[Tool] ChromIQ – a macOS and Windows GUI for ArgyllCMS printer profiling (v4.2.0)

[know]

@know,

I just downloaded the latest beta version to have a look at. Did you that pulling down the preset menu in Windows (not sure it is Windows specific) but the scroll down bar does not work: Wen I try to select the pull down bar to scroll down it chooses the first preset immediately. It seems it is static when using my mouse cursor (pressing down on the pull down bar and trying to scroll downwards does not work).

So I can only use my wheel to scroll down, it does work without any problems. Trying to push on the scroll down bar on the right of the pull down menu freezes and the system chooses the first preset and throw me out of the preset menu.

Thanks for the report... i will report the bug.
Did you get to open the "Settings for built-in presets" window, the gear-symbol button to the right of the "Select preset" field, in manual mode of Create Chart? That is the list you can select what you want as default, then export the list and give it back to me.

Also, when testing to use the preset pulldown list, or the "built-in presets" button (the middle button at the top right in the Create Chart tab), did you first load or create a project? I now there was some bugs also when no project was loaded.

 

[know]

Hi @cosinus ,
Glad you like it.

If you have a small screen, the only thing to do currently is either to increase resolution (if possible), or hide the log panel...It is a bit difficult due to all the settings, I know... We have not found a better solution than this without a major change in the layout.

View attachment 19291

Hi again @cosinus ,

From the next release (from beta 44) the fames called "Output" and "Select preset", in the Manual mode of Create Chart, will be able to collapse when needed. Thus, users like you, with a smaller screen can increase space for scrolling the layout options when needed.

Cheers,
Knut

 

[Im-back-Arno]

Dag Arno,

If needed you can cramp up to 918 patches on a single sheet of A4 (but you cannot press the pamper under the clamp of the jig for the first 4 rows as it needs the whole surface:

I have made a single sheet 918 patch target on A4 for the i1Profiler using the superior patch generator of ChromIQ emphasizing the neutral and near neutral greys:

View attachment 19271

I have measured the patch width and it is 7.2 mm (patch height is 7.5 mm).

Taking into account the headway for the clamp area (when completely inserting the paper onder the clamp and aligning completely to the right border of the scanning table) then 744 patches (24 rows of 31 patches) is about the maximum you can get for A4 (important: choose a smaller separator bar with -A0.5), but then you have to scan it slowly and steadily. A safer and more comfortable amount is 648 (24 rows of 27 patches).

I hope this will be helpful for you.

Dag Pharmacist,

Trying to keep up with this active (brilliant) thread in between travels. Got up at 2:30 this morning to catch a 6 AM flight so can’t do anything useful today but try to be helpful here.

Regarding your post and the following discussions around what a good preset set is: How is it determined what a good preset is? Presumably one that has been tested for accuracy and ease of use(?). How is color accuracy and efficiency of a patch set tested? By us users? Without wanting to sound critical or negative, it seems to me that the easy of creating patches possibly leads to too many presets without clarity (to me) on their accuracy (?). Not sure if this novice interpretation is correct, but I wonder why one would want a “quick”, “medium” and “expert” version of a patch set. Personally, if I take the effort of profiling a paper, I would always want to use the most accurate one, no?

I do second the comment by one user here that likely two batches might be needed: One optimized to provide the highest color accuracy, and one optimized for perfectly neutral B&W. For me, only the “best/expert” patch set would be needed: I do it once so I better do it as good as it can be, even when it takes 3 or 4 pages. As a control versus time/drift, a one-pager will probably suffice (?). Hence, are “handicapped” presets really needed? Who is going to use that? (Just my personal perfectionist mind speaking here).

@knut: Thanks for the heads-up that I can calibrate my scanner with very high accuracy, but “scanning” in my case means digitizing a 35mm negative or positive with silly resolution using a digital camera and a high resolution dedicated lens that I pulled from a Minolta slide scanner and stuck onto a custom-built rig (see the link in post #645, which I will not link again out of fear of being banned again). So scanning in my workflow means digitizing a negative, and yes, I do calibrate color or neutral B&W with an also custom-built trichromatic light source with narrow-band R, G, and B LEDs to simulate R4(?) paper response (see my blog).

In short, for me personally, one “perfect” patch set for color and one “perfect” patch set for B&W, both on A4, would suffice, provided that they are optimized (and verified) to provide optimal profiles. How “optimal” is defined is accurate color and neutral B&W, respectively. Ideally they would be accompanied by a one pager each which would probe the extremes to quickly determine any drift with time/inkload/paper batch/etc.

As it is today, I grab a random profile with a lot of patches and many greys (for B&W), and get a profile which seems to work, but I have no clue if another patch set will give me a better profile. As a novice or end-user I would like an indication by the experts like: Pick this one for the most accurate color profiling or most neutral B&W profiling, respectively.

Just my thoughts, intended constructively!

Arno

 

[know]

Hi @pharmacist,
ChromIQ Version 4.3.0-beta-44 is out. You can now see how the preset list looks like, and it seems to work nicely.
There is considerable less profiles in the list now, and it is configurable for anyone to change

Also, I know you are helping to make your own suggested charts into our built-in chart presets, so that they can load with all settings and the patch set engine settings used... You asked about adding an image as the clip-border. Just wanted to say that in the Expert Options section in the Manual mode Create Chart tab, there is the Clip-border content frame that contains settings to customise the clip-border. Under the pulldown list for "Content" you can select "Imported Image" to add an image. You know this, but what is new is that you can select "ChromIQ branding", which adds the ChromIQ logo, and then you can add any text that you want in the text field, and those "Inserts" you can add, for page number, patch size, or whatever, they are updated automatically. If you add an image, it is a static thing.

As an example, Select an i1Pro preset (all of them use a clip-border Notes box). Go to Expert Options section in the Manual mode Create Chart tab and to Clip-border content frame (bottom of the list). Select the Content option "Custom text example", which puts an example into the text field. Then select "ChromIQ branding" and click Generate chart.

This can then look like this (I modified the example text a little):

And @cosinus ,
you now will be able to get more space to scroll the chart layout parameters Output and Presets frames can be collapsed.

Cheers,
Knut

 

[pharmacist]

Dag Pharmacist,

Trying to keep up with this active (brilliant) thread in between travels. Got up at 2:30 this morning to catch a 6 AM flight so can’t do anything useful today but try to be helpful here.

Regarding your post and the following discussions around what a good preset set is: How is it determined what a good preset is? Presumably one that has been tested for accuracy and ease of use(?). How is color accuracy and efficiency of a patch set tested? By us users? Without wanting to sound critical or negative, it seems to me that the easy of creating patches possibly leads to too many presets without clarity (to me) on their accuracy (?). Not sure if this novice interpretation is correct, but I wonder why one would want a “quick”, “medium” and “expert” version of a patch set. Personally, if I take the effort of profiling a paper, I would always want to use the most accurate one, no?

I do second the comment by one user here that likely two batches might be needed: One optimized to provide the highest color accuracy, and one optimized for perfectly neutral B&W. For me, only the “best/expert” patch set would be needed: I do it once so I better do it as good as it can be, even when it takes 3 or 4 pages. As a control versus time/drift, a one-pager will probably suffice (?). Hence, are “handicapped” presets really needed? Who is going to use that? (Just my personal perfectionist mind speaking here).

@knut: Thanks for the heads-up that I can calibrate my scanner with very high accuracy, but “scanning” in my case means digitizing a 35mm negative or positive with silly resolution using a digital camera and a high resolution dedicated lens that I pulled from a Minolta slide scanner and stuck onto a custom-built rig (see the link in post #645, which I will not link again out of fear of being banned again). So scanning in my workflow means digitizing a negative, and yes, I do calibrate color or neutral B&W with an also custom-built trichromatic light source with narrow-band R, G, and B LEDs to simulate R4(?) paper response (see my blog).

In short, for me personally, one “perfect” patch set for color and one “perfect” patch set for B&W, both on A4, would suffice, provided that they are optimized (and verified) to provide optimal profiles. How “optimal” is defined is accurate color and neutral B&W, respectively. Ideally they would be accompanied by a one pager each which would probe the extremes to quickly determine any drift with time/inkload/paper batch/etc.

As it is today, I grab a random profile with a lot of patches and many greys (for B&W), and get a profile which seems to work, but I have no clue if another patch set will give me a better profile. As a novice or end-user I would like an indication by the experts like: Pick this one for the most accurate color profiling or most neutral B&W profiling, respectively.

Just my thoughts, intended constructively!

Arno

Dag Arno,

Profile quality is determined in the first place by visual comparison with a print made with another proifle. The most objective method is to use a verification scan with a target (there are some small patch count preset targets in ChromIQ, especially made for this. In the beginning I was surprised these small patch count presets exist in ChromIQ as for a satisfactory profile at least 300 patches are needed, until @know pointed me at this and that those small patch count presets are meant for verifcation scans) with known colours and printed on the same paper with the profile made for this paper and determine (by scanning with your spectro) the Delta-E (that in ideal circumstances should be 0 or near zero).

How about the quality: @Alan G and @hackensacker both have confirmed that my 3-page 1944 patch count preset does a very good job in generating good quality profiles (especially good neutral greys with good details in the highlights and shadows) without the need of the standard workflow starting with a primary target in ArgyllCMS to determine the neutral axis for a given paper/ink/printer combination. The reason is that in most cases the exact neutral position of the neutral axis is not exactly R=G=B and often deviates a bit (serpentine behaviour) from pure white to pure black. The primary target can determine this behaviour with help of the RGB raster as an anker. With this preconditioning profile we generate a secondary target and you will see that in most cases the "grey" patches (neutral steps) are now teinted (magenta, brownish, bluish or greenish hue) on your screen: this is the dermined measured neutral axis for your printer/ink/paper combination.

The 1944 target I have made uses a double ring near neutral ring around the pure neutral axis: any deviations from the pure neutral grey axis (swing) is thus effectively captured and determined to compensate from an eventual off-set behaviour from the pure neutral grey axis. This way you don't need a two-step method (preconditioning + final profile) to get the same results. I believe @Alan G has compared his previous modus operandi of preconditioning + final profile (using about 4 sheets of Letter with more patches in total) and my one-step 1944 target and he concluded that both yield the same quality profile with the one-step 1944 target doing slightly better: faster (no 2 days waiting) and consuming less paper and ink. I have taken this idea from the SpyderPrint Extended grays target: despite it's terrible hardware quality (problems with OBA-based papers) the underlying concept is actually very good: This extended grays target to neutralise black and white prints is based on a double near neutral ring around the grey axis and I have found to be very effective, even for printers using cheap aftermarket inks (like my Canon TS705a with no grey ink cartridge): this printer prints with a terrible sepia cast in the greys without a profile and even with a standard profile you can still see a slight sepia hue in the greys. With a double near neutral ring target this is completely eradicated. I took this extended grays concept in the sampling of my 1944 target to a higher level (more steps). The lower patch targets uses a single near neutral ring: the famous TC9.18 target uses a single near neutral ring. Still very good (medium quality), but 2 rings is much better (expert quality).

Printers with already 1 or 2 grey inks included (e.g. ET-8550 and SC-P900) will do well with a single near neutral ring target, whereas printers without any grey ink (like your R1900) probably needs a double near neutral ring target to get the best results.

Those simpler smaller patch count presets are for "standard" quality targets. This probably gives a quality similar to the Colormunki/i1Studio 50 + 50 patch workflow or the i1Profiler standard 400 patch count target. It is still a huge improvement compared to printing without a profile (especially if you are using aftermarket inks). This is the reason why ChromIQ has them as presets.

 

[pharmacist]

Today I tested my 1944 patch target for my i1Pro1 (vintage Eye-One Pro) that I previously read with my i1Pro2:

Head way is perfect: starts on white and ends on white:

Because the patch height is slightly lower (7.55 mm) than the 9 mm gap of the i1Pro1 scanning jig, the patch markers on the left are very handy to center the gap as good as possible as a visual anker to align the row as good a possible in the middle of the scanning path way of the spectro.

I did this test to see if the target can be used with both the i1Pro1 and the i1Pro2 scanning jig/board and it past the test. Here is print screen of scanning a few rows with this setup:

Same target under the i1Pro2 scanning jig/board:

Again the patch markers are very handy to align the ruler exactly on the border between 2 patches:

All these improvements lifts ArgyllCMS to a new user friendly level, not seen with the command based ArgyllCMS. I hope that @Graeme Gill , the father of ArgyllCMS is following this topic as well.

 

[know]

Today I tested my 1944 patch target for my i1Pro1 (vintage Eye-One Pro) that I previously read with my i1Pro2:

View attachment 19304

Head way is perfect: starts on white and end on white:

View attachment 19305

Because the patch height is slightly lower (7.55 mm) than the 9 mm gap of the i1Pro1 scanning jig, the patch markers on the left are very handy to center the gap as good as possible as a visual anker to align the row as good a possible in the middle of the scanning path way of the spectro.

I did this test to see if the target can be used with both the i1Pro1 and the i1Pro2 scanning jig/board and it past the test. Here is print screen of scanning a few rows with this setup:

View attachment 19306

Same target under the i1Pro2 scanning jig/board:

View attachment 19307

Again the patch markers are very handy to align the ruler exactly on the border between 2 patches:

View attachment 19308

Thanks for that test @pharmacist !
Great to see the thinking done was not wrong

 

[know]

Dag Arno,

Profile quality is determined in the first place by visual comparison with a print made with another profile. The most objective method is to use a verification scan with a target (there are some small patch count preset targets in ChromIQ, especially made for this. In the beginning I was surprised these small patch count presets exist in ChromIQ as for a satisfactory profile at least 300 patches are needed, until @know pointed me at this and that those small patch count presets are meant for verifcation scans) with known colours and printed on the same paper with the profile made for this paper and determine (by scanning with your spectro) the Delta-E (that in ideal circumstances should be 0 or near zero).

How about the quality: @Alan G and @hackensacker both have confirmed that my 3-page 1944 patch count preset does a very good job in generating good quality profiles (especially good neutral greys with good details in the highlights and shadows) without the need of the standard workflow starting with a primary target in ArgyllCMS to determine the neutral axis for a given paper/ink/printer combination. The reason is that in most cases the exact neutral position of the neutral axis is not exactly R=G=B and often deviates a bit (serpentine behaviour) from pure white to pure black. The primary target can determine this behaviour with help of the RGB raster as an anker. With this preconditioning profile we generate a secondary target and you will see that in most cases the "grey" patches (neutral steps) are now teinted (magenta, brownish, bluish or greenish hue) on your screen: this is the dermined measured neutral axis for your printer/ink/paper combination.

The 1944 target I have made uses a double ring near neutral ring around the pure neutral axis: any deviations from the pure neutral grey axis (swing) is thus effectively captured and determined to compensate from an eventual off-set behaviour from the pure neutral grey axis. This way you don't need a two-step method (preconditioning + final profile) to get the same results. I believe @Alan G has compared his previous modus operandi of preconditioning + final profile (using about 4 sheets of Letter with more patches in total) and my one-step 1944 target and he concluded that both yield the same quality profile with the one-step 1944 target doing slightly better: faster (no 2 days waiting) and consuming less paper and ink. I have taken this idea from the SpyderPrint Extended grays target: despite it's terrible hardware quality (problems with OBA-based papers) the underlying concept is actually very good: This extended grays target to neutralise black and white prints is based on a double near neutral ring around the grey axis and I have found to be very effective, even for printers using cheap aftermarket inks (like my Canon TS705a with no grey ink cartridge): this printer prints with a terrible sepia cast in the greys without a profile and even with a standard profile you can still see a slight sepia hue in the greys. With a double near neutral ring target this is completely eradicated. I took this extended grays concept in the sampling of my 1944 target to a higher level (more steps). The lower patch targets uses a single near neutral ring: the famous TC9.18 target uses a single near neutral ring. Still very good (medium quality), but 2 rings is much better (expert quality).

Printers with already 1 or 2 grey inks included (e.g. ET-8550 and SC-P900) will do well with a single near neutral ring target, whereas printers without any grey ink (like your R1900) probably needs a double near neutral ring target to get the best results.

Those simpler smaller patch count presets are for "standard" quality targets. This probably gives a quality similar to the Colormunki/i1Studio 50 + 50 patch workflow or the i1Profiler standard 400 patch count target. It is still a huge improvement compared to printing without a profile (especially if you are using aftermarket inks). This is the reason why ChromIQ has them as presets.

For information to everyone,
The charts @pharmacist made are shown in ChromIQ ending with the name "by Pharmicist".
"i1Pro · A4-1944p-3pages extended target by Pharmacist", or
"i1Pro · Letter-1944p-3pages extended target by Pharmacist"

These have a patch set that is a 10x10x10 cube with additional skin tones and some random patches. The extra focus on skin tones make this profile probably better for printing with images that have a people in them.

Based on his page layout design, I made most other i1Pro presets to match, and in the process also tried to make a equivalent 1944 patch set (but larger are also available) that is good as a standard for most cases.

The result is the two presets named:
"i1Pro · A4-1944p-3pages-Portrait-w7.5mm" or
"i1Pro · Letter-1944p-3pages-Portrait-w7.5mm"

The patch sets on these are a clean 11x11x11 cube with defining edge patches, but no extra skin tones. The higher density cube is thus more uniform, and a good alaround profile.

Both pharmacist's profiles and mine have approximately same number of grey-ramp patches and the near-neutral rings.

Here is a comparison of the two patch sets, Pharmacist's on the left and the built-in preset version "i1Pro · A4-1944p-3pages-Portrait-w7.5mm" on the right:

A disadvantage of the profiles by pharmacist, which he is currently working on to fix, is that they do not contain any layout information (only images), which make them hard to use as basis for a new layout.

Here is the layout information from the Patch Set Editor ("Edit / Create Chart Patch Set" in the tools menu) for the "i1Pro · A4-1944p-3pages-Portrait-w7.5mm" preset chart:

 

[know]

Dag Pharmacist,

Trying to keep up with this active (brilliant) thread in between travels. Got up at 2:30 this morning to catch a 6 AM flight so can’t do anything useful today but try to be helpful here.

Regarding your post and the following discussions around what a good preset set is: How is it determined what a good preset is? Presumably one that has been tested for accuracy and ease of use(?). How is color accuracy and efficiency of a patch set tested? By us users? Without wanting to sound critical or negative, it seems to me that the easy of creating patches possibly leads to too many presets without clarity (to me) on their accuracy (?). Not sure if this novice interpretation is correct, but I wonder why one would want a “quick”, “medium” and “expert” version of a patch set. Personally, if I take the effort of profiling a paper, I would always want to use the most accurate one, no?

I do second the comment by one user here that likely two batches might be needed: One optimized to provide the highest color accuracy, and one optimized for perfectly neutral B&W. For me, only the “best/expert” patch set would be needed: I do it once so I better do it as good as it can be, even when it takes 3 or 4 pages. As a control versus time/drift, a one-pager will probably suffice (?). Hence, are “handicapped” presets really needed? Who is going to use that? (Just my personal perfectionist mind speaking here).

@knut: Thanks for the heads-up that I can calibrate my scanner with very high accuracy, but “scanning” in my case means digitizing a 35mm negative or positive with silly resolution using a digital camera and a high resolution dedicated lens that I pulled from a Minolta slide scanner and stuck onto a custom-built rig (see the link in post #645, which I will not link again out of fear of being banned again). So scanning in my workflow means digitizing a negative, and yes, I do calibrate color or neutral B&W with an also custom-built trichromatic light source with narrow-band R, G, and B LEDs to simulate R4(?) paper response (see my blog).

In short, for me personally, one “perfect” patch set for color and one “perfect” patch set for B&W, both on A4, would suffice, provided that they are optimized (and verified) to provide optimal profiles. How “optimal” is defined is accurate color and neutral B&W, respectively. Ideally they would be accompanied by a one pager each which would probe the extremes to quickly determine any drift with time/inkload/paper batch/etc.

As it is today, I grab a random profile with a lot of patches and many greys (for B&W), and get a profile which seems to work, but I have no clue if another patch set will give me a better profile. As a novice or end-user I would like an indication by the experts like: Pick this one for the most accurate color profiling or most neutral B&W profiling, respectively.

Just my thoughts, intended constructively!

Arno

Thanks for your thoughts @Im-back-Arno
I think it is always a matter of making the best without going to the most extremes. Generally, the more patches the better, but what you get in improvement is less and less the higher it goes. Some swear to closer to 3000 patches, maybe because they can and have an i1iSis for A3 pages. We were contacted by a lady from Red River, where we added their patch set into the presets in ChromIQ. Their standard patch set to profile all their papers is 2052 patches, so that might be an indicator in the industry.

Take also a look at this post, where I compare @pharmacist's 1944 patch preset with my own that is in the list of presets: post-149535

To your comment on a one-pager for control: We have designed a workflow in ChromIQ that allows regular control measurements of a profile (we call it a verification run), which then outputs a report, from the tool "Measurement Report" in the tools menu.

Attached is an example report made on demo data, as I do not currently have any real data measured over time to get all the trend-graphs.
And here is a screenshot of the Measurement Report window for the attached report:

Sebastian and I hope the community can test this feature and report back the results... There are many options to customise limits for each metric checked, and there are 20 different metrics to choose from. Here is a view of the setup window that defines all the metrics and thresholds (all 20 does not fit into one screenshot, as the window scrolls):

The Help Card "Getting Started" in the Welcome to ChromIQ window (question mark icon on top-right corner of the app) gives an overview of the workflow in a diagram. I recommend having a look at that. Here is a snapshot of that diagram:

Cheers,
Knut

 

[Im-back-Arno]

Dag Arno,

Profile quality is determined in the first place by visual comparison with a print made with another proifle. The most objective method is to use a verification scan with a target (there are some small patch count preset targets in ChromIQ, especially made for this. In the beginning I was surprised these small patch count presets exist in ChromIQ as for a satisfactory profile at least 300 patches are needed, until @know pointed me at this and that those small patch count presets are meant for verifcation scans) with known colours and printed on the same paper with the profile made for this paper and determine (by scanning with your spectro) the Delta-E (that in ideal circumstances should be 0 or near zero).

How about the quality: @Alan G and @hackensacker both have confirmed that my 3-page 1944 patch count preset does a very good job in generating good quality profiles (especially good neutral greys with good details in the highlights and shadows) without the need of the standard workflow starting with a primary target in ArgyllCMS to determine the neutral axis for a given paper/ink/printer combination. The reason is that in most cases the exact neutral position of the neutral axis is not exactly R=G=B and often deviates a bit (serpentine behaviour) from pure white to pure black. The primary target can determine this behaviour with help of the RGB raster as an anker. With this preconditioning profile we generate a secondary target and you will see that in most cases the "grey" patches (neutral steps) are now teinted (magenta, brownish, bluish or greenish hue) on your screen: this is the dermined measured neutral axis for your printer/ink/paper combination.

The 1944 target I have made uses a double ring near neutral ring around the pure neutral axis: any deviations from the pure neutral grey axis (swing) is thus effectively captured and determined to compensate from an eventual off-set behaviour from the pure neutral grey axis. This way you don't need a two-step method (preconditioning + final profile) to get the same results. I believe @Alan G has compared his previous modus operandi of preconditioning + final profile (using about 4 sheets of Letter with more patches in total) and my one-step 1944 target and he concluded that both yield the same quality profile with the one-step 1944 target doing slightly better: faster (no 2 days waiting) and consuming less paper and ink. I have taken this idea from the SpyderPrint Extended grays target: despite it's terrible hardware quality (problems with OBA-based papers) the underlying concept is actually very good: This extended grays target to neutralise black and white prints is based on a double near neutral ring around the grey axis and I have found to be very effective, even for printers using cheap aftermarket inks (like my Canon TS705a with no grey ink cartridge): this printer prints with a terrible sepia cast in the greys without a profile and even with a standard profile you can still see a slight sepia hue in the greys. With a double near neutral ring target this is completely eradicated. I took this extended grays concept in the sampling of my 1944 target to a higher level (more steps). The lower patch targets uses a single near neutral ring: the famous TC9.18 target uses a single near neutral ring. Still very good (medium quality), but 2 rings is much better (expert quality).

Printers with already 1 or 2 grey inks included (e.g. ET-8550 and SC-P900) will do well with a single near neutral ring target, whereas printers without any grey ink (like your R1900) probably needs a double near neutral ring target to get the best results.

Those simpler smaller patch count presets are for "standard" quality targets. This probably gives a quality similar to the Colormunki/i1Studio 50 + 50 patch workflow or the i1Profiler standard 400 patch count target. It is still a huge improvement compared to printing without a profile (especially if you are using aftermarket inks). This is the reason why ChromIQ has them as presets.

Dag Pharmacist,

Thanks for this. I’m still on travel so can’t play around with it, but your 1944 patch preset seems what I need then. I’ll look later today if I can find it in ChromIQ and run it when I’m back home and use a more limited patch set to verify its accuracy like you say.

So the advantage of your 1944 is that it’s a one-shot calibration that gives similar or even better accuracy than a two-step calibration (precondition plus final) if my interpretation is correct, which is ideal.

One thing that I did notice with my quick trials so far was that some patches in a row have a red box around them so I rescan that line of patches but not all of them disappear. Note that in my quick tests I did so far not let the print dry overnight, but used a hairdryer to speed things up. Could the red boxes appear due to insufficient drying time, or am I scanning too fast, or are the red boxes patches simply out of the gamut that my printer is able to produce?

Thanks again for all the help!

Arno