[...]

If Klaus isn't around to talk about it then I'll just say that he did  
a _great_ job on GaMapICC. 

 I tested it on a 10.5 Intel Mac and it worked brilliantly.  
Unfortunately I upgraded to 10.6 and haven't been able to use it since  
due to Apple's undocumented removal of colorsync. [...]

No, I can't reveal anything about unannounced products.

Chris



On 11/4/09 12:44 PM, "Marco Ugolini" <email@hidden> wrote:

In a message dated 11/4/09 12:28 PM, Chris Cox wrote:

[...]

Could you reveal whether or not future versions of Photoshop will allow [...]

[...]

which reminds me.. did anything more happen with GaMapICC? I was never 
able to make it work, then was informed it would not work converting to 
CMYK profiles... any progress here? Seemed like a useful idea, and 
related to a bit of this "discussion". [...]

[...]

Could you reveal whether or not future versions of Photoshop will allow
conversions from RGB to CMYK using device links?

Thank you.

Marco Ugolini


 

Yes, we requested the tag so we could enable real workflows with devicelink profiles.
But, when we wrote our devicelink code, almost nobody was using the tag.
And when we tested the devicelink profile that did use the tag, a large percentage of them had problems. [...]

Kodak renamed it for marketing purposes.  "Reference Output Media Metric" just didn't register with photographers and printers.
Adobe just picked up what Kodak offered.


On 11/4/09 6:56 AM, "email@hidden" <email@hidden> wrote:


On the subject of standards I find it interesting that Adobe's [...]

[...]

Yes, I know the feeling - I also hate it when people can't get their facts 
right!  ;) 

 Bob Frost






  References:   

[...]

Re: When black is white (or blue is black) was: CMYK spaces used	for document creation

 Previous by thread:
Re: When black is white (or blue is black) was: CMYK spaces used fordocument creation

 Next by thread:
Re: When black is white (or blue is black) was: CMYK spaces used	for document creation

 Index(es):

 Date 
 Thread 












    Home 
    Archives 
    FAQ 
    Terms/Conditions 
    Contact 
    RSS 
    Lists 
    About 






 Visit the Apple Store online or at retail locations. 
1-800-MY-APPLE 
 Contact Apple | Terms of Use | Privacy Policy 
 Copyright © 2007 Apple Inc. All rights reserved. 

Am 04.11.2009 um 18:03 schrieb Martin Orpen:  
 On 4 Nov 2009, at 15:49, Karl Koch wrote:

[...]

No offense intended! 
You could blame Adobe again for not blocking the "perceptual" choice  
when no perceptual rendering intent is possible (matrix based  
priofile), but that may go too far ; [...]

Martin,
Why is the color of the ball in the lower right hand corner of the image so far off from the original image? Even with both types of conversions you are testing? I've created my own profile from an Epson 9800, converted using Photoshop and the results much closer to the original than both your converions (PS or Argyll). Could it be your profile instead of Photoshop that is causing the banding and colors changing completely from the original file?

Regards,
Steve

On 11/4/09 9:36 AM, "Martin Orpen" <email@hidden> wrote:

On 4 Nov 2009, at 15:19, Pylant, Brian wrote:

[...]

I see no evidence of it either.

Like I see no evidence that Adobe planned to make half the world use
US SWOP Coated CMYK even when it is completely inappropriate.

This is about Ubiquity not conspiracy theories.

I've heard people using the term "Adobe ProPhoto RGB" lots of times
whereas I've never heard anybody using the term "Adobe DonRGB" or
"Adobe BruceRGB". Kodak should have kept their name on that profile...

Regards

 

[...]

I'm glad you appreciate that.

[...]

Excellent, I thought my version was good -- if I asked the client to  
ignore the dark green background... 

 Yours has a nice blue background and green type!  
 Whose "perception" was used to turn R66 G0 B255 to R0 G81 B106? [...]

EXACTLY. Finally someone had the "courage" to bring up that point.  
 Sent from Terry's iPhone  
 On Nov 4, 2009, at 9:46 AM, Graeme Gill <email@hidden> wrote:  
Andrew Rodney wrote:

[...]

BTW: This is what it looks like when you convert using a DeviceLink in  
Photoshop CS4 and forget to switch the document profile from  
ProPhotoRGB to sRGB, the profile you had converted to: 
http://files.me.com/basicc/h70ygy 

If you look for flaws in Photoshop, here is one! [...]

Martin,  
 now we are getting somewhere and this thread becomes interesting again.  
 You show the amount of ignorance that is common with the common user!  
You try to convert from one to the other RGB working space – and both  
are matrix based. In a matrix based profile there is only one way of  
converting (not quite true, but good enough if we talk Adobe  
Photoshop): relaive colorimetric. 
You can´t expect to convert out-of-gamut colors this way, so that they  
would show detail after conversion. 

When building an ICC-DeviceLink profile, you can specify different  
renderings, e.g. image based in the case of Argyll. This seems to be  
the case in your idiot example. If I do a "normal" perceptual  
compression, this is the result: 

 http://files.me.com/basicc/aec8yi
Yes, it is easy – if you know what you are doing!  
So, you can see that there are several ways to color heaven. And, as  
Albert Einstein found out: Everything is relative ;-) 

This doesnt´say that Photoshop is flawless, but you´ll have to find  
other instances ;-) 

  Best regards,  
 Karl Koch  
 

[...]

I see no evidence of it either.  
Like I see no evidence that Adobe planned to make half the world use  
US SWOP Coated CMYK even when it is completely inappropriate. 

 This is about Ubiquity not conspiracy theories.  
I've heard people using the term "Adobe ProPhoto RGB" lots of times   [...]

[...]

I see no evidence that Adobe has ever tried to take credit for this profile,
and I don't know of anyone who refers to it as "Adobe ProPhoto RGB" (or
"Kodak ProPhoto RGB" for that matter -- it's simply "ProPhoto RGB" plain and
simple). 

Here is all the metadata from profile itself:
 [...]

About the RAW issue -- aren't you going to explain why you used the  
term RAW repeatedly when you wrote Adobe's technical paper (and an  
embarrassing number of other items) but criticise other people when  
they follow your example? 

  On 4 Nov 2009, at 14:06, Andrew Rodney wrote:

[...] [...]

[...]

Yes I know that. Point is (for the last time), Martin’s been very  
obscure in all the posts about his “process”. At least two others  
asked him initially, as did I, about how ICC profiles would produce  
such different results in the two products, wondering about CMM and so   [...]

[...]

The device links in question are ICC profiles (they're defined by
the ICC spec., they have the same header format, signature, and are
composed from the same repertoire of tags).  
 Graeme Gill.

[...]

Hogwash. I have dozens of color space options installed in Photoshop  
(and each version is different). So ALL are standards?

[...]

Its NOT Adobe’s ProPhoto RGB and never was. Its always been a Kodak  
designed color space.

[...]

It is for most of us!

[...] [...]

[...]

I'm glad you cleared that up.  
I'd really hate it if people started referring to "Kodak's ProPhoto  
RGB" as "Adobe's ProPhoto RGB". 

 Regards  
 --
Martin Orpen
Idea Digital Imaging Ltd  
  

[...]

Actually Kodak created it to be a REFERENCE Input/Output Medium Metric  
 - http://www.colour.org/tc8-05/Docs/colorspace/PICS2000_RIMM-ROMM.pdf  
 In that paper they say: -  
 "A new color encoding specification known as Reference
Output Medium Metric RGB (ROMM RGB) is defined. [...]

So it is the "language" that is driving the defaults -- not the country.
Interesting. I guess all Englishmen speak "SWOPv2". Thank's for taking the
time to go through the install procedure, Nipat.

Best / Roger

[...]

On 3 Nov 2009, at 14:43, Andrew Rodney wrote:  
 On Nov 3, 2009, at 7:35 AM, Martin Orpen wrote:

[...]

Yes, why do you bite?  
We were in the dying stages of a thread that has pretty conclusively  
shown that Adobe's choices of colour space in out of the box copies of   [...]

To test this, I've just tried installing Leopard to a partition, choosing Thailand in the setup, 
installing Photoshop CS4 from Design Premium package and choosing English (International) 
as default language. The Color Settings still set to "North America General Purpose 2" 
as Default. [...]

Nipat,  
I appreciate the input. It looks like the world is very uniform in terms of 
how color-management is understood and put in practice. 

 Thanks!  
 Danny  
 ----- Original Message ----- 
From: "Nipat Paiboonponpong" <email@hidden> 
To: <email@hidden>  [...]

[...]

Not with this one printer because it doesn’t print correctly (untagged  
data). 

Its also driver dependent too. I CAN print from the 2880 untagged  
correctly though CS4. 

 Andrew Rodney
http://www.digitaldog.net/  

[...]

Andrew, do you have numbers for printing through PS4 with the standard
workflow?
I have 3 sets of prints that I don't have time to measure tonight, but to
Mk1 calibrated eyeball look exactly the same.
I will pass on numbers tomorrow.



 

[...]

No profile whatsoever. Its simply the raw measure data of a TC918 run  
through both print paths. 

Andrew Rodney 
http://www.digitaldog.net/ 

The comparison dE values, are you simply round-tripping the profile's  
A2B numbers against the actual measurements or what? 

 Terry Wyse  
 Sent from Terry's iPhone  
 On Nov 3, 2009, at 3:54 PM, Andrew Rodney <email@hidden> wrote:  
Something else to try besides the tip of Eric’s. [...]

Good news (if it is consistently borne out - pardon my reticence).

That's a change from Leopard then.  I recently did this exact test under
Leopard (or rather I convinced a printer acquaintance of mine to do it!) and
found the photoshop printed target deviated from the preview one by more [...]

Something else to try besides the tip of Eric’s. I just tested this  
and it works on my end. Print the targets (untagged) in Preview! 

 Printing ICC Profile Targets via Apple Preview 5.0 (501)  
1. Open the target file and select “Print” 
2. Select the 'Color Matching' pull down menu from the Preview app and  
select 'EPSON Color Controls' 
3. Select your various Epson driver settings for the profile you want  
to create and ensure you've turned OFF the Epson Color Management  
under our driver menu. 

I just did this and also used Eric’s technique. The targets were  
measured right out of the printer, hence that max deltaE due to dry  
down. It appears that this is printing correctly (matching Photoshop  
using Eric’s tip). Test done with the 3880: 

 Overall - (930 colors) Andrew Rodney
http://www.digitaldog.net/  
 On Nov 3, 2009, at 10:01 AM, Armand Rosenberg wrote:

[...]

On Nov 3, 2009, at 1:26 PM, Martin Orpen wrote:  
 On 3 Nov 2009, at 17:34, Chris Murphy wrote:

[...]

You can also get better results in Photoshop using a better behaving  
profile, which is primarily what your test demonstrates. Not that they  
can be better outside of Photoshop. [...]

[...]

So we are all agreed that you can get better results if you do your  
colour conversions outside of Photoshop! 

 --
Martin Orpen
Idea Digital Imaging Ltd  
  

I came across a possibly useful set of suggestions, written by Mark 
Dubovoy, for users of Epson printers on the latest Mac OS. I thought 
this might be of interest to some of the audience here. You can read 
it here, if you wish: 
http://www.luminous-landscape.com/tutorials/solving. [...]

[...]

But you haven't actually done this, because your test is apples and  
oranges. 

In your Photoshop test, the B2A table of the Argyll produced inkjet  
output device profile was used to convert the image. Whereas in your  
Argyll devicelink test, you had used Argyll's -G flag which builds a   [...]

Martin,  
I am pretty sure that NOBODY ever has tested an AdobeRGB->FOGRA39  
DeviceLink in Photoshop CS4! 
Why? 
Because it doesn´t work! 
CS4 isn´t able to handle RGB-CMYK DLs, as I have stated in an article  
I posted on this list nearly a year ago and took some bashing from   [...]

[...]

Profile in both cases, yes. I see I wasn’t the only one “confused” by  
his assertion which should really have said from the get go: 

Getting PS to convert to the target profile during printing produce  
worse results than getting something else to convert using a device   [...]

On 3 Nov 2009, at 14:36, Andrew Rodney wrote:  
 On Nov 3, 2009, at 7:27 AM, Martin Orpen wrote:

[...]

Going back to the poor scientific mind set of yours, for the last  
time, I’ll paste: 

 On Nov 2, 2009, at 1:44 PM, Martin Orpen wrote:

[...]

You sound surprised by that? [...]

I've tried the technique and it works with my Epson 4880, but also the  
result was identical to printing the chart from Preview, as I  
described in a post a couple of days ago. 

 Regards  
 Rob Griffith

[...]

[...]

Crap, there you go again (why do I bite...)  
 Adobe made this a “standard” how? Name the standards body.  
And yes, if you are working in a Raw workflow, encoding into a color  
space smaller than ProPhoto (certainly Adobe RGB (1998)) will in many,   [...]

[...]

Going back to the poor scientific mind set of yours, for the last  
time, I’ll paste: 

 On Nov 2, 2009, at 1:44 PM, Martin Orpen wrote:

[...]

Sure, why try the same Device link in product A and Photoshop? Why try  
ICC profile in product B and Photoshop? [...]

On 3 Nov 2009, at 12:27, Roger wrote:  
 Chris,

[...]

As well as seeing plenty of SWOP CMYK making it into the proofing  
queues for ISO Coated v2 I'm also seeing more and more ProPhotoRGB  
files coming in for separation or retouching. 

Adobe have made this a "standard" for wide gamut images and there are  
plenty of photographers wandering around with the sincere belief that  
processing RAW files into anything smaller will destroy valuable data  
in their images. 

These images are one edit away from the imaginary colours which you  
can see ruining the "blue balls" of the Twenty-Eight Ball test chart. 

Somebody has to deal with this when the images are converted for print  Again I'd suggest looking outside of Photoshop to handle conversions  
from ProPhotoRGB. 

A practical solution offered by the ArgyllCMS is per-image gamut  
analysis prior to conversion. 

 --
Martin Orpen
Idea Digital Imaging Ltd  
  

[...]

Me "vague"? I think not.  
I made it quite plane that I don't use Photoshop for colour  
conversions because I get better results using Argyll. 

I also mentioned as an aside that this included prepping images for  
ink jet output. 

You and Chris then decided in the face of this "heresy" that I should  
be branded as unscientific, unable to provide a method, unable to  
provide examples and, strangely, that it was also necessary that I  
provide evidence of a bug in ACE to prove my point. 

As I've now posted the images, profiles and method all that remains is  
for you to admit that I was right and converting images outside of  
Photoshop gives superior results. 

 --
Martin Orpen
Idea Digital Imaging Ltd  
  

[...]

No one’s arguing that a profile and a device link are two different  
things and can and do produce differing results! Or that two profiles  
from different makes can produce differing results. The point Martin  
makes is using apples to apples (something he’s clearly not doing   [...]

Chris,

[...]

This is a very sensible statement. Which raises the issue of just what to
based a profile conversion's results evaluation on. What are your thoughts
on that? Would you care for using some standard ISO sRGB or Lab images
(ISO-12640), as a starting point, so we could all be on the same page? Or
would you have something else in mind?

Roger

 

The thing that bugs me is, yesterday, Peter Constable from Adobe assured us
that the choice of defaults is tied to localization. So, Nipat, did you
chose United States as your choice of country when you installed your Mac?
Otherwise, if I go by Peter's reasoning, how can you end up with SWOPv2 as [...]

A recent article on Luminous Landscape gives a workaround from Eric Chan
(from Adobe) for printing targets from Photoshop in Snow Leopard:

http://www.luminous-landscape.com/tutorials/solving.shtml
Photographer
Imaging & Visual Services
National Gallery of Art
Washington, D.C. [...]

On 3 Nov 2009, at 06:53, Martin Weberg wrote:  
2009/11/3 Martin Orpen <email@hidden>:

[...]

The printer profile is embedded in the images above. The options for  
converting in Photoshop are limited by the profile construction.  
Although it is worth noting that BPC works in Perceptual but I didn't  
use it. 

 The device link is here:

[...]

From memory: collink -qh -G -ip -cmt -dpp.  
 Regards  
 --
Martin Orpen
Idea Digital Imaging Ltd  
  

On Mon, 2 Nov 2009 14:32:37 -0700 Andrew Rodney <email@hidden>
wrote:

[...]

On Mon, 2 Nov 2009 18:12:56 -0500 Chris Murphy <email@hidden>
wrote:

[...]

On Tue, 03 Nov 2009 11:18:58 +1100 Graeme Gill <email@hidden>
wrote:

[...]

Thanks Graeme for stopping these "trolls" from degrading an exchange of
experiences and ideas into a mud fight in which they have excelled on
previous occasions...



On Sun, 1 Nov 2009 22:24:02 +0000 Martin Orpen <email@hidden>
wrote:

On 1 Nov 2009, at 15:46, dpascale wrote:

[...]

Thanks Martin for taking your time to contribute to this list.

I suppose most of the longtime lurkers/active members have no interest in
reading the computer generated reactions of the trolls...


Paul Foerts

 

Hi Martin,

2009/11/3 Martin Orpen <email@hidden>:

[...]

Big differences here. For completeness, please provide profiles and PS
steps and ArgyllCMS command line to reproduce your results.

Martin Weberg