Sign in with
Sign up | Sign in
Your question

Poor scanned color problem

Last response: in Computer Peripherals
Share
June 11, 2004 3:59:32 PM

Archived from groups: alt.comp.periphs.scanner (More info?)

I've been struggling to fully understand the poor color representation of my
scanned and later printed color photos. After looking at the printer end of
the process I'm now looking at the scanner. I recently scanned genuine
Pantone color samples
of pure cyan, magenta and yellow and then used Photoshop's eyedropper to
analyze the samples on its cymk sliders. The result for the cyan sample was
c-79% m- 21% y - 8% k- 0%.
The other colors had similar "contamination". Am I right in thinking that my
inexpensive 8 year old Epson needs to be replaced?
Anonymous
June 11, 2004 8:52:08 PM

Archived from groups: alt.comp.periphs.scanner (More info?)

"josh" <josqb@cinci.rr.com> wrote in message
news:E8hyc.78261$DG4.61908@fe2.columbus.rr.com...
> I've been struggling to fully understand the poor color representation of
my
> scanned and later printed color photos. After looking at the printer end
of
> the process I'm now looking at the scanner. I recently scanned genuine
> Pantone color samples
> of pure cyan, magenta and yellow and then used Photoshop's eyedropper to
> analyze the samples on its cymk sliders. The result for the cyan sample
was
> c-79% m- 21% y - 8% k- 0%.
> The other colors had similar "contamination". Am I right in thinking that
my
> inexpensive 8 year old Epson needs to be replaced?
>
Not necessarily.
Scanners do not produce perfect copies.

This is largely a color management problem.
Which I am not qualified to speak upon.

Scanner lamps do lose color temperature over time and use.
Borrow another scanner from a friend and see if you get better results.

quote:
The other colors had similar "contamination".

Have you tried to create pure sample color patches with Photoshop?
Then print at the best quality and on the best glossy photo paper.
Then scan those prints.

Here are two TIFF test scans to download. One is Kodak Gray and Color
patches, the other is Macbeth ColorChecker Chart. Both of these scans were
done with an AcerScan 620ST scanner.
http://www.carlmcmillan.com/Test_images.htm

One image is 2,112KB the other image is 2,887KB as reported by Windows
Explorer. Your mileage may vary.<g>

--
CSM1
http://www.carlmcmillan.com
--
Anonymous
June 12, 2004 2:43:52 AM

Archived from groups: alt.comp.periphs.scanner (More info?)

On Fri, 11 Jun 2004 11:59:32 GMT, "josh" <josqb@cinci.rr.com> wrote:

>I've been struggling to fully understand the poor color representation of my
>scanned and later printed color photos. After looking at the printer end of
>the process I'm now looking at the scanner. I recently scanned genuine
>Pantone color samples
>of pure cyan, magenta and yellow and then used Photoshop's eyedropper to
>analyze the samples on its cymk sliders. The result for the cyan sample was
>c-79% m- 21% y - 8% k- 0%.
>The other colors had similar "contamination". Am I right in thinking that my
>inexpensive 8 year old Epson needs to be replaced?
>

Can't you just adjust the colour balance (or whatever epson calls it)
in the driver config? I think even new consumer scanners will be a
little bit off on colour sensitivity, but you can usually compensate
for it.
---------------------------------------------

MCheu
Related resources
June 12, 2004 7:51:39 PM

Archived from groups: alt.comp.periphs.scanner (More info?)

Thank you Carl for considering my problem. I did what you suggested
regarding the use of photoshop to produce pure cmyk color swatches. Printed
them out and then scanned them back in. The results were no better. The
scanned so-called pure colors each show very noticeable mounts of the other
two cmy components.

Josh Page

"CSM1" <nomoremail@nomail.com> wrote in message
news:Yqlyc.4126$on.1843@newssvr24.news.prodigy.com...
> "josh" <josqb@cinci.rr.com> wrote in message
> news:E8hyc.78261$DG4.61908@fe2.columbus.rr.com...
> > I've been struggling to fully understand the poor color representation
of
> my
> > scanned and later printed color photos. After looking at the printer end
> of
> > the process I'm now looking at the scanner. I recently scanned genuine
> > Pantone color samples
> > of pure cyan, magenta and yellow and then used Photoshop's eyedropper to
> > analyze the samples on its cymk sliders. The result for the cyan sample
> was
> > c-79% m- 21% y - 8% k- 0%.
> > The other colors had similar "contamination". Am I right in thinking
that
> my
> > inexpensive 8 year old Epson needs to be replaced?
> >
> Not necessarily.
> Scanners do not produce perfect copies.
>
> This is largely a color management problem.
> Which I am not qualified to speak upon.
>
> Scanner lamps do lose color temperature over time and use.
> Borrow another scanner from a friend and see if you get better results.
>
> quote:
> The other colors had similar "contamination".
>
> Have you tried to create pure sample color patches with Photoshop?
> Then print at the best quality and on the best glossy photo paper.
> Then scan those prints.
>
> Here are two TIFF test scans to download. One is Kodak Gray and Color
> patches, the other is Macbeth ColorChecker Chart. Both of these scans were
> done with an AcerScan 620ST scanner.
> http://www.carlmcmillan.com/Test_images.htm
>
> One image is 2,112KB the other image is 2,887KB as reported by Windows
> Explorer. Your mileage may vary.<g>
>
> --
> CSM1
> http://www.carlmcmillan.com
> --
>
>
June 12, 2004 10:51:37 PM

Archived from groups: alt.comp.periphs.scanner (More info?)

Thanks for your comment. I presume you are referring to the adjustment
options offered in the Twain interface. I had played with saturation before,
but at your suggestion I tried tone correction, gamma and exposure. Nothing
seemed to help. My scanner when presented with a Pantone pure magenta color
swatch seems to see a significant cyan component and the other colors are
similarly misinterpreted.

Josh Page


"MCheu" <mpcheu@yahoo.com> wrote in message
news:qarkc0tsogl5n8luq692act59tr7v7nn04@4ax.com...
> On Fri, 11 Jun 2004 11:59:32 GMT, "josh" <josqb@cinci.rr.com> wrote:
>
> >I've been struggling to fully understand the poor color representation of
my
> >scanned and later printed color photos. After looking at the printer end
of
> >the process I'm now looking at the scanner. I recently scanned genuine
> >Pantone color samples
> >of pure cyan, magenta and yellow and then used Photoshop's eyedropper to
> >analyze the samples on its cymk sliders. The result for the cyan sample
was
> >c-79% m- 21% y - 8% k- 0%.
> >The other colors had similar "contamination". Am I right in thinking that
my
> >inexpensive 8 year old Epson needs to be replaced?
> >
>
> Can't you just adjust the colour balance (or whatever epson calls it)
> in the driver config? I think even new consumer scanners will be a
> little bit off on colour sensitivity, but you can usually compensate
> for it.
> ---------------------------------------------
>
> MCheu
Anonymous
June 13, 2004 1:44:52 AM

Archived from groups: alt.comp.periphs.scanner (More info?)

You are apparently just beginning to find out why soooo many people talk soooo da*# much
about "Color Management."

First, a couple of random thoughts about your posting, and predicament. Some of the things
you need to consider are

(1) HOW are you getting your CYMK measurement(s). All scanners ever made work natively
(internally) in RGB. If you are loading the image into PhotoShop (or any other image
editing program) and measuring the patches with the eye-dropper, WHAT are the settings in
PhotoShop for the RGB --> CYMK conversion? Note that there essentially are an INFINITE
VARIETY of these conversion factors available for this translation in PhotoShop,
PhotoPaint, or other image editing program.

(2) OK ... you've just gone to the time and expense of purchasing some Pantone color
chips. You get a 'Very Good' for effort, but you have NOT gotten the right thing to
analyze your situation. Why? Several reasons.

First:
There is always very significant "color contamination" in any and all printing inks, and
Pantone color chips are made with printing inks. What you describe is fairly normal for a
Cyan (printing) ink.

Second:
The Pantone Color Chips are not designed to be used for "calibration" applications. You
need something that is. That would probably (best case) be a calibrated IT8 target. They
are *expensive,* but available from several vendors.

A usable solution might be getting something like a Kodak Q13 Grayscale / Color Patch
set. They are *not* calibration / analytical grade, but they work pretty well for getting
you into the ball-park. Cost would be about $20 ... and they are available form some of
the larger photography suppliers ... for example, I *think* I've seen them at B&H. Another
option would be to get a Gretag-Macbeth Color Checker, which is often used in the
photography field, and is pretty consistent. Cost would be something like $70 ... and they
are available form some of the larger photography suppliers. The difficulty with both the
Q13 and The Color Checker is getting accurate numbers for what they *should* read.

Note that recently there has been a rather long (and very interesting) discussion of the
use of exactly these color "standards" in the Color Theory list hosted by Dan Margulis.
Incidentally ... any of his books are well worth reading.

http://groups.yahoo.com/group/colortheory/

As a point of reference, I just scanned the color patches on one of my Kodak Q13 Color
Patch targets. Here are the measurements in RGB ... I'm not including measurements in CYMK
because my conversion setup is probably quite different than yours.

Numbers are on the PhotoShop 0-255 scale.

Cyan Patch
R = 41
G = 159
B = 200
G = 128

Magenta Patch
R = 211
G = 35
B = 107
G = 96

Yellow Patch
R = 238
G = 215
B = 68
G = 206

Additionally ... you asked:
"... Am I right in thinking that my inexpensive 8 year old Epson needs to be replaced?
...."

I just replaced my old scanner with an Epson 4870 PRO ... and yes, (IMHO) it really *is*
worth it.



"josh" <josqb@cinci.rr.com> wrote in message
news:E8hyc.78261$DG4.61908@fe2.columbus.rr.com...
> I've been struggling to fully understand the poor color representation of my
> scanned and later printed color photos. After looking at the printer end of
> the process I'm now looking at the scanner. I recently scanned genuine
> Pantone color samples
> of pure cyan, magenta and yellow and then used Photoshop's eyedropper to
> analyze the samples on its cymk sliders. The result for the cyan sample was
> c-79% m- 21% y - 8% k- 0%.
> The other colors had similar "contamination". Am I right in thinking that my
> inexpensive 8 year old Epson needs to be replaced?
>
>
June 13, 2004 4:32:51 PM

Archived from groups: alt.comp.periphs.scanner (More info?)

To RSD99 my sincere thanks for your helpful posting.
The one part of the calibration process I'm still unclear about is which
knobs to turn to make an adjustment. I had been focusing on the Twain
interface, but you seem to be suggesting the settings in PhotoShop for the
RGB --> CYMK conversion.....is that correct?
Josh Page

"RSD99" <rsdwla.NOSPAM@gte.net> wrote in message
news:o PKyc.14683$H65.5683@nwrddc02.gnilink.net...
> You are apparently just beginning to find out why soooo many people talk
soooo da*# much
> about "Color Management."
>
> First, a couple of random thoughts about your posting, and predicament.
Some of the things
> you need to consider are
>
> (1) HOW are you getting your CYMK measurement(s). All scanners ever made
work natively
> (internally) in RGB. If you are loading the image into PhotoShop (or any
other image
> editing program) and measuring the patches with the eye-dropper, WHAT are
the settings in
> PhotoShop for the RGB --> CYMK conversion? Note that there essentially are
an INFINITE
> VARIETY of these conversion factors available for this translation in
PhotoShop,
> PhotoPaint, or other image editing program.
>
> (2) OK ... you've just gone to the time and expense of purchasing some
Pantone color
> chips. You get a 'Very Good' for effort, but you have NOT gotten the right
thing to
> analyze your situation. Why? Several reasons.
>
> First:
> There is always very significant "color contamination" in any and all
printing inks, and
> Pantone color chips are made with printing inks. What you describe is
fairly normal for a
> Cyan (printing) ink.
>
> Second:
> The Pantone Color Chips are not designed to be used for "calibration"
applications. You
> need something that is. That would probably (best case) be a calibrated
IT8 target. They
> are *expensive,* but available from several vendors.
>
> A usable solution might be getting something like a Kodak Q13 Grayscale /
Color Patch
> set. They are *not* calibration / analytical grade, but they work pretty
well for getting
> you into the ball-park. Cost would be about $20 ... and they are available
form some of
> the larger photography suppliers ... for example, I *think* I've seen them
at B&H. Another
> option would be to get a Gretag-Macbeth Color Checker, which is often used
in the
> photography field, and is pretty consistent. Cost would be something like
$70 ... and they
> are available form some of the larger photography suppliers. The
difficulty with both the
> Q13 and The Color Checker is getting accurate numbers for what they
*should* read.
>
> Note that recently there has been a rather long (and very interesting)
discussion of the
> use of exactly these color "standards" in the Color Theory list hosted by
Dan Margulis.
> Incidentally ... any of his books are well worth reading.
>
> http://groups.yahoo.com/group/colortheory/
>
> As a point of reference, I just scanned the color patches on one of my
Kodak Q13 Color
> Patch targets. Here are the measurements in RGB ... I'm not including
measurements in CYMK
> because my conversion setup is probably quite different than yours.
>
> Numbers are on the PhotoShop 0-255 scale.
>
> Cyan Patch
> R = 41
> G = 159
> B = 200
> G = 128
>
> Magenta Patch
> R = 211
> G = 35
> B = 107
> G = 96
>
> Yellow Patch
> R = 238
> G = 215
> B = 68
> G = 206
>
> Additionally ... you asked:
> "... Am I right in thinking that my inexpensive 8 year old Epson needs to
be replaced?
> ..."
>
> I just replaced my old scanner with an Epson 4870 PRO ... and yes, (IMHO)
it really *is*
> worth it.
>
>
>
> "josh" <josqb@cinci.rr.com> wrote in message
> news:E8hyc.78261$DG4.61908@fe2.columbus.rr.com...
> > I've been struggling to fully understand the poor color representation
of my
> > scanned and later printed color photos. After looking at the printer end
of
> > the process I'm now looking at the scanner. I recently scanned genuine
> > Pantone color samples
> > of pure cyan, magenta and yellow and then used Photoshop's eyedropper to
> > analyze the samples on its cymk sliders. The result for the cyan sample
was
> > c-79% m- 21% y - 8% k- 0%.
> > The other colors had similar "contamination". Am I right in thinking
that my
> > inexpensive 8 year old Epson needs to be replaced?
> >
> >
>
>
>
>
Anonymous
June 13, 2004 5:59:45 PM

Archived from groups: alt.comp.periphs.scanner (More info?)

Greetings RSD99.

You have said RGB readings, but you are listing 4 numbers.
RGBG.
What is the fourth number? I think Red Green Blue Green. If so why two
Greens?

--
CSM1
http://www.carlmcmillan.com
--
"RSD99" <rsdwla.NOSPAM@gte.net> wrote in message
news:o PKyc.14683$H65.5683@nwrddc02.gnilink.net...
> You are apparently just beginning to find out why soooo many people talk
soooo da*# much
> about "Color Management."
>
> First, a couple of random thoughts about your posting, and predicament.
Some of the things
> you need to consider are
>
> (1) HOW are you getting your CYMK measurement(s). All scanners ever made
work natively
> (internally) in RGB. If you are loading the image into PhotoShop (or any
other image
> editing program) and measuring the patches with the eye-dropper, WHAT are
the settings in
> PhotoShop for the RGB --> CYMK conversion? Note that there essentially are
an INFINITE
> VARIETY of these conversion factors available for this translation in
PhotoShop,
> PhotoPaint, or other image editing program.
>
> (2) OK ... you've just gone to the time and expense of purchasing some
Pantone color
> chips. You get a 'Very Good' for effort, but you have NOT gotten the right
thing to
> analyze your situation. Why? Several reasons.
>
> First:
> There is always very significant "color contamination" in any and all
printing inks, and
> Pantone color chips are made with printing inks. What you describe is
fairly normal for a
> Cyan (printing) ink.
>
> Second:
> The Pantone Color Chips are not designed to be used for "calibration"
applications. You
> need something that is. That would probably (best case) be a calibrated
IT8 target. They
> are *expensive,* but available from several vendors.
>
> A usable solution might be getting something like a Kodak Q13 Grayscale /
Color Patch
> set. They are *not* calibration / analytical grade, but they work pretty
well for getting
> you into the ball-park. Cost would be about $20 ... and they are available
form some of
> the larger photography suppliers ... for example, I *think* I've seen them
at B&H. Another
> option would be to get a Gretag-Macbeth Color Checker, which is often used
in the
> photography field, and is pretty consistent. Cost would be something like
$70 ... and they
> are available form some of the larger photography suppliers. The
difficulty with both the
> Q13 and The Color Checker is getting accurate numbers for what they
*should* read.
>
> Note that recently there has been a rather long (and very interesting)
discussion of the
> use of exactly these color "standards" in the Color Theory list hosted by
Dan Margulis.
> Incidentally ... any of his books are well worth reading.
>
> http://groups.yahoo.com/group/colortheory/
>
> As a point of reference, I just scanned the color patches on one of my
Kodak Q13 Color
> Patch targets. Here are the measurements in RGB ... I'm not including
measurements in CYMK
> because my conversion setup is probably quite different than yours.
>
> Numbers are on the PhotoShop 0-255 scale.
>
> Cyan Patch
> R = 41
> G = 159
> B = 200
> G = 128
>
> Magenta Patch
> R = 211
> G = 35
> B = 107
> G = 96
>
> Yellow Patch
> R = 238
> G = 215
> B = 68
> G = 206
>
> Additionally ... you asked:
> "... Am I right in thinking that my inexpensive 8 year old Epson needs to
be replaced?
> ..."
>
> I just replaced my old scanner with an Epson 4870 PRO ... and yes, (IMHO)
it really *is*
> worth it.
>
>
>
> "josh" <josqb@cinci.rr.com> wrote in message
> news:E8hyc.78261$DG4.61908@fe2.columbus.rr.com...
> > I've been struggling to fully understand the poor color representation
of my
> > scanned and later printed color photos. After looking at the printer end
of
> > the process I'm now looking at the scanner. I recently scanned genuine
> > Pantone color samples
> > of pure cyan, magenta and yellow and then used Photoshop's eyedropper to
> > analyze the samples on its cymk sliders. The result for the cyan sample
was
> > c-79% m- 21% y - 8% k- 0%.
> > The other colors had similar "contamination". Am I right in thinking
that my
> > inexpensive 8 year old Epson needs to be replaced?
> >
> >
>
>
>
>
Anonymous
June 13, 2004 8:10:56 PM

Archived from groups: alt.comp.periphs.scanner (More info?)

Unfortunately ... there's a LOT more to it than just twisting a knob.

Suggestion:
Do a web search ... using your favorite local search engine ... on the exact phrase

color management
Anonymous
June 13, 2004 8:12:27 PM

Archived from groups: alt.comp.periphs.scanner (More info?)

Sorry for the confusion ... it's Gray.

The way I make the measurement(s) in PhotoShop, the Gray value just sorta 'pops out,' so I
record it.
Anonymous
June 14, 2004 3:43:12 AM

Archived from groups: alt.comp.periphs.scanner (More info?)

Yes, I believe so.

--
DaveW



"josh" <josqb@cinci.rr.com> wrote in message
news:E8hyc.78261$DG4.61908@fe2.columbus.rr.com...
> I've been struggling to fully understand the poor color representation of
my
> scanned and later printed color photos. After looking at the printer end
of
> the process I'm now looking at the scanner. I recently scanned genuine
> Pantone color samples
> of pure cyan, magenta and yellow and then used Photoshop's eyedropper to
> analyze the samples on its cymk sliders. The result for the cyan sample
was
> c-79% m- 21% y - 8% k- 0%.
> The other colors had similar "contamination". Am I right in thinking that
my
> inexpensive 8 year old Epson needs to be replaced?
>
>
!