Computer Graphics with Photoshop - Working profile to work on images

Author

Working profile to work on images

Daniel Masse

2004-05-28, 4:14 am

Hello all !

I am a newbie at color management, and I am trying to understand...

1. I load an image, with the sRGB profile. It will be displayed in PS using
my monitor profile : the colors will be displayed as well as possible
(assuming my monitor is calibrated and profiled. It is). Correct ?

2. From what I have read, I should work on my image in the sRGB workspace,
using my monitor profile, then check the result I will obtain on my printer,
using soft-profiling. This operation then involves a number of trips between
the two profiles. Correct ?

3. I understand that soft profiling does not actually change my image : why
not make the changes to my image directly under soft-profiling ? I would
still have the advantage of the wide gamut of my workspace (sRGB). Does that
make sense ?

Thanks !

Mike Russell

2004-05-28, 4:14 am

Daniel Masse wrote:
> Hello all !
>
> I am a newbie at color management, and I am trying to understand...
>
> 1. I load an image, with the sRGB profile. It will be displayed in PS
> using my monitor profile : the colors will be displayed as well as
> possible (assuming my monitor is calibrated and profiled. It is).
> Correct ?

Absolutely.

> 2. From what I have read, I should work on my image in the sRGB
> workspace, using my monitor profile, then check the result I will
> obtain on my printer, using soft-profiling. This operation then
> involves a number of trips between the two profiles. Correct ?

Normaly I just work in sRGB, and print. The difference between the screen
and the print is normally subtle, and in any case the overall brightness and
shadow detail in the print is the same as that of the monitor.

> 3. I understand that soft profiling does not actually change my image
> : why not make the changes to my image directly under soft-profiling
> ? I would still have the advantage of the wide gamut of my workspace
> (sRGB). Does that make sense ?

No problem - but I suspect you have a setup issue if your print and monitor
are very much different from one another. To my mind, soft profiling is
good for relatively subtle issues, such as paper color, and shadow detail.

Others who use soft profiling more may have some comments as well.
--

Mike Russell
www.geigy.2y.net

Daniel Masse

2004-05-28, 7:14 am

Mike Russell wrote:
> Daniel Masse wrote:
>
> Normaly I just work in sRGB, and print. The difference between the
> screen and the print is normally subtle, and in any case the overall
> brightness and shadow detail in the print is the same as that of the
> monitor.

Actually, I want to send my pictures to be printed in a commercial lab : I
have found one lab, where I can ask them not to make any correction. They
use a Frontier, and I have their profile. The first tests I have made were
not very conclusive, and I am trying to find out what I did wrong...

As you say, the results I obtained could be considered as being acceptable,
but the printed images were not identical to what I have on my screen. Maybe
I am asking for too much...

>
> No problem - but I suspect you have a setup issue if your print and
> monitor are very much different from one another. To my mind, soft
> profiling is good for relatively subtle issues, such as paper color,
> and shadow detail.

Hmm... Maybe I am asking for too much... But also maybe the problem comes
from the fact that I am using an LCD screen - good quality, but not top.

The important thing is that you are telling me that working under
soft-proofing is not stupid : I will try that !

Thanks !

Johan W. Elzenga

2004-05-28, 12:14 pm

Daniel Masse <pas_de_spam_masse.dan@free.fr> wrote:

> Mike Russell wrote:
>
> Actually, I want to send my pictures to be printed in a commercial lab : I
> have found one lab, where I can ask them not to make any correction. They
> use a Frontier, and I have their profile. The first tests I have made were
> not very conclusive, and I am trying to find out what I did wrong...
>
> As you say, the results I obtained could be considered as being acceptable,
> but the printed images were not identical to what I have on my screen. Maybe
> I am asking for too much...

If you send the images to a lab and you have the profile of their
printer, why work in 'soft proof'? If you do that, you'll still send
them an image in another color space (sRGB), so the results still depend
on whether their conversion is the same as what you saw when looking at
'soft proof'. It _should_ be the same, but...

I would suggest you use "Convert to profile" to convert your image to
the color space of the Frontier. Next make the corrections the way you
like them and send this to the lab. Because the image is already in the
Frontier color space, the lab will just print the image the way you
supplied it.

--
Johan W. Elzenga johan<<at>>johanfoto.nl
Editor / Photographer http://www.johanfoto.nl/

Daniel Masse

2004-05-28, 12:14 pm

Johan W. Elzenga wrote:

> I would suggest you use "Convert to profile" to convert your image to
> the color space of the Frontier. Next make the corrections the way you
> like them and send this to the lab. Because the image is already in
> the Frontier color space, the lab will just print the image the way
> you supplied it.

Yes, this makes sense. However, I understand that it is much better to use
the widest possible gamut as a working space. This is why many people
recommend using Adobe98, which is wider than sRGB. If I use the color space
of the Frontier as working space, it seems to me that I might be losing a
lot of color information. Does that make sense ?

Johan W. Elzenga

2004-05-28, 12:14 pm

Daniel Masse <pas_de_spam_masse.dan@free.fr> wrote:

> Johan W. Elzenga wrote:
>
>
> Yes, this makes sense. However, I understand that it is much better to use
> the widest possible gamut as a working space. This is why many people
> recommend using Adobe98, which is wider than sRGB. If I use the color space
> of the Frontier as working space, it seems to me that I might be losing a
> lot of color information. Does that make sense ?

No, it doesn't. We are not talking about an image in general (where
indeed you should use a wide color space like AdobeRGB), we are talking
about an image that is specifically destined to be printed on a Fuji
Frontier. A Fuji Frontier has a certain color space, it cannot print any
colors outside that space. If you work in a wider space, you'll only end
up having colors that the Frontier can't print, so the print will be
different from what you see on your screen because the colors will be
remapped to the Frontier space. Whether you like it or not, the Frontier
color space is the color space that can and will be printed, no matter
what working space you used.

If you convert your image to the Frontier color space (and save it under
a DIFFERENT name, so you keep the original in AdobeRGB as well for other
purposes), you can edit and change anything, knowing that the Frontier
will be able to print anything you see.

--
Johan W. Elzenga johan<<at>>johanfoto.nl
Editor / Photographer http://www.johanfoto.nl/

Scott Southerland

2004-05-29, 11:14 pm

> If you convert your image to the Frontier color space (and save it under
> a DIFFERENT name, so you keep the original in AdobeRGB as well for other
> purposes), you can edit and change anything, knowing that the Frontier
> will be able to print anything you see.

Daniel - please forget that you've ever read this. Although you can get
away with small edits in the destination space, as Johan is suggesting,
there is one *minor* detail here. The Frontier profile may not be
gray-balanced! You should NOT edit in the destination space. There
really isn't any reason to do so if you are using the soft-proofing
feature.

Your image should remain in Adobe RGB or sRGB (whichever you have it
in) while you edit. You should turn on the soft-proofing feature, using
the Frontier profile that you've been provided. You may want to choose
between perceptual or relative colorimetric rendering intent -
whichever looks better to you. Once the soft-proof is 'on', THEN you
may safely edit your image knowing that the colors you see on screen
are printable by the Frontier. When you are happy, use the convert to
profile feature to move into the Frontier space and deliver that file
to your lab.

Don't edit in the destination space! If you'd like to test this, make a
new document, fill it with any shade of gray. Convert to the Frontier
profile (or most any other for those of you following along - SWOP will
prove this also). Now that you're in 'destination space', boost the
saturation all the way up. Do you still see gray? Probably not.

btg9923@att.net

2004-05-31, 7:14 pm

A little off topic question, but here it goes.

After editing in AdobeRGB, there are *significant* differences when soft
proofing with different media profiles. Questions:

1. Should this happen?
2. If no, what can be the cause?
3. If yes, which image should be "tweaked" before printing - the one
before soft proof, or the one being soft proofed?

Many books suggested creating a "master image" which can then be used
for different purposes, i.e. web, different media, etc. But the above
problem seems to imply there may need to be significant "tweaking" for
each media. Yes?

Scott Southerland wrote:
>
>
> Daniel - please forget that you've ever read this. Although you can get
> away with small edits in the destination space, as Johan is suggesting,
> there is one *minor* detail here. The Frontier profile may not be
> gray-balanced! You should NOT edit in the destination space. There
> really isn't any reason to do so if you are using the soft-proofing
> feature.
>
> Your image should remain in Adobe RGB or sRGB (whichever you have it
> in) while you edit. You should turn on the soft-proofing feature, using
> the Frontier profile that you've been provided. You may want to choose
> between perceptual or relative colorimetric rendering intent -
> whichever looks better to you. Once the soft-proof is 'on', THEN you
> may safely edit your image knowing that the colors you see on screen
> are printable by the Frontier. When you are happy, use the convert to
> profile feature to move into the Frontier space and deliver that file
> to your lab.
>
> Don't edit in the destination space! If you'd like to test this, make a
> new document, fill it with any shade of gray. Convert to the Frontier
> profile (or most any other for those of you following along - SWOP will
> prove this also). Now that you're in 'destination space', boost the
> saturation all the way up. Do you still see gray? Probably not.

Bill Hilton

2004-05-31, 11:14 pm

>From: btg9923@att.net

>After editing in AdobeRGB, there are *significant* differences when soft
>proofing with different media profiles. Questions:
>
>1. Should this happen?

Yes, it can easily happen even with different papers on the same printer, like
on the Epson 2200.

>3. If yes, which image should be "tweaked" before printing - the one
>before soft proof, or the one being soft proofed?

If you'll be outputting to different devices then my preferred way is to create
a layer set for each device and put all the tweaks for that soft-proof in
adjustment layers, storing the adjustment layers in one layer set. Name the
set for the paper, for example Epson2200_VFA or LightJet_glosssy.

Then just make the right layer set active when ready to output or turn them all
off to see the original image.

>Many books suggested creating a "master image" which can then be used
>for different purposes, i.e. web, different media, etc. But the above
>problem seems to imply there may need to be significant "tweaking" for
>each media. Yes?

Yes, but put the tweaks on adjustment layers and store in separate layer sets
and you've got it.

Bill

Scott Southerland

2004-06-01, 4:14 am

> 1. Should this happen?

It wouldn't be too surprising. Every paper has a different white point.
Each is going to absorb a different amount of ink and reflect a certain
amount of light. All of these factors play into how your eye sees the
colors. Those factors are taken into account in the profile as best as
possible.

> 2. If no, what can be the cause?

The quality of the soft-proof depends on the quality of the profile.
Each printer is slightly different. There is the possibility that your
printer makes color slightly differently than another - which would
make a difference in your printed colors.

> 3. If yes, which image should be "tweaked" before printing - the one
> before soft proof, or the one being soft proofed?

I'm not sure I understand. When you edit an image with soft-proof
turned on, you're still changing the numbers of the 'before' image. The
difference is that Photoshop is showing you the results based on the
output profile you've chosen. After you tweak the look using a
soft-proof, you still need to convert your image to the profile you
were using to soft-proof. You can do this using Image > Mode > Convert
to Profile or do it from the print with preview dialog. Bill had good
suggestions about creating different layers for different papers. In
that case, it would be best to use the Print w/ Preview dialog and
select the profile there.

> Many books suggested creating a "master image" which can then be used
> for different purposes, i.e. web, different media, etc. But the above
> problem seems to imply there may need to be significant "tweaking" for
> each media. Yes?

Yes. Each media has characteristics that affects how our eyes see
color. The goal is to characterize each of these, create profiles to
describe those differences, and then use the profiles to make the color
look the same on each type of output. It may not always be a
'significant' tweak. There are many factors involved.

btg9923@att.net

2004-06-01, 12:14 pm

Thanks Bill and Scott. You both confirmed my suspicion. Bill's
suggestion to create different layer sets for different "tweaks" is a
great solution.

Scott Southerland wrote:
>
>
> It wouldn't be too surprising. Every paper has a different white point.
> Each is going to absorb a different amount of ink and reflect a certain
> amount of light. All of these factors play into how your eye sees the
> colors. Those factors are taken into account in the profile as best as
> possible.
>
>
> The quality of the soft-proof depends on the quality of the profile.
> Each printer is slightly different. There is the possibility that your
> printer makes color slightly differently than another - which would
> make a difference in your printed colors.
>
>
> I'm not sure I understand. When you edit an image with soft-proof
> turned on, you're still changing the numbers of the 'before' image. The
> difference is that Photoshop is showing you the results based on the
> output profile you've chosen. After you tweak the look using a
> soft-proof, you still need to convert your image to the profile you
> were using to soft-proof. You can do this using Image > Mode > Convert
> to Profile or do it from the print with preview dialog. Bill had good
> suggestions about creating different layers for different papers. In
> that case, it would be best to use the Print w/ Preview dialog and
> select the profile there.
>
>
> Yes. Each media has characteristics that affects how our eyes see
> color. The goal is to characterize each of these, create profiles to
> describe those differences, and then use the profiles to make the color
> look the same on each type of output. It may not always be a
> 'significant' tweak. There are many factors involved.

"significant" is very subjective. I have been using Epson Enhanced Matte
on a 1280. Both the soft proof and prints match the monitor quite well,
and only needs tweaking once a while. But when I try to soft proof with
a third party's matte and profile, it looks so different that I have to
make major curve moves to match the monitor.

Is there any way to alleviate this during the editing workflow? For
example, does it make sense to periodically soft proof to different
profiles during the editing workflow, and try to shoot for a "middle
ground" in the edits? The goal is to come up with a final master image
that will only need "minor" tweaks for the different media later on.

Daniel Masse

2004-06-04, 12:14 pm

Scott Southerland wrote:
> Daniel - please forget that you've ever read this. Although you can
> get away with small edits in the destination space, as Johan is
> suggesting, there is one *minor* detail here. The Frontier profile
> may not be gray-balanced! You should NOT edit in the destination
> space. There really isn't any reason to do so if you are using the
> soft-proofing feature.

Scott, I apologize : I realize that I have not thanked you for your advice !

Point well taken : I should definitely work on my image in sRGB or Adobe 98.

> Your image should remain in Adobe RGB or sRGB (whichever you have it
> in) while you edit. You should turn on the soft-proofing feature,
> using the Frontier profile that you've been provided. You may want to
> choose between perceptual or relative colorimetric rendering intent -
> whichever looks better to you. Once the soft-proof is 'on', THEN you
> may safely edit your image knowing that the colors you see on screen
> are printable by the Frontier. When you are happy, use the convert to
> profile feature to move into the Frontier space and deliver that file
> to your lab.

Your advice is also very clear : I have had discussions with several people,
who always told me to work in the monitor space, and check the results by
soft-proofing (the image being in sRGB or Adobe RGB). I always thought that
it would be a lot easier to work directly with the soft-proofing turned on.
Thank you for your confirmation !

Thanks again !

Dan

Scott Southerland

2004-06-05, 7:14 pm

> But when I try to soft proof with
> a third party's matte and profile, it looks so different that I have to
> make major curve moves to match the monitor.

Unless that profile was created custom for *your* printer and paper
then it is very possible.

> The goal is to come up with a final master image
> that will only need "minor" tweaks for the different media later on.

Exactly, and this is depedent on the quality of your profiles. If your
soft proof is showing you one thing and your printer does another,
then there is either a problem with your printer (either hardware or
driver) or a problem with the profile not being a good match for your
printer.

Scott Southerland

2004-06-05, 7:14 pm

> Scott, I apologize : I realize that I have not thanked you for your advice !

No need to apologize! I'm glad to help.