Pages: 1 2 [3] 4 5 6   Go Down

Author Topic: Silverfast Advanced Targets ?  (Read 23652 times)

joofa

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 544
Re: Silverfast Advanced Targets ?
« Reply #40 on: April 20, 2018, 03:42:19 pm »

Joofa, you have made four posts to this thread, none of which has had the slightest bearing on the topic and the first of which was wholly unnecessary personal attack.

Be relevant or be silent.

Jeremy

I don't understand. I have been silent. It is either Andrew Rodney, and now you, who keep bringing this 'silent' and 'on-topic' stuff again.

From your title I can see you that you are a moderator. Though, I don't know you. I joined this forum from Michael Reichman time when he was alive. Though I don't participate here much. I know it is under your purview as an administrator to keep order. However, a quick glance shows that small deviations always happen (of course not just limited to this forum) and judging by history I don't think all of us, including yourself, have been consistent.

In any case, as I said I'm silent on this topic, but you better put the dog on a leash also.  :)
Logged
Joofa
http://www.djjoofa.com
Download Photoshop and After Effects plugins

GWGill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 609
  • Author of ArgyllCMS & ArgyllPRO ColorMeter
    • ArgyllCMS
Re: Silverfast Advanced Targets ?
« Reply #41 on: April 20, 2018, 08:38:58 pm »

I would still encourage you to reconsider your decision.
It's not my decision, it's theirs.
Logged

digitaldog

  • Sr. Member
  • ****
  • Online Online
  • Posts: 20883
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: Silverfast Advanced Targets ?
« Reply #42 on: April 25, 2018, 01:21:26 pm »

It's tagging with each profile which is what's so odd.
I do have a conference call with LaserSoft tomorrow but I think I may have been using the settings incorrectly (so Mark, here is a chance to comment).
I updated the settings in CMS preferences and now I do see a major difference in the scans in terms of deltaE and saturation with Standard vs. Advanced profile. Below are the settings I used, different from the first test where the dE's were nearly the same. This setting seems to be 'correct' in that I believe only the conversions are to the scanner profiles but I find it odd that they are not embedded as shown. When I examine the differences in the two scans, the deltaE is quite large!

--------------------------------------------------
dE Report
Number of Samples: 208500
Delta-E Formula dE2000
Overall - (208500 colors)
--------------------------------------------------
Average dE:   7.29
    Max dE:  17.86
    Min dE:   0.00
 StdDev dE:   3.05


Best 90% - (187649 colors)
--------------------------------------------------
Average dE:   6.79
    Max dE:  10.94
    Min dE:   0.00
 StdDev dE:   2.79


Worst 10% - (20851 colors)
--------------------------------------------------
Average dE:  11.77
    Max dE:  17.86
    Min dE:  10.94
 StdDev dE:   0.86


--------------------------------------------------
--------------------------------------------------
Also see a visual of the differences with ColorThink Pro below. Again, I need the ear of LaserSoft engineers and/or Mark to verify that this is the best way to scan the same print.
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

digitaldog

  • Sr. Member
  • ****
  • Online Online
  • Posts: 20883
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: Silverfast Advanced Targets ?
« Reply #43 on: April 25, 2018, 01:26:18 pm »

The scans in 3D gamut movie.
Red dots are Advanced Scanner Profile.
Green are Standard Scanner Profile.
Low rez but it should give an idea of how much larger the gamut is of the scan with the larger color gamut profile. IF (big if) the settings I used for scanning are correct.
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12512
    • http://www.markdsegal.com
Re: Silverfast Advanced Targets ?
« Reply #44 on: April 25, 2018, 01:57:05 pm »

Hi Andrew,

I can't respond on this without knowing your SilverFast scan settings in detail. Can you do some screen grabs and post them? That said, their Auto IT8 process is supposed to work internally without the user needing to do anything other than load the target, trigger the procedure and follow their instructions. For rescanning the target with the new profile if I remember correctly this should be done with the same settings in CMS that are active for the Auto IT8 procedure, which I think should be <none> wherever possible, because profiling is supposed to characterize the native behaviour of the scanner in "scanner space", but best to check with LSI when you speak with them and do let us know.

Mark
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."

digitaldog

  • Sr. Member
  • ****
  • Online Online
  • Posts: 20883
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: Silverfast Advanced Targets ?
« Reply #45 on: April 25, 2018, 02:01:47 pm »

Hi Andrew,
I can't respond on this without knowing your SilverFast scan settings in detail. Can you do some screen grabs and post them?
The CMS settings below, what else do you need?
There's something kind of funky about these scanner profiles too. The scans are untagged as shown (why the settings below show Working Space CIE Lab is odd, they are most certainly RGB). When I open them in Photoshop, I get the untagged profile warning. If I select the profile and ask to open them, they still come in untagged. The Assign Profile command works however. If instead of picking the scanner profile when opening, instead I pick say ProPhoto RGB, they open tagged as such. So PS seems to have issue with the profiles being tagged in the dialog that appears when opening untagged doc's. Odd.
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12512
    • http://www.markdsegal.com
Re: Silverfast Advanced Targets ?
« Reply #46 on: April 25, 2018, 02:09:34 pm »

Sorry, I think you forgot to add-in the image of the CMS settings. I don't think you can show the CMS options being used by the Auto profiling process, but you can show the settings active when you made the verification scan of the target (with the new profile loaded - one hopes). If you could attach a photo of these CMS settings it may help decipher the rest of what's going on there.
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."

digitaldog

  • Sr. Member
  • ****
  • Online Online
  • Posts: 20883
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: Silverfast Advanced Targets ?
« Reply #47 on: April 25, 2018, 02:13:04 pm »

Sorry, I think you forgot to add-in the image of the CMS settings. I don't think you can show the CMS options being used by the Auto profiling process, but you can show the settings active when you made the verification scan of the target (with the new profile loaded - one hopes). If you could attach a photo of these CMS settings it may help decipher the rest of what's going on there.
I'm confused but then I don't know this app as well as I should. The CMS settings are below #43. The settings when making the profile isn't an issue; it made them just fine. It's the settings for using them that's confusing. Like the setting below for working space CIE-Lab which results in an RGB document (untagged) shown below again.
Isn't this what we need to setup after profile creation?
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

digitaldog

  • Sr. Member
  • ****
  • Online Online
  • Posts: 20883
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: Silverfast Advanced Targets ?
« Reply #48 on: April 25, 2018, 02:25:57 pm »

Goofy but...
IF I set "Output" dropdown to any RGB working space, the Input (scanner) profile is shown and can be configured. But I don't want an RGB working space.
IF I toggle the scanner profile I wish in the Input dropdown while RGB working space is on, then pick None, the Input dropdown grays out but the profile is seen there. But is it used?
With such a setting, I'm shown that the embedded profile is the scanner profile. I believe I tried this kind of setting, toggling "Output" to switch from Standard to Advanced profile and ended up with that tiny dE result so I figured this was all wrong. Anyway, here's what I'm talking about in terms of actually getting the scanner profile to show as being embedded but again, the input dropdown is grayed out and how one configures it to change is a rather odd GUI maneuver.
Hopefully in our conference call tomorrow, the LaserFast folks can clear this up....  :o
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12512
    • http://www.markdsegal.com
Re: Silverfast Advanced Targets ?
« Reply #49 on: April 25, 2018, 02:28:03 pm »

Yes indeed - the settings in 47 are what you need to make scans using the profile you've created.

In the top three rows that are headed "Color Management":

Input-working space is OK as you have it.
Working space-monitor should say <colorsync>.
Working Space-Output should say <RGB>

Then in the next section headed "Profiles":

Input should be the new profile you just created.
Internal should be <Adobe RGB (1998)> - (My experience - this works best with their software and the scanners I've used it with)
Grey is fin as you have it.
Output/printer is fine as you have.
Rendering intent: As you have it, or change to RelCol to taste (I use RelCol)

Then in the bottom section "Embedded ICC profiles:

If everything is set as above, Profile to embed" should say Adobe RGB (1998).

Then when you open the scan in Photoshop, assuming your Photoshop default colour working space is ProPhoto and your warnings are switched on, you should get a profile mkismatch warning. You can keep ARGB(98) or convert as you see fit.


Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12512
    • http://www.markdsegal.com
Re: Silverfast Advanced Targets ?
« Reply #50 on: April 25, 2018, 02:32:20 pm »

Goofy but...
IF I set "Output" dropdown to any RGB working space, the Input (scanner) profile is shown and can be configured. But I don't want an RGB working space.
IF I toggle the scanner profile I wish in the Input dropdown while RGB working space is on, then pick None, the Input dropdown grays out but the profile is seen there. But is it used?
With such a setting, I'm shown that the embedded profile is the scanner profile. I believe I tried this kind of setting, toggling "Output" to switch from Standard to Advanced profile and ended up with that tiny dE result so I figured this was all wrong. Anyway, here's what I'm talking about in terms of actually getting the scanner profile to show as being embedded but again, the input dropdown is grayed out and how one configures it to change is a rather odd GUI maneuver.
Hopefully in our conference call tomorrow, the LaserFast folks can clear this up....  :o

In the very first line under Color Management you disabled Colorsync and that's why it isn't letting you put a scanner profile into the first line of the Profiles section.

Why don't you want an RGB working space?
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."

digitaldog

  • Sr. Member
  • ****
  • Online Online
  • Posts: 20883
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: Silverfast Advanced Targets ?
« Reply #51 on: April 25, 2018, 02:33:28 pm »

If everything is set as above, Profile to embed" should say Adobe RGB (1998).
I don't want Adobe RGB (1998)! I don't want to funnel the color into something that 'small' for testing the differences in the two input profiles.
The new target has a wider color gamut and gamut volume. I want to see the differences in that input profile vs. the standard and smaller input profile alone. So I want two scans in the scanner RGB color space, not an RGB working space. Then I can (and have) plotted the two scans in CTP. Bringing a working space into the mix adds too much 'noise' into this process.
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12512
    • http://www.markdsegal.com
Re: Silverfast Advanced Targets ?
« Reply #52 on: April 25, 2018, 03:25:14 pm »

I don't want Adobe RGB (1998)! I don't want to funnel the color into something that 'small' for testing the differences in the two input profiles.
The new target has a wider color gamut and gamut volume. I want to see the differences in that input profile vs. the standard and smaller input profile alone. So I want two scans in the scanner RGB color space, not an RGB working space. Then I can (and have) plotted the two scans in CTP. Bringing a working space into the mix adds too much 'noise' into this process.

OK, understood. Now, going back to your settings, in the Color Management section (top), the row "Working space-output" is basically useful for softproofing a print that would be made from the scan. For that one, normal practice when printing to RGB printers would be the set it to RGB, which allows you in the Profiles section to select a printer profile for soft-proofing. So forget that setting - it's not relevant to this exercise.

For the Internal profile in the Profiles section, my pull-down menu does not allow me to set a non-RGB profile such as CIE-Lab. So if you want the least possible gamut compression from a smaller RGB working space, I would suggest setting it to ProPhoto (of course still RGB but big) or to <none>. If set to <none> it also reverts the Input Profile (top row of settings) to None, and produces a scan that embeds the scanner profile. So the internal working space becomes the scanner space as defined by the scanner profile. If you do the same with the old target and compare the two scans in CTP it should give you the comparative information you are looking for on gamut volume without impacts of a small RGB tunnel.
« Last Edit: April 25, 2018, 03:29:01 pm by Mark D Segal »
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12512
    • http://www.markdsegal.com
Re: Silverfast Advanced Targets ?
« Reply #53 on: April 25, 2018, 03:30:46 pm »

I should add, I don't think gamut volume is the key issue here. I think it's more important to test for accuracy of colour rendition. I have these new targets as well and my forthcoming "to do" list includes testing them, so we can compare notes as it goes along.
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."

digitaldog

  • Sr. Member
  • ****
  • Online Online
  • Posts: 20883
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: Silverfast Advanced Targets ?
« Reply #54 on: April 25, 2018, 03:39:02 pm »

I should add, I don't think gamut volume is the key issue here. I think it's more important to test for accuracy of colour rendition.
So scan a color target and compare the RGB values to the TDF?
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12512
    • http://www.markdsegal.com
Re: Silverfast Advanced Targets ?
« Reply #55 on: April 25, 2018, 04:01:39 pm »

So scan a color target and compare the RGB values to the TDF?

Forgot to mention in CMS for accuracy testing purposes I think the RI should likely be set to Abs Col.

For accuracy testing, the exercise I have in mind is to (1) create a new profile using their advanced profiling target, whether reflective or transparent, (2) with that new profile active, make a scan of the same target used for creating the profile. (3) Extract the colour information of the scanned image and (4) compare it to the reference values of the scanner target. Most likely CTP can be used for this, but I haven't put my mind to it yet. Otherwise, a more selective manual process is doable, as described in Appendix 3 of my book (confined to the grayscale patches). BTW, the IT8 process did achieve pretty accurate results when I tested this on three scanners back in 2011/2012.

One of the first problems I detected just as we're bantering back and forth here this afternoon is that I see no way to give the profiles I create a name. If true, this is THE PITS. One MUST be able to distinguish between iterations of custom profiles or custom profiles from canned profiles to make any of this tractable. You may wish to raise this with LSI when you talk to them. In the old days, one was given the option to name the custom profile. The new one doesn't even distinguish itself by date. Either that, or they are storing it somewhere unusual that I can't find.
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."

digitaldog

  • Sr. Member
  • ****
  • Online Online
  • Posts: 20883
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: Silverfast Advanced Targets ?
« Reply #56 on: April 25, 2018, 04:07:17 pm »

I was able to name the profiles differently.
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12512
    • http://www.markdsegal.com
Re: Silverfast Advanced Targets ?
« Reply #57 on: April 25, 2018, 04:11:16 pm »

Where did that option come up when you did the profiling?
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."

digitaldog

  • Sr. Member
  • ****
  • Online Online
  • Posts: 20883
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: Silverfast Advanced Targets ?
« Reply #58 on: April 25, 2018, 04:32:03 pm »

Where did that option come up when you did the profiling?
Yup. I stand corrected and yes, it's a mess. Just tired again, wants to overwrite the original and that original has a unique name I must have provided in the ColorSync utility. But what's odd is, I'm asked do I want to overwrite (I say yes, I have a copy) and it doesn't do so. The default name comes up SF_R (Epson Perfection blabla). Viewing the Localized description tag in the CS utility, I see that I must have updated the names there. So yeah, this is a very bad behavior on their part!
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12512
    • http://www.markdsegal.com
Re: Silverfast Advanced Targets ?
« Reply #59 on: April 25, 2018, 04:34:25 pm »

Yes indeedy. :-)
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."
Pages: 1 2 [3] 4 5 6   Go Up