Digital BW, The Print

I wondering about the actual techniques people are using to scan 
targets, for example the QTR Random Gray 51.  The issues I can think of 
are ambient light, patch versus strip, speed on strip scanning, random 
versus "linear" targets.  Perhaps there are others.

Thanks,
Walt

Random is best. Strip is fine.  Patch is painful.  The software will tell
you if you're going too fast.  Ambient light is not a problem.  The ICC spec
calls for black backing not white.

> From: wwodets <odets@...>
> Reply-To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
> Date: Thu, 20 Oct 2005 15:44:56 -0000
> To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
> Subject: [Digital BW] i1 Target Scanning technique . . .
> 
> I wondering about the actual techniques people are using to scan
> targets, for example the QTR Random Gray 51.  The issues I can think of
> are ambient light, patch versus strip, speed on strip scanning, random
> versus "linear" targets.  Perhaps there are others.
> 
> Thanks,
> Walt
>

When working on a curve description, I strip scan the 21-gray target.
For the final readings, I use patch method, which seems to be more
repeatable, to create the icc profile.  For 21 patches, that seems fastest,
for > 21 patches multiple reads and average using measuretool would be
faster.

Best regards,
John Moody

-----Original Message-----
From: DigitalBlackandWhiteThePrint@yahoogroups.com
[mailto:DigitalBlackandWhiteThePrint@yahoogroups.com]On Behalf Of wwodets
Sent: Thursday, October 20, 2005 11:45 AM
To: DigitalBlackandWhiteThePrint@yahoogroups.com
Subject: [Digital BW] i1 Target Scanning technique . . .

I wondering about the actual techniques people are using to scan
targets, for example the QTR Random Gray 51.  The issues I can think of
are ambient light, patch versus strip, speed on strip scanning, random
versus "linear" targets.  Perhaps there are others.

Thanks,
Walt




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

Steve Kale wrote:
>  The ICC spec
> calls for black backing not white.

Which is a sensible thing when you have text or images printed 
at the other side of the paper too but not for photography 
etc. Better use two or three extra sheets of the same paper 
you printed on underneath the target. That's what the majority 
of the color gurus do.

Other recommendations:

Reading the target three times or even better print three 
targets and measure them + averaging the results is something 
I now do. I print three on wide paper and have one of the 
targets turned 180 degrees in the print.

Drying 12 to 24 hours before measuring. I use an old hair 
dryer (cap model) mounted on a box to speed the drying process.

Change an old HP plotter so it can drive your Spectrometer if 
you get bored of measuring or if your hand isn't moving 
consistently. Drinking a glass of wine is another method.



                    --
           Ernst Dinkla


www.pigment-print.com
(         unvollendet         )

> From: Ernst Dinkla <E.Dinkla@...>
> Reply-To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
> Date: Thu, 20 Oct 2005 18:20:18 +0200
> To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
> Subject: Re: [Digital BW] i1 Target Scanning technique . . .
> 
> Steve Kale wrote:
>>  The ICC spec
>> calls for black backing not white.
> 
> Which is a sensible thing when you have text or images printed
> at the other side of the paper too but not for photography
> etc. Better use two or three extra sheets of the same paper
> you printed on underneath the target. That's what the majority
> of the color gurus do.

There's a lot of debate about this.  The spec calls for black backing
regardless but this is one of those things that will probably change...

Steve, Ernst, John-

Thanks for all the responses.  I have been using the random target, 
51 patches, strip read, the occasional glass of wine (but not two) 
and a white card overlaid with a piece of the same paper.

I have not been averaging multiple targets or reads, though I've done 
consecutive profiles and overlaid then and find them extremely 
close.  I imagine that this would have to be done by manually 
calculating the averages for each of the 51 steps and then entering 
them in the Measure Tool text file.  Is that correct, or is there 
something sleeker I'm not thinking of?

A sample of three is probably better than a sample of one, but 
*statistically* a sample of three (compared to thousands of prints 
done with the profile) is probably meaningless (small "N"). 

Thanks,
Walt  





--- In DigitalBlackandWhiteThePrint@yahoogroups.com, Steve Kale 
<stevekale@b...> wrote:
>
> 
> 
> 
> > From: Ernst Dinkla <E.Dinkla@c...>
> > Reply-To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
> > Date: Thu, 20 Oct 2005 18:20:18 +0200
> > To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
> > Subject: Re: [Digital BW] i1 Target Scanning technique . . .
> > 
> > Steve Kale wrote:
> >>  The ICC spec
> >> calls for black backing not white.
> > 
> > Which is a sensible thing when you have text or images printed
> > at the other side of the paper too but not for photography
> > etc. Better use two or three extra sheets of the same paper
> > you printed on underneath the target. That's what the majority
> > of the color gurus do.
> 
> There's a lot of debate about this.  The spec calls for black 
backing
> regardless but this is one of those things that will probably 
change...
>

I have not averaged.  I don't think the Average module in MT works without
the dongle.

> From: wwodets <odets@...>
> Reply-To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
> Date: Thu, 20 Oct 2005 17:19:21 -0000
> To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
> Subject: Re: [Digital BW] i1 Target Scanning technique . . .
> 
> Steve, Ernst, John-
> 
> Thanks for all the responses.  I have been using the random target,
> 51 patches, strip read, the occasional glass of wine (but not two)
> and a white card overlaid with a piece of the same paper.
> 
> I have not been averaging multiple targets or reads, though I've done
> consecutive profiles and overlaid then and find them extremely
> close.  I imagine that this would have to be done by manually
> calculating the averages for each of the 51 steps and then entering
> them in the Measure Tool text file.  Is that correct, or is there
> something sleeker I'm not thinking of?
> 
> A sample of three is probably better than a sample of one, but
> *statistically* a sample of three (compared to thousands of prints
> done with the profile) is probably meaningless (small "N").
> 
> Thanks,
> Walt  
>

wwodets wrote:

> Thanks for all the responses.  I have been using the random target, 
> 51 patches, strip read, the occasional glass of wine (but not two) 
> and a white card overlaid with a piece of the same paper.
> 
> I have not been averaging multiple targets or reads, though I've done 
> consecutive profiles and overlaid then and find them extremely 
> close.  I imagine that this would have to be done by manually 
> calculating the averages for each of the 51 steps and then entering 
> them in the Measure Tool text file.  Is that correct, or is there 
> something sleeker I'm not thinking of?

My youngest son wrote a small command line averaging program 
for the PC but for color targets. It might be usable for the 
51 steps target but I have to check that. Printing two or 
three different targets on a wide format roll at the same time 
with one of them 180 degrees turned  is a sensible thing to do 
as the printer is used that way and by that the drying time 
between strokes etc corresponds more to daily practice.  I 
sometimes observe a small difference in color at the start of 
a printjob, having the patches reversed on one target reduces 
the problem + printing another file just before it helps too. 
Cleaning the heads to get all the nozzles is good too but you 
should at least run a normal print before printing the target 
to get the head in its normal printing condition.

BTW, Epson has a hidden advice in the Colorbase FAQ that says 
no speed printing (bidirectional) is color consistent due to 
the reversed order of laying down of the hues and the 
increased dotgain.  So even for their x800 printers they give 
that advice. I have never before seen that advice but always 
suspected that something happens in the reversed laying down 
order.. Can't recall right now what the subject line was in 
the FAQ that covers it but ithe same subject line appears 
twice in the FAQ with different underlying texts.


> A sample of three is probably better than a sample of one, but 
> *statistically* a sample of three (compared to thousands of prints 
> done with the profile) is probably meaningless (small "N"). 

When offset presses are  profiled the profile people select a 
quantity of say 12 target prints of a print run. Based on 
their knowledge at what time the variations in the printrun 
occur. The quantity of sheets produced on an offset press with 
the same profile must be a 100x more though than what is done 
with one profile on an inkjet printer. I'm trying to reduce 
reading errors and print variation at the same time. I have to 
say that the profiles I made since adopting that method are 
more consistent than the ones before.

                    --
           Ernst Dinkla


www.pigment-print.com
(         unvollendet         )