Luminous Landscape Forum

Raw & Post Processing, Printing => Colour Management => Topic started by: Doug Gray on August 19, 2017, 08:13:17 pm

Title: I1Isis 2 XL, the good, bad, and ugly
Post by: Doug Gray on August 19, 2017, 08:13:17 pm
Well, after getting tired trying to get more out of an I1Pro 2 than it perhaps can offer, I bought the XL. and am putting it though its paces to understand its capabilities as well as limits. Initial tests show a few things:

1. The instrument produces highly repeatable chart readings. The Isis default patch set, 957 patches, fits on letter size sheet. Successive reads produced average dEs of 0.04, and a max dE of .45 on glossy paper. This compares to about .20  with a max of 1.9 with the I1Pro 2 on a 918 patch set with two sheets. Much of this is likely due to the precision feeding and automatic registration of the Isis. So it is reading almost exactly the same locations on the patches. Reading patches printed on different sheets shifting the image slightly produces variations similar to the I1Pro.

2. I hand measured a set of 64 RGB patches evenly spread through the RGB gamut with a distance of 85 (see patchtool) on an Isis target and compared them to the Isis measurements. The average dE was .45 and max was .9.  This was better than I expected as my I1Pro 2 is over 5 years old and has never been factory calibrated. Remarkable, really.

Now for the bad news.

The Isis moves paper using bottom wheels spaced about 2 cm apart and top wheels that are spaced 4.2 cm apart and the top ones can leave "tracks." [added: The 4.2 cm "top wheels" are fixed and seem to be for the purpose of keeping the paper flat]  For glossy prints that have dried for 24 hours, the tracks are nearly invisible and there is no significant impact on the Isis readings. But for thick Luster papers the wheels appear to flatten the paper and create a region, about 1mm wide, where the colors appear, and measure, a more saturated color. For instance a light yellow patch (RGB 255,255,170) produced a b* of 32.40 but on a patch with a tread running through it the b* was 35.54.  Wow!

However, that paper, a Baryta paper, is 300gsm and 16mils thick (.41 mm) which is just under the max thickness of .45 mm the XL can handle. It is likely a smaller effect with thinner paper and is not a significant effect with glossy.

The good news is that the treads are spaced such that it only affects one of every 7 columns so one possible solution is to create a specific patch set that fills the columns were the tread wheels track with blank patches. This would decrease the max number of useable patches by 1/7th but would be pretty solid. This is pretty easy to do with a program script to manipulate CGAT files.

A second approach would be to shift the paper to either the left or right such that the wheel marks were centered on the transition where one patch goes to the next. This should work quite well if the paper alignment is sufficiently repeatable. It has to pretty much ignore readings within 1.3mm or so of that transition to work.  I'm currently testing that.

Here's a cell snap  showing the wheel marks on Finestra BaryTa Fine Art 16mil paper. This was taken in sunlight arranged so that the reflectance showed the wheel marks. There are extra tracks a cm or so apart in the image as I read the paper offset about 1 cm to see the effect of additional tracks.

As an aside, the effect is most noticeable on more saturated colors and not the neutral grays. The dE 2000 differences are much smaller. < 50%. So I doubt this would be visible in photo prints. Overall, it's still better, on average, than manual scanning with an I1 Pro2 and far better with glossy.
Title: Re: I1Isis 2 XL, the good, bad, and ugly
Post by: danstart17 on August 20, 2017, 04:50:38 am
Wow, that's pretty bad.

I'm just getting started with colour management and got the i1pro2, are you using i1profiler to read the charts or a 3rd party?
Title: Re: I1Isis 2 XL, the good, bad, and ugly
Post by: Doug Gray on August 20, 2017, 11:37:16 am
Wow, that's pretty bad.

Not really that bad. The light was set up to show the tracks. It's actually not that obvious on the prints.

I'm just getting started with colour management and got the i1pro2, are you using i1profiler to read the charts or a 3rd party?

The Isis 2 XL comes with I1Profiler which can be used to create and print charts.

It does not include a license for I1Profiler profile making s/w, but if you have an I1 Pro and I1Profiler already it will use that license. It can also read targets and save scans in CGAT format for use with other profiling software such as Argyll.
Title: Re: I1Isis 2 XL, the good, bad, and ugly
Post by: Doug Gray on August 20, 2017, 04:40:07 pm
I made a target of 58, all light yellow RGB (255,255,170), patches. Printed it and scanned with the Isis 2. This makes it easy to both see the wheel tracks and plot the "yellowness" change from the track impressions. There are two rows of 29 patches, all the same light yellow.

Attached is a snapshot showing specular surface reflection as well as a Matlab plot of the b* component of the measured Lab values of each row of patches. The graph has been stretched so that the excursions of b* align with the image. The wheel mark excursions are from about .5 to about 1.5 dE over the average baseline. When carefully measured in tiny steps with an I1 Pro, the peaks go up to over b*=35. Clearly the Isis is averaging over some portion of the 6mm patch size and not just reading it at the peak so expanding the patch width should provide improvement. But how much remains to be seen.

Also, the secondary wheels are every 12 mm and leave visible tracks too but aren't showing up on the measurements.


Additional Info:
I ran the test with a patch width of 6.0 and 7.0mm on glossy media and let dry only 30 minutes. There is a very slight evidence of change from the wheels with the glossy. And, it turns out the 7.0 width just happens to align fairly closely with the patch transition edges.

Much better results.

First and second scan comparisons:

6.0 mm:  Average=.04 dE, Worst case=.37dE
7.0 mm:  Average=.03 dE, Worst case=.07dE

Please note these are dE1976 values. The dE2000 values are only 1/3rd as much which is effectively negligible. These errors are now miniscule in comparison to comparing the two print charts which vary an average of .25 dE1976 or .10 dE2000.

Looking at the error locations spatially, the 6.0 mm set aligned with the wheel marks. The 7.0 mm patch set max errors were randomly distributed. However, they were quite small and quite possibly just due to intrinsic variation in Isis readings.

Wheel mark issues aside, I'm very impressed with the repeatability of the instrument and I now have a process for making excellent profiles on the thicker semigloss/luster media as well.
Title: Re: I1Isis 2 XL, the good, bad, and ugly
Post by: Doug Gray on August 21, 2017, 08:04:15 pm
I made some tests to determine the region in the patches that are actually measured. The patches are not measured evenly throughout the length and width but are center weighted with low sensitivity to colors that are near the edges. This is, of course, desirable.

To determine what portion and location of the patches were measured I made a series of horizontal and vertical thin white lines spaced such that, when read, each patch would pick up that potion of luminance contributed by the white line. The lines were evenly spaced from the left edge to the right edge over 30 horizontal black patches. Thus each line is .2 mm further right on each successive column for the 6 mm wide patches and .236 mm for the 7 mm wide patches. The ideal position was calculated and the lines were placed at the nearest pixel column on a 600 DPI grid, the native DPI for the Canon 9500 II.

The prints for the default 6 mm wide and an increased 7 mm wide patches were scanned twice with the ISIS 2. The "Y" value (linear luminance) was read and integrated and plotted to show the accumulated luminance contributed as a function of distance from the patch center.

Each chart shows the cumulative luminance vertically as well as horizontally. The solid and dashed blue lines represent the vertical, 6 mm cumulative luminance response v position. The green line is the horizontal for the 6 mm wide patches and the yellow line is the 7 mm wide patch responses.

Summary:
95% of the 6 mm patch color is read within about a 2.8 mm width reasonably centered on the patches. 95% of the 7 mm patch color is read within about a 3.2 mm width.

Interestingly, the vertical response is elongated. I suspect this is due to the horizontal scan microstepping, which appears to be around (.5 mm) but it may also be due to differences in the optics. Curiously, the vertical measurement distance comes somewhat closer to adjacent patch edges and they vary in centering somewhat more than the horizontal exhibits. This suggests there may be some reason to increase the vertical spacing beyond the minimums.

I hope this isn't too geeky but I wanted to get a better sense of how to determine whether and when the wheel tracks can be ignored.

Here's how to use the graphs to determine the effect of the wheel imprints on thick, lustre type media.

look at the graph, either 7mm or 6mm (default) where the tracks occur. They are about 1 mm wide. Use the ratio, relative to 1, of the start and end of the wheel imprints. If they are 1mm offset from the center than only about 25% of the light is used there so a dE of 3 would be reduced to a dE of about .8 .  If the offset from the center was 1.5 mm then the impact would be much less, about 10%. So the dE error would be closer to .3.  OTOH, if the wheel tracks were dead center, the impact would be about 60% so the dE error would come in around 1.8.
Title: Re: I1Isis 2 XL, the good, bad, and ugly
Post by: Ethan Hansen on August 25, 2017, 04:07:17 pm
Doug - Your observations about patch size vs. dE variations are in line withwhat we found (http://forum.luminous-landscape.com/index.php?topic=119277.40). In the earlier post I neglected to mention our experience with thicker papers. Delicate surfaces definitely create problems with the iSis and thick, stiff paper makes matters worse. We still use Spectroscans for such stocks.
Title: Re: I1Isis 2 XL, the good, bad, and ugly
Post by: Doug Gray on August 25, 2017, 08:44:28 pm
Doug - Your observations about patch size vs. dE variations are in line withwhat we found (http://forum.luminous-landscape.com/index.php?topic=119277.40). In the earlier post I neglected to mention our experience with thicker papers. Delicate surfaces definitely create problems with the iSis and thick, stiff paper makes matters worse. We still use Spectroscans for such stocks.

Ethan,

I appreciate your sharing the test results you've run. So far I've not had any registration or recognition problems with the ISIS 2 XL but I'm only a week and a half into it. OTOH, I do not use ACPU as the identity transform technique works well and I do need precision in the patch spacing to avoid the track problem. Since XRite's I1Profiler refuses to print from the app I have no choice but to save the tiff file. It's become my preferred approach anyway since I can text label relevant info into the target image in an unused region.

So far it's the leading "spacers" that shift the color readings. If they happen to leave tracks near the center there can be a large impact. It's a negligible effect with glossy's and so far with even thick matte, but the thicker lustre/semigloss Baryta I have it's a real problem. There are also "wheels" further back that grab the paper and these microstep. While they can leave a slightly visible specular difference I haven't seen any significant color reading shift from them on any media I've tested so far.  I really like the Baryta for it's great dynamic range, low OBs, and excellent brightness (L=98). I've found a simple approach that eliminates the problem. Since the spacing is 42mm, setting the patch width at either 6mm or 7mm (both divide evenly into 42mm) then simply printing with a carefully selected horizontal offset causes all the lines to be very near the transition region. Even if the paper alignment isn't perfect and the tracks are up to 1mm offset from the patch transition region the effect is reduced to well under .1 dE.  Verifying my empirical observations was the purpose of the sliding, .5mm white window test described earlier.

As an aside, my gold standard profile test is to print Lab patches then measure them and run statistics on the measured v requested values. This checks the BtoA1 tables whereas just checking the reported Lab values from a profile generating patch scan reports the results of the AtoB1 tables. For various reasons, and especially near gamut boundaries, the AtoB1 tables are more accurate than the BtoA1 tables and actually printing a photo uses the BtoA1 tables. Hence the need to print known colors, measure them, and run statistics on them.

Doing this "gold standard" profile test was easy with patchtool and the I1Pro2 but not with the ISIS which imposes various alignment items in it's target image. So I've worked on a way around that.

I've made a Matlab script that reads the Isis tif file and replaces the patch RGB values with a set from a CGATs file. I convert the reference Lab values to device space using the profile under test in Abs. Col. then call the script to "fix" a generic tif Isis file with those device RGB values. It also treats the patches as collections of 4 grouped pixels and interpolates fractional values. For instance a R value of 150.3 will put 150 in three pixels and 151 in the fourth. This effects a 10 bit printer resolution. Then the image is printed w/o color management (it's already done and in device space) and Isis scans the patches and records the Lab values. Then the measured ones are compared with the requested Lab values. So now that process is not only doable like the I1Pro 2, but much faster and more consistent than the I1Pro 2 process.


Title: Re: I1Isis 2 XL, the good, bad, and ugly
Post by: Wayne Fox on August 27, 2017, 06:40:23 pm
Is this being caused by a mechanical change in the iSis 2 vs. the original model?  I’ve never seen issues with this effect in my iSis, but then again I’ve really never looked at things this closely.
Title: Re: I1Isis 2 XL, the good, bad, and ugly
Post by: Doug Gray on August 27, 2017, 10:40:48 pm
Is this being caused by a mechanical change in the iSis 2 vs. the original model?  I’ve never seen issues with this effect in my iSis, but then again I’ve really never looked at things this closely.

It's a fairly small effect. Overall the Isis 2 is extremely consistent. Successive passes normally produce average dEs of .05 or less. That high repeatability, which is much better than their specification (.10), is what made this problem evident. Even then it really only affected fairly thick, Baryta pearl type surface media where the 4.2 cm spacers rubbed against the surface.

I see a variation of up to 2.0 dE with an average of .4 just from running patches randomized to different locations on the same paper and that is unrelated to the track issue. The Isis reads a fairly small area and most of the color data is from a 2mm wide portion in the patch center. This can likely be explained by small differences in the head to paper spacing and possibly tiny differences in the particular nozzles in use over that small region.

As an aside, the larger dE excursions from this, when measured with Delta E 2000, instead of Delta E 1976 are 1/2 to 1/3 the amplitude so, perceptually it's pretty much a non-issue. For instance, one of the color patches that shows the greatest tendency is a highly saturated yellow with a b* close to 100 and this is where the dE 2000 is much less sensitive than the dE 1976.

But, because of the high repeatability of the Isis, and a desire to refine some specialized charts to measure things like imager flatness and registration to the image plane I'm pursuing it just to understand and eliminate or mitigate the effect of some of the variables.
Title: Re: I1Isis 2 XL, the good, bad, and ugly
Post by: Ethan Hansen on August 28, 2017, 01:11:04 pm
Is this being caused by a mechanical change in the iSis 2 vs. the original model?  I’ve never seen issues with this effect in my iSis, but then again I’ve really never looked at things this closely.

It certainly is worse with the iSis 2. The original iSis sometimes leaves crush marks on thick, delicate papers, with matte being affected most. The single iSis v2 we have feeds more consistently but this comes at the cost of more mechanical squishing on the paper surface.
Title: Re: I1Isis 2 XL, the good, bad, and ugly
Post by: unesco on August 28, 2017, 04:13:44 pm
I see a variation of up to 2.0 dE with an average of .4 just from running patches randomized to different locations on the same paper and that is unrelated to the track issue. The Isis reads a fairly small area and most of the color data is from a 2mm wide portion in the patch center. This can likely be explained by small differences in the head to paper spacing and possibly tiny differences in the particular nozzles in use over that small region.
I would say that this is rather inconsistency of printer having slightly different results in different location of the page than the measurement variation. Have you tried to measure full page of the identical patches?
Title: Re: I1Isis 2 XL, the good, bad, and ugly
Post by: Doug Gray on August 28, 2017, 05:21:21 pm
I would say that this is rather inconsistency of printer having slightly different results in different location of the page than the measurement variation. Have you tried to measure full page of the identical patches?

I agree. So far my tests indicate very little variation in measurements made in identical places. It's quite remarkable how consistent the readings are. But the tests also show that only a fairly tiny part of each patch is measured, in effect a portion between 2mm and 3mm square. So that would exacerbate any location or paper variation.

Good idea on the full page test. I've just made one with all the patches using the RGB value that shows the largest variation. Interesting to see if there is some location pattern.
Title: Re: I1Isis 2 XL, the good, bad, and ugly
Post by: Doug Gray on August 28, 2017, 06:32:20 pm
Here's the results. The statistics are consistent with my randomized set of 64 patches repeated and re-randomized 8 times.

I changed the ink on my 9800 to PK so I could do decent comparison tests with glossy on the 9500. The 9800 exhibits about 30% less variation than the 9500.

The color that exhibited the most variation on the 9800, is a salmon color, device space RGB (255,170,170), Average dE .5, max 1.8.

I used patchtool to compare two scans. One was fed in 2 cm offset from the other to identify effects from the ISIS v effects from the inked paper itself. There are no observable effects from feeding the paper in with an offset.

Indeed, there is an interesting vertical patterning which is likely something the printer is doing, possibly from the way vacuum is applied creating slight variation in head to paper spacing.

The two scans also differed by an average dE of .10 with a maximum of the 841 patches of .31.

Assuming more or less a normal distribution the measurement variance was 1/25th of the variance due to media/ink variation. I suspect most of that is due to slight, sub mm,  shifts in where the readings are taken on each patch.

So the Isis is proving to be extremely consistent. However, because of the relatively small portion of each patch that the sample is taken from it seems that the very best profiles are best made from the use of multiple, randomizing patch sets. I1Profiler offers two variations. unrandomized and randomized but it's fixed. Averaging scans from the two would be much better than just re-reading a single set.
Title: Re: I1Isis 2 XL, the good, bad, and ugly
Post by: GWGill on August 28, 2017, 08:46:29 pm
It's a fairly small effect. Overall the Isis 2 is extremely consistent. Successive passes normally produce average dEs of .05 or less.
I'd attribute higher stability to the nature of the light source more than (or as well as) mechanical repeatability. With ideal mechanical repeatability (i.e. leaving the instrument in the same spot), I found that the ColorMunki is far better than the i1Pro due to the better stability of the (temperature compensated) LED light source over the incandescent lamp.
Title: Re: I1Isis 2 XL, the good, bad, and ugly
Post by: Doug Gray on August 28, 2017, 09:06:28 pm
I'd attribute higher stability to the nature of the light source more than (or as well as) mechanical repeatability. With ideal mechanical repeatability (i.e. leaving the instrument in the same spot), I found that the ColorMunki is far better than the i1Pro due to the better stability of the (temperature compensated) LED light source over the incandescent lamp.

Absolutely agree. I too have seen the I1Pro shift it's white point depending on how often one takes samples or has held down the strip reading button.

OTOH, I'm very pleased with how well it correlates with the Isis. I wouldn't have been surprised if the average difference in Lab readings was over dE 2. Especially given that I haven't had the I1Pro 2 calibrated since I bought it and the earlier I1s I have, a uv Cut and regular M0, are significantly different than the I1 Pro 2 with an ave dE on a Colorchecker of > 1.7 on the uV cut and 2.2 on the older M0 model. dE2000 on the ones furthest off is much closer, fortunately.

Also, the Isis does use a true, uV cut and, optionally, simulates M1 and M0 with a second pass using, presumably, a uV LED. The readings with high OB paper and M2 and M1 mode for both are remarkably consistent. I do need to test for that anomaly I saw with the I1Pro2 where there would be unexplained if infrequent, overly large responses to the uV LED messing up both the M1 and M2 values.
Title: Re: I1Isis 2 XL, the good, bad, and ugly
Post by: Doug Gray on August 28, 2017, 09:40:23 pm
I would say that this is rather inconsistency of printer having slightly different results in different location of the page than the measurement variation. Have you tried to measure full page of the identical patches?

unesco,

I would like to thank you once again. It turns out your suggestion is a highly productive one. For one thing it's voided one hypothesis I had, which was that Isis's small area of patch reading contributed significantly to the dE variation I saw. It doesn't. Very good news but now I have to re-align my thinking towards almost all the observed error being due to the printer's paper handling and not variation in the printer head or the Isis instrument.

I'm now looking at the Canon 9500 II and will be summarizing the relevant data soon. In particular it suggests a mechanism that could work much better than just averaging, redundant, randomly positioned patches.
Title: Re: I1Isis 2 XL, the good, bad, and ugly
Post by: Stephen Ray on August 29, 2017, 01:53:15 am
Doug,

Can you imagine how consistently the ISIS would handle canvas media?
Title: Re: I1Isis 2 XL, the good, bad, and ugly
Post by: unesco on August 29, 2017, 01:27:44 pm
unesco,

I would like to thank you once again. It turns out your suggestion is a highly productive one. For one thing it's voided one hypothesis I had, which was that Isis's small area of patch reading contributed significantly to the dE variation I saw. It doesn't. Very good news but now I have to re-align my thinking towards almost all the observed error being due to the printer's paper handling and not variation in the printer head or the Isis instrument.

I'm now looking at the Canon 9500 II and will be summarizing the relevant data soon. In particular it suggests a mechanism that could work much better than just averaging, redundant, randomly positioned patches.

Doug, thank you, I am really happy I could help - as an ex-researcher I quite often concentrate on details which looks tiny, but can have significant impact... in this case my tries with Epson 3880 and QTR curve design with my Munki made me think that repeatibility of patch print is the major cause for my curves not being smooth. Even 3 prints of 51 grayscale patches with different distribution on page and then averaging gave results close to spline smoothing of one print. And I talk about visual examination of bull-eye gradient smoothness, not just measurement. Anyway, let's keep researching this complicated printing world ;-)
Title: Re: I1Isis 2 XL, the good, bad, and ugly
Post by: Ethan Hansen on August 29, 2017, 02:44:23 pm
Can you imagine how consistently the ISIS would handle canvas media?

That depends on the media. Thin, fairly stiff canvas with minimal texture and a low-sheen finish works well. Flexible canvas gets mangled in the feed mechanism. Rough textured or glossy surfaces do not read accurately. You need a spherical geometry rather than 45/0 for best results on fabric. Failing that, a polarizing filter helps, Neither is possible with the iSis.
Title: Re: I1Isis 2 XL, the good, bad, and ugly
Post by: Ethan Hansen on August 29, 2017, 08:12:56 pm
I'm now looking at the Canon 9500 II and will be summarizing the relevant data soon. In particular it suggests a mechanism that could work much better than just averaging, redundant, randomly positioned patches.

Doug - Welcome to the color measurement rabbit hole!

You will find color variations from different sources. Photo inkjets will, as you've seen, exhibit variation depending on how the printer/paper combination works. Your hypothesis that the paper hold-down is playing a part sounds reasonable. We see this frequently on wide format Epsons, where there are distinct areas of the page that print slightly differently than others. Laser-exposed silver halide printers (Fuji Frontier, Noritsu QSS, etc.) can show a difference usually from one side of the page to another.

Our targets have a number of repeated patches to detect this variability. We also track the behavior of similar colors across the page (e.g. target values with consistent, small offsets) to see if they are consistent. We preprocess each file before feeding to the profiler. The key is to determine whether variations are (a) significant and if so (b) is the printer setup within normal parameters.

Looking at variation, it is worth first checking the level of variability from the iSis itself. A straightforward test is to measure a blank page. Simply print the positioning bar and side markers and measure. Using an OEM paper (Canon or Epson) with a lustre surface will give uniformity well within what the iSis can resolve. We also printed the iSis fiducial marks on pvc sheets of various colors to see how the iSis behaves on colors other than white. Gather short-term repeatability data as well as measurements a day or two apart. You'll derive a baseline for instrument performance.

Start customizing the canned targets and preprocessing. We toyed with the idea of creating targets with identical patches per page but different layouts. You'll need to average the data first (stay in the spectral domain!) before feeding the summarized data file to i1Profiler. We ended up not going this route for a couple of reasons. First, we typically did not read variations sufficiently far from the iSis noise floor to be worth the doubled measurement times. Also, we have the luxury for most of the profiles we build of having hundreds of the same model of printer to build a reference form. At that point, a significant portion of the profile generation hinges on eigenvectors rather than only the raw data.
Title: Re: I1Isis 2 XL, the good, bad, and ugly
Post by: Doug Gray on August 30, 2017, 01:49:39 am
You will find color variations from different sources. Photo inkjets will, as you've seen, exhibit variation depending on how the printer/paper combination works. Your hypothesis that the paper hold-down is playing a part sounds reasonable. We see this frequently on wide format Epsons, where there are distinct areas of the page that print slightly differently than others. Laser-exposed silver halide printers (Fuji Frontier, Noritsu QSS, etc.) can show a difference usually from one side of the page to another.
Yes, I see large, repeatable variations, mostly horizontal, on my 9800. These are not present on my 9500 II desktop. OTOH, the 9500 exhibits a sort of "warm up" phenomina where the dE changes almost 3 dE when printing a fairly high saturation red. This occurred after the 9500 had been unused for 24 hours. Checking it against after a 2 hour pause showed about a 1 dE "warm up" period. Nothing odd is showing in the image and nozzle checks have always been clean unless it sits more than a few weeks. The 9800's horizontal patterning is quite consistent, and changes only slightly when altering the vacuum or single/dual direction modes. It did improve about 20% or so using the narrow platen selection. The 9500's ink flow warmup issue is visible in the Patchtool image I attached. As you can see, the first rows have pretty significant dE variation.
Quote

Our targets have a number of repeated patches to detect this variability. We also track the behavior of similar colors across the page (e.g. target values with consistent, small offsets) to see if they are consistent. We preprocess each file before feeding to the profiler. The key is to determine whether variations are (a) significant and if so (b) is the printer setup within normal parameters.
My initial test was with a 512 patch set that included 64 distributed RGB patches randomly placed and repeated 8 times. The fact there was significant variation in patches of the same color compared to reading the same sheet and comparing readings of the same patches kicked me off down this rabbit hole.

Quote
Looking at variation, it is worth first checking the level of variability from the iSis itself. A straightforward test is to measure a blank page. Simply print the positioning bar and side markers and measure. Using an OEM paper (Canon or Epson) with a lustre surface will give uniformity well within what the iSis can resolve. We also printed the iSis fiducial marks on pvc sheets of various colors to see how the iSis behaves on colors other than white. Gather short-term repeatability data as well as measurements a day or two apart. You'll derive a baseline for instrument performance.
Funny you should mention that. I did just that earlier today out of curiosity as to whether there was some sort of trend from the start of reading patches as the sheet scan progresses. The I1 Pro shows a significant drift (.2 to .3 L*) and drift from the illuminant has the largest impact at high luminance. Turned out to be less than .1 from start to end. Frankly, I was rather pleasantly surprised.
Quote

Start customizing the canned targets and preprocessing. We toyed with the idea of creating targets with identical patches per page but different layouts. You'll need to average the data first (stay in the spectral domain!) before feeding the summarized data file to i1Profiler. We ended up not going this route for a couple of reasons. First, we typically did not read variations sufficiently far from the iSis noise floor to be worth the doubled measurement times. Also, we have the luxury for most of the profiles we build of having hundreds of the same model of printer to build a reference form. At that point, a significant portion of the profile generation hinges on eigenvectors rather than only the raw data.
Averaging the data is pretty straightforward. I use Matlab as I find it's easier but I understand I1Profiler will average files if you select both and drag and drop. Haven't tried it though.

Curious as to how you use eigenvectors in profile generation. I've kicked around doing some principal factor analysis to see if the spectral functions could be reduced (mostly - SVD here we go) to a linear group in 3 space. and then one should be able to find a matrix to convert linearized scanner RGB to XYZ. I doubt it would work with the 9500 since that has a green and red ink as well as the CYMs. OTOH, the 9800 just uses CYM and lightened version of C and M so it might have a better shot
[/quote]
Title: Re: I1Isis 2 XL, the good, bad, and ugly
Post by: Doug Gray on August 30, 2017, 11:30:47 am
Looking at variation, it is worth first checking the level of variability from the iSis itself. A straightforward test is to measure a blank page. Simply print the positioning bar and side markers and measure. Using an OEM paper (Canon or Epson) with a lustre surface will give uniformity well within what the iSis can resolve.

I wanted to test both the consistency of the Isis and get some idea of the production consistency of the (cheap) Costco glossy media I'm using to refine processes before expanding into OEM or specialty media. So I pulled two sheets from the new Costco package and rotated 1 relative to the other 180 degrees which flips top/bottom and right/left. These were then printed with the registration marks and form a 841 patch set 29 by 29 with the default 6 mm patch sizes.

Two Isis read scans were done with each paper. The variation, small as it is between the two papers, is far smaller comparing two scans with the same paper.  Yes!

This says volumes about the excellent Isis repeatability and good positioning to the registration marks. Very low noise baseline.

About .11 ave dE between the two papers and .02 dE on the same paper. See attachments.

The 4 scans were done back to back with the fast, 1 pass (M2) mode over a 10 minute interval.
Title: Re: I1Isis 2 XL, the good, bad, and ugly
Post by: Doug Gray on August 30, 2017, 01:16:43 pm
>> I see large, repeatable variations, mostly horizontal, on my 9800. These are not present on my 9500 II desktop. OTOH, the 9500 exhibits a sort of "warm up" phenomina where the dE changes almost 3 dE when printing a fairly high saturation red. This occurred after the 9500 had been unused for 24 hours.

Test protocol: For each printer, the Canon 9500 II and the Epson 9800, a page of 512 patches containing 8, randomly located, groups of 64 max spaced RGB triplets was printed. The pages were scanned and each set of 8 patch samples of the same colors were read for both printers. The RGB triplet with the maximum average deltaE on the 9500 and another on the 9800 was used to create a 841 patch set in a square, 29x29, of 6 mm identical patches. These were printed and scanned with the Isis. In both cases the major contributor to Delta E was b* variation so that was the variable studied.

The scans, a matrix of 29 rows by 29 columns was averaged horizontally to determine vertical variation and averaged vertically to determine horizontal variation. Each of these produced a length 29 vector. Attached charts labeled "Horizontal Variation" is a plot of the 29 points, (left to right paper in profile) from averaging each column. "Vertical Variation" charts were row averaged producing a vertical column vector. The 29 points is the average  b* values from top to bottom.

A note about the error bars. These are the standard deviation of the set of 29 samples from which that point on the chart was determined.  Thus, they represent the error bars of samples orthogonal to the plot line. When the horizontal variation is greater than the vertical variation the error bars shown on the vertical plot will also be higher and vice versa. Please note that these error bars are per sample that makes up the point. The error bars of the averaged point itself would be a bit over 80% narrower.


Note the strong shift in *b in the top portion of the Canon 9500 chart representing the average  change vertically. It's useful to compare that to the Patchtool 9500 chart dE chart posted earlier.

Also, the Epson 9800, in a reverse of the Canon 9500, shows more variation horizontally than vertically. This is consistent with what Ethan noted about wide format Epson printers.
Title: Re: I1Isis 2 XL, the good, bad, and ugly
Post by: Doug Gray on August 30, 2017, 05:35:00 pm
More good stuff from the I1 Isis 2.

In order to rule out whether the Isis was reading some portion of adjacent patches I made a special target file with 841 (29x29) 6 mm black patches. Then I added white stripes .25 mm inset from all patch edges. This changes the actual black patch sizes to 5.5 mm square with white filling the gaps between patches. This maximizes light leakage and makes it easy to detect since the reflectance of the black patches is very low. At DMax of L*=4 (my 9800's BP) leakage of .01% would produce an increase of .9 L*. Running this chart, even with a slight angle tilt, showed no signs of adjacent patch light leakage.

I've included a tif image of the target which anyone can use with an Isis. Just scan it into a target of 841 patches using the defaults and a standard letter size profile page. Save the results as a CGATS files and read it with a text editor or graph the L* values in Excel.

I am way more comfortable now that any variations I see are actually on the media and not leakage from adjacent patches.
Title: Re: I1Isis 2 XL, the good, bad, and ugly
Post by: Doug Gray on September 02, 2017, 05:14:01 pm
Building on some of the observations and comments, I created two, 9800 glossy profiles using the Isis default target of 957 patches with the following differences:

Target 1: The patches were duplicated then randomized producing a two page target but with the patches randomly scattered.

Target 2: The patches from one default target were duplicated then the columns were relocated. Each column was placed at the max and min horizontal deviation (sorted col extremes). As these columns were filled the two extremes were removed and the process repeated. This also yields 2 target pages where each of the default patches are duplicated though at positions where the horizontal inking variation tends to cancel.

Targets were printed, scanned, and profiles made. Then my 389 patch set of in-gamut LAB colors was printed using the two profiles. This differs from things such as the dE report from ColorThink which reports on the reverse lookup of RGB values and compared them to the target's scanned LAB values. This uses the AtoB1 tables in ICC profiles. Printing photos (colorimetric) does not use AtoB1 tables and only uses the BtoA1 tables.

Here are the statistics on the two, 9800 profiles which use exactly the same patches and page size but with the patches re-ordered to cancel the effect of the predictable portion of horizontal positioning inking sensitivity. Results in Delta E 2000.

CG BASE: dE's-> ave:0.46  median:0.42  <90%:0.75    chart max: 1.36
CG OPT:   dE's-> ave:0.39  median:0.35  <90%:0.67    chart max: 1.16

I consider both of these results to be well within what I would expect for a high quality profile given the relatively small number of unique (957) patches.

The BASE profile benefits from averaging targets with randomly scattered patches while the OPT benefits from strategically placing the patches such that most of the observed horizontal inking variation in the wide format 9800 cancels.


Side Notes:
The inking variation occurs most frequently on channels with a lot of added yellow. Magenta and cyan dominant colors show less pattering. Speculation is that there is only one "Y" ink whereas there are light versions of cyan and magenta.

Also, this technique of compensating for horizontal variation only improves the profile accuracy to the extent the profile is generated from affected patches . Variation in the printed image itself will exhibit this variation but no longer has the risk of compounding variation in creating the profile and variation in printing using the profile.
Title: Re: I1Isis 2 XL, the good, bad, and ugly
Post by: Doug Gray on October 02, 2017, 05:26:32 pm
I am way more comfortable now that any variations I see are actually on the media and not leakage from adjacent patches.

No longer!

I scanned a single sheet of 2920, 6mm x 6mm patches on a glossy 13x19"  3 patches were way off. 8, 10 and 11 dE compared to 3 scans I had made the day before respectively. All others were reasonably close. Ave dE was .12.

Those three patches were all on the right most column and each spaced with a good row in between. Interestingly, those patches were printed with RGB 78,0,0  0,78,0,  0,0,78 and are quite dark on the Espon 9800 with L* of 7 to 10. The patches to the left were all similar in L*. Doing the math, it appears that light was leaking from the white margins. It would take a tiny misregistration of about to produce that sort of shift. As large as it is, it's only the effect of about .5% light contamination from "white." Misregistration errors are far more damaging to darker patches than lighter ones. The same error at L*=50 would only produce about 1 dE difference.

The difference between this and the other scans, which match nicely, is that I enabled dual scan mode. This measures 2 rows with a white light, then, backs up a row and measures the same two rows with a uV LED. Also, each row is measured in alternating directions, right to left, step a row, left to right, step ... repeat.

There are other factors that should be considered. The I1Isis does not use the color changes from one patch to another to determine boundaries. They are purely a function of dividing up the total row length into precise, segments and building a window that should exclude leakage from adjacent patches. This makes the patch spacing linearity extremely critical. This is magnified when the patches range over a long distance as is the case of the 13x19, single page, 2920 patch print.

Recommendations:

When profiling larger than US letter size and you have the choice, select Profile, not Landscape, when using letter size. The longer the rows, the more error potential when interpolating where those patch boundaries are.

Use 6.5 or 7.0 mm spacing, in steps of .5 mm. This is because other fractions produce patch sizes that vary. The patch pixel width on the print is .25 mm and so .2 or .3 mm slider settings will result in different patch sizes across a row. Dumb.

Let the print dry. There could be subtle dimension changes due to variations in inking and drying. While the colors stabilize in a few minutes, the dimensions may not. I've seen some limited evidence of this.

I need to look into this more closely but I believe these recommendations will avoid most of the registration issues.

Edit for minor dim corrections.

Edit: Explanation found!
I tried to repeat scans on the target that showed the dE > 10 deviations and was not successful until I noticed that the position of the Isis XL together with the use of very a large target was the problem.

I had placed the Isis XL on a table about 28" in width, closer to exit edge. I had also select dual scan mode - I normally scan with just M2. Because the paper is fairly long at 19 inches, a portion of it would bend when extending past the table edge. This increases the drag and makes the drag asymmetrical. When in the dual scan mode the scanning motion is to read with a white lamp left to right then step forward and read right to left.  Then the device reverses paper direction and steps back to the previous scanned row where the process is repeated with the uV LED. This exacerbates the effect of asymmetrical drag gradually producing an offset that resulted in enough leaked light from adjacent patches to produce the larger error. The asymmetrical drag is made worse because the 13x19" 250 gsm paper's weight and inertia is quite an increase over my normal targets that are US Letter size and lie flat on both sides of the Isis.

Subsequently, I have changed the orientation so that the paper always lies flat and doesn't extend past the  tables with these long papers. Since then there has been no recurrence.
Title: Re: I1Isis 2 XL, the good, bad, and ugly
Post by: Ethan Hansen on October 04, 2017, 05:04:10 pm
Use 6.5 or 7.0 mm spacing, in steps of .5 mm. This is because other fractions produce patch sizes that vary. The patch pixel width on the print is .25 mm and so .2 or .3 mm slider settings will result in different patch sizes across a row. Dumb.

Dumb indeed. We build our iSis targets without i1Profiler. The resolution of 40 pixels/cm (101.6 px/in) produces a patch resolution of +/- 0.25mm. And i1Profiler does not accept values less than 0.1mm in normal configuration.

What we do instead is make targets at (usually) 300 ppi. This gives +/- 0.085 mm patch resolution. A typical printer will not introduce much additional sizing error. Note that we picked 300 ppi as this is the resolution used by the drivers for our main customers. Other printers vary, and you may well choose a target dimension of 360 ppi for an Epson inkjet.

Entering the resulting patch sizing into i1Profiler is easiest if you switch the units on the test chart page to inches. Magically sizing resolution switches to 0.001 inches (or 0.0254 mm), making it an order of magnitude better than the baseline behavior.

The total length of each row is sized to the number of patches * the patch size specified in i1Profiler. This means that individual patches may vary by 1 pixel; at 300ppi this is not enough to cause measurement errors. For example if you want a patch size of 6.69mm, this translates to 79 pixels at 300 ppi. Let's assume we want rows of 30 patches (7.9 inches). The total row length should be 30 [patch] * 6.69 [mm/patch] * 1/25.4 [in/mm] * 300 [px/in] = 2370 pixels (rounded down). One patch in the row will have only 78 pixels while the rest have 79.

The end result of these shenanigans is fewer read errors for a given patch size.
Title: Re: I1Isis 2 XL, the good, bad, and ugly
Post by: Doug Gray on October 04, 2017, 05:53:43 pm
Dumb indeed. We build our iSis targets without i1Profiler. The resolution of 40 pixels/cm (101.6 px/in) produces a patch resolution of +/- 0.25mm. And i1Profiler does not accept values less than 0.1mm in normal configuration.

What we do instead is make targets at (usually) 300 ppi. This gives +/- 0.085 mm patch resolution. A typical printer will not introduce much additional sizing error. Note that we picked 300 ppi as this is the resolution used by the drivers for our main customers. Other printers vary, and you may well choose a target dimension of 360 ppi for an Epson inkjet.

Entering the resulting patch sizing into i1Profiler is easiest if you switch the units on the test chart page to inches. Magically sizing resolution switches to 0.001 inches (or 0.0254 mm), making it an order of magnitude better than the baseline behavior.

The total length of each row is sized to the number of patches * the patch size specified in i1Profiler. This means that individual patches may vary by 1 pixel; at 300ppi this is not enough to cause measurement errors. For example if you want a patch size of 6.69mm, this translates to 79 pixels at 300 ppi. Let's assume we want rows of 30 patches (7.9 inches). The total row length should be 30 [patch] * 6.69 [mm/patch] * 1/25.4 [in/mm] * 300 [px/in] = 2370 pixels (rounded down). One patch in the row will have only 78 pixels while the rest have 79.

The end result of these shenanigans is fewer read errors for a given patch size.

Interesting you mentioned this. I'm in the process of creating Isis chart image files using matlab with arbitrary DPI (720 for my Epson, 600 for the Canon). I've finished the code for generating arbitraryf patch sizes from 6 to 7 mm width and height. I'm now adding another feature, creating a sliding window across the patch, vertically or horizontally, which will allow me to characterize the specific variability and effective scanned patch sizes. RGB values will be set to one value on the left and another on the right. The dividing line will be a ratio fro 0 to 1. Same for the vertical.

I am most curious what it will show. I expect it to provide clear results for maximizing comfort level of profile quality together with getting the most patch density.