Pages: 1 2 3 [4] 5 6   Go Down

Author Topic: Silverfast Advanced Targets ?  (Read 10739 times)

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 15630
    • http://www.digitaldog.net/
Re: Silverfast Advanced Targets ?
« Reply #60 on: April 25, 2018, 04:42:13 pm »

Getting back to accuracy, it's clear to me, IF I scanned the Epson 3880 print correctly that one scanner profile has a wider color gamut than the other. As do the scans! So if that's true, the lower gamut input profile alone is going to be less 'accurate' if our goal is to capture what's on the print.

Now comparing a scan of the target to the TDF does tell us about accuracy within the gamut of the profile and target. But then what happens if we place a wider gamut original onto the scanner? The printer I used has a pretty darn wide color gamut. The scanner really doesn’t have a color gamut but the profile from the target used to build that profile does and it's 'limited'. So color accuracy here would need to define what the scanner can capture no? And with the Standard target, I can't capture as wide a gamut from a print than I can with the Advanced target; again IF (big if) my scan settings are correct.

What one might have to do is create a custom target that mimics both of the IT8's but printed on a much wider gamut output device. These are RA4 prints right? An Epson has a wider color gamut.
We'd need to measure each patch and produce a custom TDF and maybe build a profile from that if possible. At the very least, we could use the profiles from those supplied by LaserSoft, scan our custom IT8 from the Epson and compare that to the custom TDF of the Epson print. That's a LOT of work! By all means Mark, go for it.  ;D
Logged
Andrew Rodney
Author “Color Management for Photographers"

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12511
    • http://www.markdsegal.com
Re: Silverfast Advanced Targets ?
« Reply #61 on: April 25, 2018, 05:03:11 pm »

Getting back to accuracy, it's clear to me, IF I scanned the Epson 3880 print correctly that one scanner profile has a wider color gamut than the other. As do the scans! So if that's true, the lower gamut input profile alone is going to be less 'accurate' if our goal is to capture what's on the print.

Now comparing a scan of the target to the TDF does tell us about accuracy within the gamut of the profile and target. But then what happens if we place a wider gamut original onto the scanner? The printer I used has a pretty darn wide color gamut. The scanner really doesn’t have a color gamut but the profile from the target used to build that profile does and it's 'limited'. So color accuracy here would need to define what the scanner can capture no? And with the Standard target, I can't capture as wide a gamut from a print than I can with the Advanced target; again IF (big if) my scan settings are correct.

What one might have to do is create a custom target that mimics both of the IT8's but printed on a much wider gamut output device. These are RA4 prints right? An Epson has a wider color gamut.
We'd need to measure each patch and produce a custom TDF and maybe build a profile from that if possible. At the very least, we could use the profiles from those supplied by LaserSoft, scan our custom IT8 from the Epson and compare that to the custom TDF of the Epson print. That's a LOT of work! By all means Mark, go for it.  ;D

The gamut volume of the scanner profile I think it created for me (that damn naming business) is over 2.3 million according to CTP. The widest gamut volume I have achieved out of any inkjet printer, close to 1 million, is using Ilford Gold Fibre Silk (or similar) in my Epson SC-P5000 - which can print a much wider gamut than an Epson 3880. So roughly 1 million is the ultimate constraint on usable gamut volume if one is scanning in order to make prints from the scans. Ignoring gamut shape, which is OK - we're talking about the critical constraint being less than half the potential input volume. Back-up to the media being scanned - what do you think the equivalent gamut volume is of a colour negative or a colour transparency or a reflective image? I'll bet it's nowhere close to 2.3 million, so I don't think one needs to worry about this. Just to give you a feel for it, here is a view of the unique colours in the LSI advanced reflective target (seen as points) mapped against the gamut volume of the custom scanner profile. The word "unmanaged" means I did not impose an RGB colour space on the scan. I think it speaks for itself. So I'm not concerned about any accuracy measurements being impacted by gamut constraints when comparing a scan of the target to the TDF all staying  within the gamut of the profile.

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

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12511
    • http://www.markdsegal.com
Re: Silverfast Advanced Targets ?
« Reply #62 on: April 25, 2018, 05:14:39 pm »

Here's another example that gives me more confidence. This is a very saturated version of a CC target (to be used for a forthcoming article on printer stress testing) whose colours come very close to the boundaries of the widest printer/paper profiles I have ever created, a print of which I scanned and plotted against the scanner profile. You can see that even for this there is no contest - you can throw just about anything at that profile and it won't clip.
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 15630
    • http://www.digitaldog.net/
Re: Silverfast Advanced Targets ?
« Reply #63 on: April 25, 2018, 05:58:12 pm »

The gamut volume of the scanner profile I think it created for me (that damn naming business) is over 2.3 million according to CTP. The widest gamut volume I have achieved out of any inkjet printer, close to 1 million, is using Ilford Gold Fibre Silk (or similar) in my Epson SC-P5000 - which can print a much wider gamut than an Epson 3880.
I think you're thinking of this in an odd way. I'll tell you why.
First, gamut volume without context isn't that useful. Suppose I tell you that Joe weights 185 pounds while Sam is 170. Until you find out that Sam is 5 feet tall, or Joe is 6 foot five, you don't know which is over or under weight.
Next, these are just containers of a fixed size. Without pixels, they contain no useful data.
Now let's look at what I've done and hopefully with correct scanner settings (I'll know more tomorrow).
I have two scanner profiles and the Advanced Target has a larger color gamut than the Standard Target but what happens when the rubber meets the road; I make two identical scans and plot that data! The actual scanned data from the Advanced Target has a larger color gamut than the same scan settings made with the Standard Target. That's simply factual if we are to believe the 3D gamut plots of the actual scanned data in CTP. I've provided a video of the two.
So accuracy aside, the print scanned with the Advanced Target has a larger color gamut than the Standard Target. Doesn't matter what the gamut volume reported for two ICC profiles state or the gamut volume of a printer profile.
Do you have CTP (specially Pro) and know how to load images, build a color list and plot that? That's what I suggest you do and with a pretty low rez scan unless you have hours for CTP to plot this data. The only difference should be what input profile is selected and you shouldn’t scan into an RGB working space. Again, it's very possible my settings are wrong but that doesn’t explain how one scan has a larger color gamut than the other.
« Last Edit: April 25, 2018, 06:09:20 pm by andrewrodney »
Logged
Andrew Rodney
Author “Color Management for Photographers"

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 15630
    • http://www.digitaldog.net/
Re: Silverfast Advanced Targets ?
« Reply #64 on: April 25, 2018, 06:00:11 pm »

Here's another example that gives me more confidence. This is a very saturated version of a CC target (to be used for a forthcoming article on printer stress testing) whose colours come very close to the boundaries of the widest printer/paper profiles I have ever created, a print of which I scanned and plotted against the scanner profile. You can see that even for this there is no contest - you can throw just about anything at that profile and it won't clip.
SCAN the target, plot the scan. NOT the gamut of the printer profile; the print which was scanned. Big difference.
Logged
Andrew Rodney
Author “Color Management for Photographers"

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12511
    • http://www.markdsegal.com
Re: Silverfast Advanced Targets ?
« Reply #65 on: April 25, 2018, 07:10:06 pm »

I think you're thinking of this in an odd way. I'll tell you why.
First, gamut volume without context isn't that useful. Suppose I tell you that Joe weights 185 pounds while Sam is 170. Until you find out that Sam is 5 feet tall, or Joe is 6 foot five, you don't know which is over or under weight.
Next, these are just containers of a fixed size. Without pixels, they contain no useful data.
Now let's look at what I've done and hopefully with correct scanner settings (I'll know more tomorrow).
I have two scanner profiles and the Advanced Target has a larger color gamut than the Standard Target but what happens when the rubber meets the road; I make two identical scans and plot that data! The actual scanned data from the Advanced Target has a larger color gamut than the same scan settings made with the Standard Target. That's simply factual if we are to believe the 3D gamut plots of the actual scanned data in CTP. I've provided a video of the two.
So accuracy aside, the print scanned with the Advanced Target has a larger color gamut than the Standard Target. Doesn't matter what the gamut volume reported for two ICC profiles state or the gamut volume of a printer profile.
Do you have CTP (specially Pro) and know how to load images, build a color list and plot that? That's what I suggest you do and with a pretty low rez scan unless you have hours for CTP to plot this data. The only difference should be what input profile is selected and you shouldn’t scan into an RGB working space. Again, it's very possible my settings are wrong but that doesn’t explain how one scan has a larger color gamut than the other.

Yes of course, I have and use routinely ColorThink Pro and I'm very familiar with building and using their colour worksheets, extracting colours from image files, plotting imgae data against profile volume,m etc, etc, and if I run into arcane issues where I think there's yet another trick in that rich application I phone Pat or Steve for a bit of help which they are wonderful at giving, so we can comfortably set that question aside.

We can also set aside concern about "oddity" in what I'm doing, because I know exactly what I'm doing and it's very straightforward - I know both the size and shape of the "container" from CTP's graph and statistics for the profile. I can then see whether the image data over-spills any boundaries of the profile. If it does, I known I have a gamut compression issue on my hands; if it doesn't, I know the profile is not going to constrain anything smaller than what I mapped against it, which is massive in the last screen grab I posted above.

I have no problem with what you are proposing, but that is addressing a different question than the one I'm exploring. You are exploring how the previous and current LSI profiles compare, while I'm looking at whether the profile gamut of the current version is large enough to handle the media that could be thrown at it. Different matters, both good questions in their own right.

In the illustration I provided in my last illustrated post, the points in the diagram are a scan of the printed colour values of a colorchecker variant (using very wide gamut paper/printer) to be explained in a forthcoming article. Suffice to point out here what I said about its characteristics above. The most saturated color values in that print considerably exceed the most saturated patches in the LSI target. I'd use the LSI target as scan media for doing the round-tripping in the context of an accuracy test we were discussing above, but as a test to challenge what the scanner profile made from that target can accommodate by way of media to be scanned without clipping, the CC variant is a more interesting test print. Horses for courses.
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 15630
    • http://www.digitaldog.net/
Re: Silverfast Advanced Targets ?
« Reply #66 on: April 25, 2018, 07:17:33 pm »

Yes of course, I have and use routinely ColorThink Pro and I'm very familiar with building and using their colour worksheets, extracting colours from image files, plotting imgae data against profile volume,m etc, etc, and if I run into arcane issues where I think there's yet another trick in that rich application I phone Pat or Steve for a bit of help which they are wonderful at giving, so we can comfortably set that question aside.
Gamut volume per se isn't telling; that's my point.
Quote
We can also set aside concern about "oddity" in what I'm doing, because I know exactly what I'm doing and it's very straightforward - I know both the size and shape of the "container" from CTP's graph and statistics for the profile. I can then see whether the image data over-spills any boundaries of the profile. If it does, I known I have a gamut compression issue on my hands; if it doesn't, I know the profile is not going to constrain anything smaller than what I mapped against it, which is massive in the last screen grab I posted above.
So how to explain that the newer Advanced target produces a wider gamut scan than the standard? Seems to be the case. I'm not plotting containers, I'm plotting the scans. The only logical difference I can come to is that the Advanced Target produces a wider gamut scan. Viewing the gamut of the profiles (printer or scanner) is useful to a degree. But again, when the rubber meets the road, IF my settings are correct, the newer target should produce a more accurate scan because it's not clipping colors that Standard Target must be clipping, no?
Logged
Andrew Rodney
Author “Color Management for Photographers"

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12511
    • http://www.markdsegal.com
Re: Silverfast Advanced Targets ?
« Reply #67 on: April 25, 2018, 07:36:47 pm »

Gamut volume per se isn't telling; that's my point. So how to explain that the newer Advanced target produces a wider gamut scan than the standard? Seems to be the case. I'm not plotting containers, I'm plotting the scans. The only logical difference I can come to is that the Advanced Target produces a wider gamut scan. Viewing the gamut of the profiles (printer or scanner) is useful to a degree. But again, when the rubber meets the road, IF my settings are correct, the newer target should produce a more accurate scan because it's not clipping colors that Standard Target must be clipping, no?

I'm using gamut volume in comparison to other data, and as such it's a component of a story that IS telling. There are reasons why CTP provides calculations of gamut volume - it's useful.

I think both approaches are valid - i.e. whether you get a wider gamut scan from one profile versus the other, or whether the size and shape of the profiles are different enough to be significant. One doesn't see the significance with that information alone, and perhaps that's what you are getting at, in which case I would agree; but I do not agree that knowing the size and shape of the profile gamut doesn't provide useful information. And your last line is exactly the point - the profile made from the newer target should enable a more accurate scan if it is not clipping colours that the profile made from the Standard target would clip. To see this a priori, the two approaches are useful. Well, let's see what you learn from the LSI folks tomorrow - that could be an interesting call.
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 15630
    • http://www.digitaldog.net/
Re: Silverfast Advanced Targets ?
« Reply #68 on: April 26, 2018, 10:35:46 am »

OK I got some clarification from Lasersoft
Man is their GUI super confusing. But I believe I've set the scans tested correctly with the exception of the profiles being embedded which isn't necessary for the gamut comparisons (I'm handing off Lab values).

First, one must first toggle Input to ColorSync THEN select the scanner profile then again select None for it to show up (grayed out) and be embedded. Otherwise, if one selects ColorSync, <RGB> and input to the scanner profile, embedded shows for no reason I can understand, ProPhoto RGB! See screen captures below.

<blockquote>
On Apr 26, 2018, at 8:05 AM, Jan-Willem Rossée <silverfast.com> wrote:

AR: I am curious why the scanner profiles isn't embedded ("profile to embed <none>"). Hopefully this is the correct setup such input is each scanner profile and that's it in terms of color space conversions.

Jan: That's because you chose CIELab under "Working space --> Output". Set this to RGB and the scanner profile will be embedded. Let me know if you'd still like to chat.

This is the correct setting as per your instructions. Input is grayed out, it MUST be selected FIRST (screen capture #2 below this one) for me to see the scanner profile embedded.
Again, Please verify that the settings just below are correct (they seem to be, confugiring them is far more difficult than it should be):


To even get the scanner profile to show up, one must select RGB for output, THEN the scanner profile, THEN ColorSync to None or ProPhoto RGB shows up:
« Last Edit: April 26, 2018, 10:42:54 am by andrewrodney »
Logged
Andrew Rodney
Author “Color Management for Photographers"

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 15630
    • http://www.digitaldog.net/
Re: Silverfast Advanced Targets ?
« Reply #69 on: April 26, 2018, 10:54:56 am »

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.
Here's the fix: In preferences, go to the auto tab and select Custom ICC profile name and then you'll be asked to name it!

Logged
Andrew Rodney
Author “Color Management for Photographers"

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12511
    • http://www.markdsegal.com
Re: Silverfast Advanced Targets ?
« Reply #70 on: April 26, 2018, 02:40:16 pm »

Here's the fix: In preferences, go to the auto tab and select Custom ICC profile name and then you'll be asked to name it!

Yes of course (says he) page 73 of my book - it's right there. Why didn't I think of this? Because in previous times I had it checked by default and never thought of it again. Then with application updates and preference resets it got unchecked at some point when I wasn't creating profiles. Me oh my. There shouldn't even be such a choice. One should be forced to name a custom profile as a matter of course. So thanks for reminding me  :-)
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12511
    • http://www.markdsegal.com
Re: Silverfast Advanced Targets ?
« Reply #71 on: April 26, 2018, 03:05:34 pm »

OK I got some clarification from Lasersoft
Man is their GUI super confusing. But I believe I've set the scans tested correctly with the exception of the profiles being embedded which isn't necessary for the gamut comparisons (I'm handing off Lab values).

First, one must first toggle Input to ColorSync THEN select the scanner profile then again select None for it to show up (grayed out) and be embedded. Otherwise, if one selects ColorSync, <RGB> and input to the scanner profile, embedded shows for no reason I can understand, ProPhoto RGB! See screen captures below.

<blockquote>
On Apr 26, 2018, at 8:05 AM, Jan-Willem Rossée <silverfast.com> wrote:

AR: I am curious why the scanner profiles isn't embedded ("profile to embed <none>"). Hopefully this is the correct setup such input is each scanner profile and that's it in terms of color space conversions.

Jan: That's because you chose CIELab under "Working space --> Output". Set this to RGB and the scanner profile will be embedded. Let me know if you'd still like to chat.

This is the correct setting as per your instructions. Input is grayed out, it MUST be selected FIRST (screen capture #2 below this one) for me to see the scanner profile embedded.
Again, Please verify that the settings just below are correct (they seem to be, confugiring them is far more difficult than it should be):


To even get the scanner profile to show up, one must select RGB for output, THEN the scanner profile, THEN ColorSync to None or ProPhoto RGB shows up:

Andrew,

Let's get Working Space-Output out of the way first: You have three choices: <RGB>, <Colorsync> and <Cie-Lab>.

If you select <RGB>, it means in effect that you aren't using that setting for anything but scanning, and it shows the embedded profile (under Embedded ICC Profiles) as the one you selected as the Internal Profile (which is Silverfastese meaning to the rest of us Color Working Space, such as ProPhoto ARGB(98) etc.).

If you select <ColorSync> it lets you select a printer profile (in Output/Printer)because this choice assumes you want to SoftProof your eventual print in SilverFast before scanning the media (for making scan settings copacetic with your intended print; the company philosophy is that you should be able to do everything you need to do in SilverFast to get a fine print from a scan).

If you select <CIE-Lab> it means you don't want to embed any profile, so it makes "Profile to embed" = <none>.

The only places that control whether a scanner profile can show up are as I said further above: your selection at Input-Working Space where you have two choices: <colorsync> or <none>.

If you select <colorsync> as Input-Working Space, then under Profiles Input you can select your scanner profile.

If you select <none> as Input-Working Space, it disables being able to select a scanner profile.

That's pretty much "all there is too it". :-)
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 15630
    • http://www.digitaldog.net/
Re: Silverfast Advanced Targets ?
« Reply #72 on: April 26, 2018, 03:10:43 pm »

Andrew,
Let's get Working Space-Output out of the way first: You have three choices: <RGB>, <Colorsync> and <Cie-Lab>.
I asked LaserSoft but haven't heard back yet: why do we need any of the color management dropdown boxes at all (Input> working space, working> space monitor, working space> Output) with the three options (ColorSync, None or in one, CieLab)? Seems pointless. Everything seems to be set-able below.
Moving on, here's more oddity.

Now examine this one example below.
Input set to scanner profile (fine).
Internal set to Adobe RGB (1998) which is fine.
Embedded profile set to AR's 3880EFP. Where on earth did you get that from? Output/printer is set to none. That's just wrong (or again, massively confusing). Makes no sense. And the little video they provide is of really no help in figuring out what is really going on here.
Logged
Andrew Rodney
Author “Color Management for Photographers"

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 15630
    • http://www.digitaldog.net/
Re: Silverfast Advanced Targets ?
« Reply #73 on: April 26, 2018, 03:13:28 pm »

Further comments to them, maybe you can explain:

Why does the user even need a None setting instead of ColorSync or ICM? Without them, the graying out of the profiles section wouldn't be an issue either. I seriously cannot see why anyone would need to alter anything in this area but please enlighten me.

I asked you this in the last post, forgive all this copy and past to them....

Input should only show the possible scanner profiles. Which it somewhat appears to do (I see input profiles that are not associated with the scanner but that's to be expected). So that dropdown is fine.

Internal dropdown makes zero sense to me. Internal means what? It shows RGB working space, call it that. I believe what you're asking for is the RGB color space for conversions FROM input such as Adobe RGB (1998) etc.

Gray I presume is only accessed if and when someone makes a grayscale scan, that's fine.

So that leaves output. I see both RGB and CMYK output spaces so it's again confusing since you have 'Internal' which doesn't gray out IF someone selects a printer profile. IOW, I can have Interal set to Adobe RGB (1998) and Output set to some CMYK output space. Are you really converting from Scanner RGB ( Input) to RGB working space (Adobe RGB (1998)) to output (CMYK)? Seems utterly pointless. You can convert from Scanner RGB to Output directly no? If so, then gray out Output/Printer when the user selects a profile in Input. Gray out Input if the user selects an output color space in Printer.

Even more odd, If I select Input (where you actually use the correct term, RGB working space not Input) to none, Input grays out; why? Again I don't see the need for that entire area. Next, if I do select ColorSync there, I can select the scanner profile below, and with Internal, gray and output set to none, I'm shown an output profile being embedded. Bug? No difference if I now select Adobe RGB (1998) in Internal.

Then you have Rendering Intent. I can pick Adobe RGB (1998) and Perceptual but you're not going to produce that conversion as there's no such table in these simple profiles. IF Input is set to Adobe RGB (1998), and Output is set to None, that drop down should gray out yet it doesn't.


And people say Photoshop's color settings are difficult to understand?  ;)
Logged
Andrew Rodney
Author “Color Management for Photographers"

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12511
    • http://www.markdsegal.com
Re: Silverfast Advanced Targets ?
« Reply #74 on: April 26, 2018, 03:30:13 pm »


Embedded profile set to AR's 3880EFP. Where on earth did you get that from? Output/printer is set to none. That's just wrong (or again, massively confusing). Makes no sense. And the little video they provide is of really no help in figuring out what is really going on here.

"I" didn't get that at all. The application on your computer did. !!  :-).

But I tried to replicate the result your copy of the application pulled up and I understand what happens. When you select <Colorsync> as Working Space-Ouput and you have in a previous session selected your 3880EFP as Output/Printer profile, the "Profile to embed" will show up as that printer profile. Now, if you thereafter change Output/Printer to <none>, the embedded profile remains as the previously selected printer profile, in your case 3880EFP, probably because it isn't programmed to know what to do once you disable "Output/Printer" by selecting <none> in that box. Now please don't shoot the messenger. I'm not saying this makes any generic sense, but that is simply a description of how the application behaves in these circumstances.

As to why you have all those choices under "Color Management": They are giving the user full control over the color management set-up in one place: choice of working space, choice of monitor profile, choice of scanner profile, choice of printer profile. Once you've made those choices in this section, the second section underneath it then lets you select profiles according to the logic of what you selected in the upper section. The final section is confirming information about the embedded working space emerging from those choices.
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12511
    • http://www.markdsegal.com
Re: Silverfast Advanced Targets ?
« Reply #75 on: April 26, 2018, 03:37:12 pm »

Further comments to them, maybe you can explain:

Why does the user even need a None setting instead of ColorSync or ICM? Without them, the graying out of the profiles section wouldn't be an issue either. I seriously cannot see why anyone would need to alter anything in this area but please enlighten me.

I asked you this in the last post, forgive all this copy and past to them....

Input should only show the possible scanner profiles. Which it somewhat appears to do (I see input profiles that are not associated with the scanner but that's to be expected). So that dropdown is fine.

Internal dropdown makes zero sense to me. Internal means what? It shows RGB working space, call it that. I believe what you're asking for is the RGB color space for conversions FROM input such as Adobe RGB (1998) etc.

Gray I presume is only accessed if and when someone makes a grayscale scan, that's fine.

So that leaves output. I see both RGB and CMYK output spaces so it's again confusing since you have 'Internal' which doesn't gray out IF someone selects a printer profile. IOW, I can have Interal set to Adobe RGB (1998) and Output set to some CMYK output space. Are you really converting from Scanner RGB ( Input) to RGB working space (Adobe RGB (1998)) to output (CMYK)? Seems utterly pointless. You can convert from Scanner RGB to Output directly no? If so, then gray out Output/Printer when the user selects a profile in Input. Gray out Input if the user selects an output color space in Printer.

Even more odd, If I select Input (where you actually use the correct term, RGB working space not Input) to none, Input grays out; why? Again I don't see the need for that entire area. Next, if I do select ColorSync there, I can select the scanner profile below, and with Internal, gray and output set to none, I'm shown an output profile being embedded. Bug? No difference if I now select Adobe RGB (1998) in Internal.

Then you have Rendering Intent. I can pick Adobe RGB (1998) and Perceptual but you're not going to produce that conversion as there's no such table in these simple profiles. IF Input is set to Adobe RGB (1998), and Output is set to None, that drop down should gray out yet it doesn't.


And people say Photoshop's color settings are difficult to understand?  ;)

Andrew, I'm assuming from all this that you did not have your conference call with them today as originally intended; is that correct? Because if you did they probably would have been able to address all these questions. I have an appointment outside the house shortly, so I cannot get into working out all the issues you are raising here just now, and in any case it would be good to see the feedback from Kiel. So how about we do this: wait and see how they answer these questions - please do fill us in here with what you learn from them, and then if there remain any residuals that need further discussion we can get back down to it tomorrow, either here in the first instance, or bilaterally over Skype and Teamviewer and then feedback the conclusions here for whatever audience is looking in on all this discussion.

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

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 15630
    • http://www.digitaldog.net/
Re: Silverfast Advanced Targets ?
« Reply #76 on: April 26, 2018, 03:48:20 pm »

But I tried to replicate the result your copy of the application pulled up and I understand what happens. When you select <Colorsync> as Working Space-Ouput and you have in a previous session selected your 3880EFP as Output/Printer profile, the "Profile to embed" will show up as that printer profile. Now, if you thereafter change Output/Printer to <none>, the embedded profile remains as the previously selected printer profile, in your case 3880EFP, probably because it isn't programmed to know what to do once you disable "Output/Printer" by selecting <none> in that box. Now please don't shoot the messenger. I'm not saying this makes any generic sense, but that is simply a description of how the application behaves in these circumstances.
That's just messed up Mark. The settings are the settings. When no profile is selected in Output/Printer as shown, there should be no profile that was selected at any time showing up.



Quote
As to why you have all those choices under "Color Management": They are giving the user full control over the color management set-up in one place: choice of working space, choice of monitor profile, choice of scanner profile, choice of printer profile. Once you've made those choices in this section, the second section underneath it then lets you select profiles according to the logic of what you selected in the upper section. The final section is confirming information about the embedded working space emerging from those choices.
Again, that make sense to have two areas. Color management and then Profiles. And depending on what you do in the upper area (color management), stuff below gets grayed out. Overly complicated. ColorSync and/or RGB for Output, why? Remove it all; then in Internal (which should be called working space), pick an RGB working space; period. Ditto for Output/Printer.
How they could take a Photoshop concept and make it ever more complicated and redundant is beyond me.


Four dropdown menu's only:
  • Input (select scanner profile).
  • Internal (call it RGB working space and allow the option to pick RGB working spaces). Gray out Output/Printer as you're asking to scan into an RGB working space. Gray out Rendering Intent, you can't alter RI anyway.
  • Gray (for grayscale profiles IF scanning in gray).
  • Output/Printer (only printer profiles, gray out Internal when selected).
Embed ICC Profile: no need, embed what the user selects above. If you want to show that, fine.
Logged
Andrew Rodney
Author “Color Management for Photographers"

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 15630
    • http://www.digitaldog.net/
Re: Silverfast Advanced Targets ?
« Reply #77 on: April 26, 2018, 03:52:11 pm »

Andrew, I'm assuming from all this that you did not have your conference call with them today as originally intended; is that correct?
Correct, only a series of emails from Jan.
Logged
Andrew Rodney
Author “Color Management for Photographers"

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12511
    • http://www.markdsegal.com
Re: Silverfast Advanced Targets ?
« Reply #78 on: April 26, 2018, 03:54:59 pm »

I'll be interested to read their feedback on your suggestions.
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 15630
    • http://www.digitaldog.net/
Re: Silverfast Advanced Targets ?
« Reply #79 on: April 26, 2018, 03:58:33 pm »

Here's all we need. In this example, pick scanner profile, desire to scan into RGB working space and embed profile (no option NOT to, just informational):
Note: gray could be configured only accessed when scanning gray but now Embed profile info is iffy.
RI grayed out when picking RGB working spaces. (wish PS would do so too).
Pick Output/Printer, RGB working space grays out, RI is accessible.
Done.
Logged
Andrew Rodney
Author “Color Management for Photographers"
Pages: 1 2 3 [4] 5 6   Go Up