, or making a single
channel profile from a grayscale ink setup in Photoshop.
It was intended for special custom mono inksets with different
densities of black inks to make the scale.
This required using curves in Photoshop on those channels to assign
image tonalities to ink densities properly and handle the difficult
transitions between them.
This is not within the capability of normal icc profiles, whose
primnary intent is mapping color between devices, it knows nothing of
grays and what we may want to do with them.

Therefore, at some point people began working out ways to edit those
necessary ink channel curves into output profiles so it would not have
to be done in Photoshop on the image, all you'd have to do is hit
print and select your profile, and the curve would be applied on the
fly. Long ago e did it on both RGB profiles for the 3000 and
PressReady CMYK profiles using crude methods.
The first this showed up commercially was Cone's iQuads, and
Piezography icc.
Bottom line is, unless you are using a special monochromatic ink set
like many of the MIS sets, that require the application of some kind
of curve to "partition" the inks, you have no need to concern yourself
with this at all.
In fact, it's only relevant to usingan RGB driver (usually), and not
relevant to a B&W driver like QTR or Epson's ABW.
I hope that helps.
Tyler

 --- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Tom Husband"
<tom.husband@...> wrote:
>
> Thanks Steve.  I'm understanding some of this but most is over my head.
> I'll have to dig into this more.
> 
> On 1/13/07, Steve Kale <stevekale@...> wrote:
> >
> > Saving a curve as an ICC is easy.  Go to Edit-> Color Settings->
Spot ->
> > Custom Dot Gain and then create the curve you want then OK.  Click on
> > Custom
> > Dot Gain again and Save Spot.  This of course creates an ICC
profile but
> > doesn±t help you with identifying the coordinates of the appropriate
> > transfer curve.  I prefer to use QTR Create ICC. I print a step
wedge with
> > my preferred Epson Adv B&W settings and then create an ICC profile
from
> > that
> > with QTR Create ICC.
> 
> 
> Now I create ICC profiles using QTR Create ICC-RGB and an ABW
created step
> wedge and then print through Qimage.
> 
>   A transfer curve that accounts for the vagaries of the
> > output, white point compensation (media relativity) and black point
> > compensation is calculated and embedded in an ICC profile.
> 
> 
> I don't understand this bit.   A transfer curve that accounts for the
> vagaries of what?  The ICC profile you just created?  Then do you create
> another ICC profile using the numbers from the first one and the
transfer
> curve?
> 
>   In addition,
> > full colour information from my spectro readings is recorded in
the ICC
> > tag
> > used for soft proofing.  Not only do I get good and easy luminance
> > management at printing but I get full colour soft proofing.  Quite
cool.

> 
> 
> I don't have a spectro but do have a densitometer.  Am I still OK?
> 
> Thanks again,
> 
> Tom
> 
> 
> [Non-text portions of this message have been removed]
>

> From: Tom Husband <tom.husband@...>
> Reply-To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
> Date: Sat, 13 Jan 2007 16:31:49 -0800
> To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
> Subject: Re: [Digital BW] Emedding Photoshop Curves in ICCs
> 
> Thanks Steve.  I'm understanding some of this but most is over my head.
> I'll have to dig into this more.
> 
> On 1/13/07, Steve Kale <stevekale@...> wrote:
>> 
>> Saving a curve as an ICC is easy.  Go to Edit-> Color Settings-> Spot ->
>> Custom Dot Gain and then create the curve you want then OK.  Click on
>> Custom
>> Dot Gain again and Save Spot.  This of course creates an ICC profile but
>> doesn±t help you with identifying the coordinates of the appropriate
>> transfer curve.  I prefer to use QTR Create ICC. I print a step wedge with
>> my preferred Epson Adv B&W settings and then create an ICC profile from
>> that
>> with QTR Create ICC.
> 
> 
> Now I create ICC profiles using QTR Create ICC-RGB and an ABW created step
> wedge and then print through Qimage.

Yes because Qimage needs a colour file (as I understand it) you need to use
the RGB version of QTR Create ICC so that it has 3 channels in the output
tag.
> 
>   A transfer curve that accounts for the vagaries of the
>> output, white point compensation (media relativity) and black point
>> compensation is calculated and embedded in an ICC profile.
> 
> 
> I don't understand this bit.   A transfer curve that accounts for the
> vagaries of what?  The ICC profile you just created?  Then do you create
> another ICC profile using the numbers from the first one and the transfer
> curve?

3 things:

(a) the luminance ramp produced by your particular ABW settings (stimulus
response behaviour)
(b) adjustments to (a) for paper white
(c) adjustments to (a) for ink black

> 
>   In addition,
>> full colour information from my spectro readings is recorded in the ICC
>> tag
>> used for soft proofing.  Not only do I get good and easy luminance
>> management at printing but I get full colour soft proofing.  Quite cool.
> 
> 
> I don't have a spectro but do have a densitometer.  Am I still OK?

If you only have a densitometer you can not record colour information and so
can't use this facility.  However, the luminance management side of things
is fine.

Sounds like you're in good shape using QTR's ICC tool.

Regards

Steve

Thanks Steve and Tyler.  I was probably thinking too much about all this.
I'll stick with what I'm doing now for a while.

Tom Husband


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