Ad
Ad
Ad
Pages: [1]   Bottom of Page
Print
Author Topic: New Hahnamühle ICC Profiles  (Read 5561 times)
paulbk
Sr. Member
****
Offline Offline

Posts: 457



« on: May 05, 2006, 06:17:35 AM »
ReplyReply

It appears that Hahnamühle has updated(?) the factory ICC profiles for Photo Rag.
The new profiles are renamed and are smaller.
Old ---- HahnPhotoRagUC4000.icc,  File Size: 1418 kb
New ---- HFA4000PhotoRag.icc,  File Size: 836 kb
Get profiles here

Or am I missing something?
« Last Edit: May 05, 2006, 06:35:19 AM by paulbk » Logged

paul b. kramarchyk
Barkhamsted, Connecticut, USA
francois
Sr. Member
****
Offline Offline

Posts: 6432


« Reply #1 on: May 05, 2006, 06:28:24 AM »
ReplyReply

Quote
It appears that Hahnamühle has updated(?) the factory ICC profiles for Photo Rag.
The new profiles are renamed and are smaller.
Old:  HahnPhotoRagUC4000.icc,  File Size: 1418 kb
New:  HFA4000PhotoRag.icc,  File Size: 836 kb
Get profiles here

Or am I missing something?
[a href=\"index.php?act=findpost&pid=64581\"][{POST_SNAPBACK}][/a]
I download the ones for the Epson 2100/2200 and they are different! I made the comparison with ColorSync Utility but I haven't had the time to do a few prints to see real-world differences.
Logged

Francois
colourperfect
Jr. Member
**
Offline Offline

Posts: 75


WWW
« Reply #2 on: May 07, 2006, 01:14:07 PM »
ReplyReply

They appear to be much smaller which suggests fewer patches were used to create them. This doesnt necessarlliy mean they are not very good but it does make you wonder.

Ian

http://profiles.colourperfect.co.uk
Logged
digitaldog
Sr. Member
****
Online Online

Posts: 8040



WWW
« Reply #3 on: May 07, 2006, 04:10:25 PM »
ReplyReply

Quote
They appear to be much smaller which suggests fewer patches were used to create them. This doesnt necessarlliy mean they are not very good but it does make you wonder.

Ian

http://profiles.colourperfect.co.uk
[a href=\"index.php?act=findpost&pid=64716\"][{POST_SNAPBACK}][/a]

Could also be less data in private tags. If you built a profile using ProfileMaker which embeds all the measured spectral data, it would be larger in size than a differing package that doesn't even from the same number of patches. But of course, you just got to test the two and look at the results.
Logged

Andrew Rodney
Author “Color Management for Photographers”
http://digitaldog.net/
Brian Gilkes
Sr. Member
****
Offline Offline

Posts: 431


WWW
« Reply #4 on: May 07, 2006, 07:05:33 PM »
ReplyReply

I wrote a lenghthy post and thankfully for readers I lost it in a crash.
Briefly I have printed test charts using Hahnemule's old and new profiles and with a custom profile. These were analysed by Les Walkling who had also built the custom profile.
The new H profile seems soundly built, unlike those of some other manufacturers.
The custom profile shows significant gamut improvement and different printer linearities.
Heavier inkloads have so far failed to improve blacks to the level we would be happy with.
This is the case with all matte art papers so far tested.
Perhaps Canon or Epson may pull something out of the hat later this year.
Recommendation : Obtain an expertly built custom profile and print on.
Cheers,
Brian,
www.pharoseditions.com.au
Logged
thompsonkirk
Full Member
***
Offline Offline

Posts: 205


WWW
« Reply #5 on: May 08, 2006, 04:59:18 PM »
ReplyReply

Seconding Brian's recommendation:

"Obtain an expertly built custom profile and print on."

I tried the new profile for HPR & 4000, & it did just fine printing PhotoDisks & grayscales.  But in an actual print it did 'the green thing': as happens consistently with the Epson driver & generic profiles, it generated muddled greens that undo tonal gradations & detail in trees & grasses.  

It compared quite unfavorably to an inexpensive InkJetArt custom profile.  

Kirk
Logged
JeffKohn
Sr. Member
****
Offline Offline

Posts: 1671



WWW
« Reply #6 on: May 08, 2006, 10:52:52 PM »
ReplyReply

Quote
Seconding Brian's recommendation:

"Obtain an expertly built custom profile and print on."

I tried the new profile for HPR & 4000, & it did just fine printing PhotoDisks & grayscales.  But in an actual print it did 'the green thing': as happens consistently with the Epson driver & generic profiles, it generated muddled greens that undo tonal gradations & detail in trees & grasses. 

It compared quite unfavorably to an inexpensive InkJetArt custom profile. 

Kirk
[a href=\"index.php?act=findpost&pid=64831\"][{POST_SNAPBACK}][/a]
I think you'll find this isn't so much an issue with generic profiles as with profiles created with GMB ProfileMaker Pro or Eye One Match. (I believe InkJetArt is no longer using PMP due to licensing restrictions placed on the created profiles).

This problem seems to be especially bad when the profile is created with the Epson driver in ICM mode with no color adjustments, probably more so on the 2400 than the 4800/7800/9800 due to the complete lack of linearization when printing in this mode. I've found that for this configuration, profiling with Color Controls - Adobe RGB yields a much better profile.

Of course there's no guarantee that purchasing a custom profile from somebody will solve the problem unless they're using software other than PMP/i1, because most people still seem to think that Color Controls mode is to be avoided and that ICM mode will produce better profiles. My experience has found that for matte papers such as HPR, this just isn't true.
« Last Edit: May 08, 2006, 10:54:49 PM by JeffKohn » Logged

paulbk
Sr. Member
****
Offline Offline

Posts: 457



« Reply #7 on: May 09, 2006, 05:32:20 AM »
ReplyReply

Quote
...due to the complete lack of linearization when printing in this mode. I've found that for this configuration, profiling with Color Controls - Adobe RGB yields a much better profile.[a href=\"index.php?act=findpost&pid=64867\"][{POST_SNAPBACK}][/a]
Jeff,
This is very interesting. I use an Epson 4000 (not 4800), and this may help with my profile problem. I’ve had many custom profiles made. Most are “okay” but all tend to be a little too dark with muddy shadow detail (not unacceptable, but not what I think the printer is capable of).

Just to be clear...., Are you saying that when I print profile targets I should NOT use No Color Management in the Advanced - Printer Color Management dialog. Rather, I should use the sRGB selection. (fyi.. the 4000 does not have "Adobe RGB" selection in the driver dialog, I'm guessing that sRGB will do the same thing?)

paul
Logged

paul b. kramarchyk
Barkhamsted, Connecticut, USA
digitaldog
Sr. Member
****
Online Online

Posts: 8040



WWW
« Reply #8 on: May 09, 2006, 08:20:28 AM »
ReplyReply

Quote
I think you'll find this isn't so much an issue with generic profiles as with profiles created with GMB ProfileMaker Pro or Eye One Match. (I believe InkJetArt is no longer using PMP due to licensing restrictions placed on the created profiles).
[a href=\"index.php?act=findpost&pid=64867\"][{POST_SNAPBACK}][/a]

Hum... I have to admit I haven't seen this issue you report with ProfileMaker or EyeOne Match, at least not with the current versions where GMB cleaned up their color engine. Now I find those packages and Monaco PROFILER produce roughly similar qualities in the profiles. There were smoothness issues in older versions of the GMB engine. If you have an older profile and the newer software, you might try regenerating the profile (drag and drop; all the measured data is contained in the profile).

GMB still has licensing of their profiles to the masses but it costs pretty big bucks. The newer Epson profiles are built from ProfileMaker list time I checked.
Logged

Andrew Rodney
Author “Color Management for Photographers”
http://digitaldog.net/
JeffKohn
Sr. Member
****
Offline Offline

Posts: 1671



WWW
« Reply #9 on: May 09, 2006, 11:04:38 AM »
ReplyReply

Quote
Jeff,
This is very interesting. I use an Epson 4000 (not 4800), and this may help with my profile problem. I’ve had many custom profiles made. Most are “okay” but all tend to be a little too dark with muddy shadow detail (not unacceptable, but not what I think the printer is capable of).

Just to be clear...., Are you saying that when I print profile targets I should NOT use No Color Management in the Advanced - Printer Color Management dialog. Rather, I should use the sRGB selection. (fyi.. the 4000 does not have "Adobe RGB" selection in the driver dialog, I'm guessing that sRGB will do the same thing?)

paul
[a href=\"index.php?act=findpost&pid=64888\"][{POST_SNAPBACK}][/a]
Yes, I'm saying that when I print profile target I set the print driver to "Color Controls", with the Color Mode set to "Adobe RGB", Brightness=+1, Contrast=-1, and all of the color-specific sliders at -1 (got these settings from Julia Borg).

What I've found is that using these print settings to profile, not only do I get a larger gamut particularly in the shadows, but shadow detail also seem to be a bit better.

Since you mention that you're using a 4000 I should note that part of why people have avoided Color Controls mode in the past is because it supposedly reduced the color gamut of the printer. You may find that choosing this mode will yield a smaller overall gamut, I don't know. On the Epson 2400 there doesn't seem to be any penalty in gamut but that could be because of the new "Adobe RGB" selection. Even if this configuration does have a smaller gamut overall you may find that for certain images it produces better results.
Logged

JeffKohn
Sr. Member
****
Offline Offline

Posts: 1671



WWW
« Reply #10 on: May 09, 2006, 11:24:00 AM »
ReplyReply

Quote
Hum... I have to admit I haven't seen this issue you report with ProfileMaker or EyeOne Match, at least not with the current versions where GMB cleaned up their color engine. Now I find those packages and Monaco PROFILER produce roughly similar qualities in the profiles. There were smoothness issues in older versions of the GMB engine. If you have an older profile and the newer software, you might try regenerating the profile (drag and drop; all the measured data is contained in the profile).
Maybe the problem isn't just PMP/i1, but the profiles I've seen where the problem was particularly bad always seemt to be created with those packages. I don't have any experience with the other software packages (switching is just too expensive, but don't get me started on the pricing of these packages that's a whole 'nother thread). I do agree that later releases have gotten better, but seems that the Logo engine still favors saturation over luminosity, with the result being that darker tones become indistinguishable and you end up with a murky mess.

With my Epson 2400 this seems to be particularly a problem with dark greens, followed by browns. The greens can be a real problem when printing landscape shots. The problem is much worse on rag papers with their limited dmax, in fact with the profile I recently created for the new Innova F-Type Gloss it's almost a complete non-issue; the gamut and dmax of that paper are pretty incredible.

Quote
GMB still has licensing of their profiles to the masses but it costs pretty big bucks. The newer Epson profiles are built from ProfileMaker list time I checked.
I personally think this is absolutely ridiculous, what right do they have to assert any terms on the profile that I created? To me this is no different than if Nikon tried to tell me what I could or could not do with images taken with my D2x. I really wish somebody with deep pockets would challenge this in court, as I think they would probably win.
Logged

opgr
Sr. Member
****
Offline Offline

Posts: 1125


WWW
« Reply #11 on: May 09, 2006, 11:31:08 AM »
ReplyReply

Quote
Yes, I'm saying that when I print profile target I set the print driver to "Color Controls", with the Color Mode set to "Adobe RGB", Brightness=+1, Contrast=-1, and all of the color-specific sliders at -1 (got these settings from Julia Borg).

What I've found is that using these print settings to profile, not only do I get a larger gamut particularly in the shadows, but shadow detail also seem to be a bit better.


But how do you print the target from Photoshop? Attached AdobeRGB and print with colormanagement?

The problem with using additional color conversions in the printer driver is that it shifts the primary and secundary colors in the target RGB away from the output primaries. This may significantly hinder correct behavior definition for the profiling software.

Would you have profile plots available to help us see the differences in gamut your experiencing?

The problem with the dark, saturated greens is actually a problem in the blackpoint compensation algorithms. Following is a gamut plot for L = 14 of a 4000 on GMG paper:



As can be seen, the a & b axes cross exactly at the edge of gamut. Unfortunately on the green side of colors. The profiling software now has to redefine what it considers neutral to make perceptual matches. But you can imagine that the smallest shift in mapping can have a significant impact on the colorperception because your moving around neutral. Note that the darkest color has L = 5 or 6. Anything between L=14 and L=min does not have an equivalent for true neutral, and you can't really make greens.

The trick is to compensate the dark saturated greens to a lighter, saturated version. The profiling software may do this in the Perceptual table, or otherwise you may do this yourself with Photoshop. The difficult part is that the slightest corrections have a significant visual/perceptual impact in that color range.
Logged

Regards,
Oscar Rysdyk
theimagingfactory
opgr
Sr. Member
****
Offline Offline

Posts: 1125


WWW
« Reply #12 on: May 09, 2006, 11:39:11 AM »
ReplyReply

Quote
I personally think this is absolutely ridiculous, what right do they have to assert any terms on the profile that I created?

I think this is primarily meant so OEMs pay a fee when providing a profile with a device. Idem paper manufacturers. Otherwise they would be out-of-business real quick. Additionally, I don't think you created the profile, you just provided the measurements. Their software, which you only use under license, really created the profile, is it not?
Logged

Regards,
Oscar Rysdyk
theimagingfactory
digitaldog
Sr. Member
****
Online Online

Posts: 8040



WWW
« Reply #13 on: May 09, 2006, 11:41:35 AM »
ReplyReply

Quote
Maybe the problem isn't just PMP/i1, but the profiles I've seen where the problem was particularly bad always seemt to be created with those packages. I don't have any experience with the other software packages (switching is just too expensive, but don't get me started on the pricing of these packages that's a whole 'nother thread). I do agree that later releases have gotten better, but seems that the Logo engine still favors saturation over luminosity, with the result being that darker tones become indistinguishable and you end up with a murky mess.

Well I'm using the newer ColorFul option which you might like or not (there's also Chroma Plus and Classic). And the issues could simply be the gamut of the device.

I agree that setting the driver to something other than NCA can help but I would submit at the expense of gamut. For some images however, it's a worthwhile trade off. The newer K3 printers are showing improvements in linearity with NCA settings but they have a long way to go.

Quote
I personally think this is absolutely ridiculous, what right do they have to assert any terms on the profile that I created? To me this is no different than if Nikon tried to tell me what I could or could not do with images taken with my D2x. I really wish somebody with deep pockets would challenge this in court, as I think they would probably win.
[a href=\"index.php?act=findpost&pid=64919\"][{POST_SNAPBACK}][/a]

An end user like you and I can build and sell or give away a profile within our EULA but a big company like Epson (or a paper company in business to profit from the profile to the masses) has to pay a fee. I'm not at liberty to tell you this fee but it's not outrageous. Such companies could easily afford to do this.
Logged

Andrew Rodney
Author “Color Management for Photographers”
http://digitaldog.net/
JeffKohn
Sr. Member
****
Offline Offline

Posts: 1671



WWW
« Reply #14 on: May 09, 2006, 12:08:42 PM »
ReplyReply

Quote
I think this is primarily meant so OEMs pay a fee when providing a profile with a device. Idem paper manufacturers.
No disagreement, but I still think it's unreasonable.

Quote
Additionally, I don't think you created the profile, you just provided the measurements. Their software, which you only use under license, really created the profile, is it not?
That might be true if there was no expertise at all involved in making a profile (maybe - but maybe not). However I think most of us would agree that not all profiles are created equally, so that would seem to indicate that the expertise of the person creating the profile plays a role.


Quote
An end user like you and I can build and sell or give away a profile within our EULA but a big company like Epson (or a paper company in business to profit from the profile to the masses) has to pay a fee. I'm not at liberty to tell you this fee but it's not outrageous. Such companies could easily afford to do this.
I could maybe give away or sell a profile to an invididual, but I think if I wanted to post a profile I had created for some new paper on a web forum such as this one for the benefit of other forum members, GMB might have something to say about that.

Maybe Epson can afford to pay the fee, but apparently some small business felt that they could not. I know there has been at least one case (if not more) of a retailer having to pull profiles they had created for the benefit of their customers (for papers that didn't have manufacturer-supplied profile). The fact that InkjetArt swtiched from GMB software to X-Rite would lead me to believe that X-Rite either doesn't restict users in this way, or at least has more favorable terms. (admittedly I haven't looked into it too closely).
Logged

digitaldog
Sr. Member
****
Online Online

Posts: 8040



WWW
« Reply #15 on: May 09, 2006, 12:18:09 PM »
ReplyReply

Quote
I could maybe give away or sell a profile to an invididual, but I think if I wanted to post a profile I had created for some new paper on a web forum such as this one for the benefit of other forum members, GMB might have something to say about that.

I don't think there would be an issue. I'm not a lawyer and don't enjoy deciphering EULAs and in the end, you'd be posting a generic profile anyway.
Logged

Andrew Rodney
Author “Color Management for Photographers”
http://digitaldog.net/
JeffKohn
Sr. Member
****
Offline Offline

Posts: 1671



WWW
« Reply #16 on: May 09, 2006, 01:32:16 PM »
ReplyReply

Quote
But how do you print the target from Photoshop? Attached AdobeRGB and print with colormanagement?
I print the target from Qimage with application ICM disabled.

Quote
The problem with using additional color conversions in the printer driver is that it shifts the primary and secundary colors in the target RGB away from the output primaries. This may significantly hinder correct behavior definition for the profiling software.
Judging from the posts here I have a feeling both you and Andrew may understand the technical side of these matters more thoroughly than I do so you may be correct. However based on my testing I'd say any hindrance caused by this approach is less significant than the hindrance caused by the non-linearity of the Epson 2400 driver in ICM/NCA mode. The 2400 just puts down too much ink when used in ICM/NCA mode in my experience.

Quote
Would you have profile plots available to help us see the differences in gamut your experiencing?
I like to use this image in conjunction with soft-proof and gamut warning in PS to evaluate similar profiles as it seems to do a good job of showing subtle differences that aren't always apparent in some of the other gamut viewers I've seen. It also seems to coincide with what I see when soft-proofing real-world images as well as actually printing them. Here's a screen cap, showing my profile in Color Controls mode on the left, and Hahnamuehl's newest 2400 PR profile on the right. As you can see my profile has a larger gamut pretty much across the entire L range, but the difference is most apparent in the bottom 3 slices which represent L values of 10, 20, and 30.

http://www.pbase.com/image/59932664/original
Logged

JeffKohn
Sr. Member
****
Offline Offline

Posts: 1671



WWW
« Reply #17 on: May 09, 2006, 01:36:52 PM »
ReplyReply

Quote
I agree that setting the driver to something other than NCA can help but I would submit at the expense of gamut.
That's the conventional wisdom, but I'm just not seeing it with the 2400. My CC/Adobe-RGB profiles have consistently larger gamuts than the generic ICM/NCA profiles released by the manufacturers. See image in my post above.

Quote
The newer K3 printers are showing improvements in linearity with NCA settings but they have a long way to go.
Yes and I have a feeling that the improvements have been greater for the "pro" models than for my 2400, so it bears considering that my experiences may not directly translate to somebody using a 7800.
Logged

Marcst
Newbie
*
Offline Offline

Posts: 12


« Reply #18 on: May 10, 2006, 05:39:25 AM »
ReplyReply

Quote
Maybe the problem isn't just PMP/i1, but the profiles I've seen where the problem was particularly bad always seemt to be created with those packages. I don't have any experience with the other software packages (switching is just too expensive, but don't get me started on the pricing of these packages that's a whole 'nother thread). I do agree that later releases have gotten better, but seems that the Logo engine still favors saturation over luminosity, with the result being that darker tones become indistinguishable and you end up with a murky mess.

AFAIK, this was/is a PM issue. The actual Version 5.0.5b features a completely reworked “Colorful” gamut mapping. The changelog lists “better details in dark green colors”. Slightly muddy greens still appear with “relative colormetric” but overall the problem is definitely “half-way fixed”. I never experienced muddy greens with X-Rites Pulse – I'm really sure it was/is a GMB-problem.
Logged
Brian Gilkes
Sr. Member
****
Offline Offline

Posts: 431


WWW
« Reply #19 on: May 14, 2006, 12:17:19 AM »
ReplyReply

Colour controls with gamma 2.2produced more neutral greyscale than NCA with the K2 inkset , but there was some nonlinearity in hue. With the K3 inkset ink linearity is greatly improved and NCA serves us well.
Brian
www.pharoseditions.com.au
Logged
Pages: [1]   Top of Page
Print
Jump to:  

Ad
Ad
Ad