Shadow Improvement stumper?

[Mick Sang]

I have been reading this forum for many years. At no time did it ever occur to me that I might someday post a question that would be a "stumper." I am certainly stumped and so far at least, this appears to be the case for the participants of this forum, since close to 100 folks have read my first post entitled "Inconsistent results in shadow detail from profiles made via M3 readings" without a single response. But, I would appreciate any thoughts from the very knowledgeable printmakers on this site. So, I thought I'd try once more:

Recently, I purchased a Barbieri LFP S3 to create ICC profiles using M3 (polarizer) based readings on various matte inkjet papers. Readings were collected from targets using Barbieri Gateway and subsequently imported to i1Profiler for the creation of the print profiles. My first attempt was to reprofile Epson Hot Press Bright which I compared against the original profile that I made 2 years back using an i1Isis-1. The resulting M3 profile was as amazing as has been described in other posts in this forum i.e. much lower L* D-max, much larger colour gamut and clearly more shadow separation in the print from the M3 based profile versus a print from the old M0 and M2 profile.
 
Next, I chose to reprofile Hahnemuhle Museum Etching. This time the new M3 profile would be compared against one previously made with an Isis-2 (M0, M1 and M2). In short, while the new M3 profile also showed the same kind of aforementioned results in ColorThink Pro, there was clearly less shadow detail in the print as compared to the print from the M0,1,2 profile. For all prints the perceptual intent was used. Neither perceptual nor Rel Col rendering intents made any improvement to the M3 print. So, now I'm wondering if the improvement that is seen in the M3 print for Hot Press Bright is really due to the difference between the readings from the new Barbieri LFP versus an older i1Isis-1.

I know that at least a few of you have experience with this method of ICC profiling. I recall having read in one past post by Josh-H that he had to laboriously adjust the rotation of the polarizing filter to find the exact sweet spot for it. But, the filters on my LFP have registration pins with corresponding sockets on the head into which the pins seat. So, there is no way that I know of to adjust the rotation of the filter. It should be in that sweet spot. At least one would hope. I'm aware that any improvement in shadow detail is negligible for gloss stocks. But, I was under the impression that all matte stocks would gain some improvement. Certainly the profile shows it. So, my question is does anyone have any idea what is going on here?

Then, a couple of days after the above was posted, I profiled another paper - Epson Legacy Etching using the Barbieri LFP with its polarizing filter. I ran the chart readings through i1Profiler for an M3 based ICC profile. Again, the resulting profile had a much larger gamut and the black point was L2 versus L16 from the previous profile which had been made using an i1Isis-1. BUT, the result in print showed NO improvement at all. It was virtually identical to the print made through the old profile.

So, after three M3 profiles, we have one with obvious improvements to the shadows, one with obviously less shadow separation and detail and another with no apparent benefit at all in print. It's a draw. But, in every case, the profiles appeared to be drastically improved.  I hope someone out there can provide a clue as to why this is happening. What am I missing? What can I do to improve the results in print?

Last week, I made an M3 profile for Hahnemuhle Torchon which contains OBAs as does Hot Press Bright. This time, the profile as viewed in ColorThink showed the same huge improvements as all the others and there was a slight improvement in the M3 print versus the print made from a profile created using i1Isis-2 (M0, 1, 2). I should add that each print pair was made on the same printer (either Epson 7900, SC500 or SP4900). So, now I wonder if these improvements have some relationship to the OBA content. In short, in view of the results that had been noted by others in this forum in the past which were based upon the same process when compared to the results that I have been getting, I am stumped!

[Doug Gray]

I have been reading this forum for many years. At no time did it ever occur to me that I might someday post a question that would be a "stumper." I am certainly stumped and so far at least, this appears to be the case for the participants of this forum, since close to 100 folks have read my first post entitled "Inconsistent results in shadow detail from profiles made via M3 readings" without a single response. But, I would appreciate any thoughts from the very knowledgeable printmakers on this site. So, I thought I'd try once more:

Recently, I purchased a Barbieri LFP S3 to create ICC profiles using M3 (polarizer) based readings on various matte inkjet papers. Readings were collected from targets using Barbieri Gateway and subsequently imported to i1Profiler for the creation of the print profiles. My first attempt was to reprofile Epson Hot Press Bright which I compared against the original profile that I made 2 years back using an i1Isis-1. The resulting M3 profile was as amazing as has been described in other posts in this forum i.e. much lower L* D-max, much larger colour gamut and clearly more shadow separation in the print from the M3 based profile versus a print from the old M0 and M2 profile.

Next, I chose to reprofile Hahnemuhle Museum Etching. This time the new M3 profile would be compared against one previously made with an Isis-2 (M0, M1 and M2). In short, while the new M3 profile also showed the same kind of aforementioned results in ColorThink Pro, there was clearly less shadow detail in the print as compared to the print from the M0,1,2 profile. For all prints the perceptual intent was used. Neither perceptual nor Rel Col rendering intents made any improvement to the M3 print. So, now I'm wondering if the improvement that is seen in the M3 print for Hot Press Bright is really due to the difference between the readings from the new Barbieri LFP versus an older i1Isis-1.

I know that at least a few of you have experience with this method of ICC profiling. I recall having read in one past post by Josh-H that he had to laboriously adjust the rotation of the polarizing filter to find the exact sweet spot for it. But, the filters on my LFP have registration pins with corresponding sockets on the head into which the pins seat. So, there is no way that I know of to adjust the rotation of the filter. It should be in that sweet spot. At least one would hope. I'm aware that any improvement in shadow detail is negligible for gloss stocks. But, I was under the impression that all matte stocks would gain some improvement. Certainly the profile shows it. So, my question is does anyone have any idea what is going on here?

Then, a couple of days after the above was posted, I profiled another paper - Epson Legacy Etching using the Barbieri LFP with its polarizing filter. I ran the chart readings through i1Profiler for an M3 based ICC profile. Again, the resulting profile had a much larger gamut and the black point was L2 versus L16 from the previous profile which had been made using an i1Isis-1. BUT, the result in print showed NO improvement at all. It was virtually identical to the print made through the old profile.

There is one area that M3 profiles excel and that is when the paper is difficult to profile in deep shadows. Also, profiles constructed from M3 use a bit more of the LUTs than M0/1/2 profiles. This can often yield better profile shadow separation. Especially on things with rough textures that produce a high noise level with M0/1/2 spectros.

But there is another issue potentially at play. Any lint or dust particles present when scanning can reduce shadow resolution. And it can be significant. Dark patches are by far the most sensitive to this effect (see a recent post by me and a response from Chromix's support). The variation you see may well be from this problem.

So, after three M3 profiles, we have one with obvious improvements to the shadows, one with obviously less shadow separation and detail and another with no apparent benefit at all in print. It's a draw. But, in every case, the profiles appeared to be drastically improved.  I hope someone out there can provide a clue as to why this is happening. What am I missing? What can I do to improve the results in print?

The larger gamut and improved DMax in an M3 profile is fiction. Pure and simple. The gamut isn't larger and the actual DMax (comparing a printed black patch using M0-2 v M3) is also the same.

More problematically, M3 profiles exclude reflected light off of small, specular surfaces on the paper. This results in a gamut expansion from which the profile is made but, unless you view prints using polarizing filters, you will see a slightly desaturated print because normal illumination will have the additional, specular reflections. This is easy to check. Just print a ColorChecker image using the M3 profile and one made using regular Ms. The M3 print's colors will be slightly desaturated.

Last week, I made an M3 profile for Hahnemuhle Torchon which contains OBAs as does Hot Press Bright. This time, the profile as viewed in ColorThink showed the same huge improvements as all the others and there was a slight improvement in the M3 print versus the print made from a profile created using i1Isis-2 (M0, 1, 2). I should add that each print pair was made on the same printer (either Epson 7900, SC500 or SP4900). So, now I wonder if these improvements have some relationship to the OBA content. In short, in view of the results that had been noted by others in this forum in the past which were based upon the same process when compared to the results that I have been getting, I am stumped!

Shouldn't have anything to do with OBAs but be aware that M3 is also uV cut in addition to polarizing filters. Best to compare with I1iSis 2 M2 profiles. Scanning to include M0/1 as well takes twice as long.

[arobinson7547]

In the past, I remember noticing that Profiles made with a Spectro LFP Series 3 reported a much highrer L* reading then an i1 Pro Rev D. UV cut.

Also during Reading a Color Strip for Proof verification the Barbieri and the i1 disagreed enough the cause the Proof verification to fail. (If Profiled and matched using one device and verified using another.


This is Inter-Instrument agreement. Two different devices, two different manufactures, perhaps different technologies.

I think this is the cause of your prints visually matching but different readings.

I also think that it is good to remember that it's Prints that are Produced/Sold/Admired and not the technology behind producing then; meaning to say, let your eyes be the ultimate judge and not raw data and numbers. That second part is, just a guide and one of the many tools used to aid in getting where you want to be.

(I'm not gonna touch the M3 Stuff)

Remember with the LFP S3  you have no filter/uv filter and Pol filter. It may be helpful to just observe the affects of using those three states, within your Barbieri/Gateway before comparing and considering the differences between instruments/manufactures.

compare your spectral measurements straight from Gateway in Colorthink, using those three modes. And look at your Prints using Profiles built with the three modes, first.

I personally hate the trade off of Pol/UV filter, cause you can't use them both, at he same time. What would the results be? (you gotta wonder)

[Doug Gray]

I personally hate the trade off of Pol/UV filter, cause you can't use them both, at he same time. What would the results be? (you gotta wonder)

M3 requires both polarization filters and uV cut. Probably because the polarization filters attenuate uV too. It's hard to make polarizing filters that cover a wide spectrum. So the constraint that M3 requires uV cut may just be a practical one.

Edited to add:
Another reason comes to mind. Fluorescence is not polarized and is not a specular effect.

[GWGill]

In short, in view of the results that had been noted by others in this forum in the past which were based upon the same process when compared to the results that I have been getting, I am stumped!

It's a truism that profiles are useful only to the degree that the measured values represent what we actually see.

This is why graphic arts color measurement uses 45/0 degree measurement, and not the 0/d geometry that is used to measure colorant content - we regard color as the appearance of the non specular portions of a surface.

So the question has to be asked :- in what way is using M3 better measuring the appearance of your prints ?
Is there something special about the illumination used, or the conditions it is viewed under  - i.e.
that the illumination is polarized, or that the viewing conditions filter out some polarization ?

Because the typical purpose of M3 measurements is to measure wet ink values for print process control, NOT to better match visual color. Looking at the measurement values and noticing lower L* values is worse than meaningless if that's not the actual visual L* value. It's worse because it causes the profile generated values to be wrong.

This is what Ray Cheydleur of X-Rite has to say on M3:

M3 – POLARIZATION
Colour Assessment independent of the surface
• Polarization reduces reflections caused by the surface reflection or bronzing
• On method of density comparison between wet and dry inks.
• It also removes UV equivalent to M2.
• Is used in ISO 12647 for as an option for density process control.
Attention: There is no viewing condition that matches this measurement condition

This is not to say that measurements made with standard M0/M1/M2 conditions are always right - they may not perfectly represent the appearance of paper with particular textures and other surface characteristics, but there is certainly no reason to think that M3 should be used to generate color profiles, or that M3 is somehow better than other measurement modes.

[ Speculation - for some types of paper some (unspecified, color dependent) mixture of polarized and un-polarized measurement values may be more realistic. Or a combination of 45/0 and 0/d measurements. Figuring that out is a research project for someone though! Turning it into something practical is yet another project for someone. ]

[Rainforestman]

Hi Mick,

I also have a Barbieri LFP S3 and I like making M3 profiles and I print on an Epson P9070 (P9000 model in Australia).  I love the profiles that the Barbieri makes.  One thing that I have found from looking at the profiles in worksheet mode in ColorThinkPro is that the white points with M3 can be a bit on the warm side, so I take an M1 reading with an iPro2 and edit the white point in the profile whilst optimising it.  To do that I use Basiccolor IMprove before taking the data into Basiccolor Print.

I appreciate that there are a range of opinions on M3.  It’s an area of private research that I want to work on.

Recently, I purchased a Barbieri LFP S3 to create ICC profiles using M3 (polarizer) based readings on various matte inkjet papers. Readings were collected from targets using Barbieri Gateway and subsequently imported to i1Profiler for the creation of the print profiles. My first attempt was to reprofile Epson Hot Press Bright which I compared against the original profile that I made 2 years back using an i1Isis-1. The resulting M3 profile was as amazing as has been described in other posts in this forum i.e. much lower L* D-max, much larger colour gamut and clearly more shadow separation in the print from the M3 based profile versus a print from the old M0 and M2 profile. 

I know that at least a few of you have experience with this method of ICC profiling. I recall having read in one past post by Josh-H that he had to laboriously adjust the rotation of the polarizing filter to find the exact sweet spot for it. But, the filters on my LFP have registration pins with corresponding sockets on the head into which the pins seat. So, there is no way that I know of to adjust the rotation of the filter. It should be in that sweet spot. At least one would hope. I'm aware that any improvement in shadow detail is negligible for gloss stocks. But, I was under the impression that all matte stocks would gain some improvement. Certainly the profile shows it. So, my question is does anyone have any idea what is going on here?

With regards to Josh Holko mentioning the polariser had to be rotated - I posted a reply to his reply on that earlier post.  Like your polariser the Barbieri has a pin on the filter that locates the polariser precisely.  It would be a nightmare to try and align that manually and I wouldn’t trust results if you had to do that.

Last week, I made an M3 profile for Hahnemuhle Torchon which contains OBAs as does Hot Press Bright. This time, the profile as viewed in ColorThink showed the same huge improvements as all the others and there was a slight improvement in the M3 print versus the print made from a profile created using i1Isis-2 (M0, 1, 2). I should add that each print pair was made on the same printer (either Epson 7900, SC500 or SP4900). So, now I wonder if these improvements have some relationship to the OBA content. In short, in view of the results that had been noted by others in this forum in the past which were based upon the same process when compared to the results that I have been getting, I am stumped!

I have also profiled Torchon using M3 and a range of other papers - mainly Canson.

I’d be happy to compare notes with you offline from the forum here - maybe we can swap some profiles and measurement data.  Maybe we can then report back to the forum with some joint findings. 

[Rainforestman]

M3 requires both polarization filters and uV cut. Probably because the polarization filters attenuate uV too. It's hard to make polarizing filters that cover a wide spectrum. So the constraint that M3 requires uV cut may just be a practical one.

Edited to add:
Another reason comes to mind. Fluorescence is not polarized and is not a specular effect.

Agree with you Doug.

ISO13655:2017 states that M3 filters must also satisfy the M2 condition and that a spectrophotometer with tungsten illumination should have UV filtration with the following transmittance characteristics:

— greater than 0,85 in the visible range above 420 nm;
— less than 0,50 at 410 nm;
— less than 0,10 at 400 nm;
— less than 0,01 at 395 nm.

[Mick Sang]

Thank you all for your input. It has been helpful.

But there is another issue potentially at play. Any lint or dust particles present when scanning can reduce shadow resolution. And it can be significant. Dark patches are by far the most sensitive to this effect (see a recent post by me and a response from Chromix's support). The variation you see may well be from this problem.

Thank you, yes, I read this post and found it very interesting. I have seen slight improvements to the print results through all my new M3 profiles except for one (Hahnemuhle Museum Etching) where the M3 profile showed noticeably less shadow detail / separation than the print through the older M0, M1 and M2 profile. I will therefore clean the Babrieri thoroughly, lightly brush off all the the charts, reread them all and report the results.

The larger gamut and improved DMax in an M3 profile is fiction. Pure and simple. The gamut isn't larger and the actual DMax (comparing a printed black patch using M0-2 v M3) is also the same.

I was aware of the fact that the Dmax as shown in ColorThink was "fiction." But, I was not sure about the gamut, although I presumed it was also not to be trusted. Your input helps to confirm that. Nevertheless, to me, it was only important insofar as it served to confirm that the readings were in fact being recognized as polarizer based. As this was my first go at these readings some confirmation was helpful. Otherwise, I realise that d-max is d-max and the readings would not change the density. It would only change the way it was being "seen" by the spectrophotometer.

There is one area that M3 profiles excel and that is when the paper is difficult to profile in deep shadows.

So far all the papers I have profiled this way have been relatively smooth except for the Torchon whose texture is large or lumpy for lack of a more suitable term. Next I will profile William Turner. That should fit the bill re your comment.

I also think that it is good to remember that it's Prints that are Produced/Sold/Admired and not the technology behind producing then; meaning to say, let your eyes be the ultimate judge and not raw data and numbers. That second part is, just a guide and one of the many tools used to aid in getting where you want to be.

Thank you for this reminder. I appreciate it. But, I am not one to lose sight of what we're after here. I do get deeply into this stuff but only with the goal of achieving improvements in the print clearly in my sights at all times.

This is Inter-Instrument agreement. Two different devices, two different manufactures, perhaps different technologies.

Yes, thank you again. I'm not comparing readings. I am comparing different printed results through different ICC profiles made from different processes. I know that the difference between spectros is playing a part. But, since X-Rite doesn't make an i1Isis with M3 capability, I have to use the Barbieri with its polarizer.

there is certainly no reason to think that M3 should be used to generate color profiles, or that M3 is somehow better than other measurement modes

Personally, I have never thought that M3 should be used to create all profiles. But, I do think that M3 and specifically polarizer readings can provide better shadow rendering than would otherwise be available through the usual means. Plenty of evidence of this has been provided over the past few years on this forum and elsewhere. To me, it is a tool. If any tool improves the result that I achieve in print, I will adopt it.

I am aware that the standard thinking regarding M3 readings is that their place is specific to fluorescent inks and wet vs dry inks etc. But, it has been documented that it can result in improved shadow rendering particularly on matte papers. That said, while I am seeing obvious improvements, I had thought that I would see vastly more dramatic results.

I also have a Barbieri LFP S3 and I like making M3 profiles and I print on an Epson P9070 (P9000 model in Australia).  I love the profiles that the Barbieri makes.  One thing that I have found from looking at the profiles in worksheet mode in ColorThinkPro is that the white points with M3 can be a bit on the warm side, so I take an M1 reading with an iPro2 and edit the white point in the profile whilst optimising it.  To do that I use Basiccolor IMprove before taking the data into Basiccolor Print.

I have seen the warmer result from the white point through quarter and midtones. I would love to pursue a means to improve upon that to achieve the white point of an M0,M1 and M2 profile through the quarter and midtones with the M3 shadows would be ideal or so it seems at this point.

With regards to Josh Holko mentioning the polariser had to be rotated - I posted a reply to his reply on that earlier post.  Like your polariser the Barbieri has a pin on the filter that locates the polariser precisely.  It would be a nightmare to try and align that manually and I wouldn’t trust results if you had to do that.

Thank you, yes, I remember your post well. It is because of your post in fact that I made it abundantly clear to my sales rep before purchasing my LFP that the pin registration was available. There was speculation that old models apparently didn't have it. Otherwise, perhaps it was broken off from Josh Holko's device.

I'd be happy to continue this discussion off line and compare notes and swap profiles, although I suspect that you are likely eons ahead of me. So, I'm not sure what I could offer to you. I will send a note.

Once again, I thank all of those who have replied so far. This is a new and exciting avenue for me to pursue. I will get back with the results of a reprofile from M3 readings after my cleaning of the Babieri to try to mitigate any interference from lint.

[digitaldog]

I was aware of the fact that the Dmax as shown in ColorThink was "fiction." But, I was not sure about the gamut, although I presumed it was also not to be trusted.

Nope, here's why:
http://www.colorwiki.com/wiki/Color_Management_Myths_26-28#Myth_26
Trust it!

[Mick Sang]

Nope, here's why:
http://www.colorwiki.com/wiki/Color_Management_Myths_26-28#Myth_26
Trust it!

Generally, I do trust it. I was referring to the fact that M3 profiles in ColorThink show such a huge increase in Dmax vs. the M0,M1 and M2 profiles and gamuts that are 200,000 plus colours larger. For example L14 to L2 when the actual corresponding density readings for matte papers are in the area of 1.8 vs. 2.6 for gloss / semi-gloss papers. So, in itself, the Dmax L* value of the M3 profile is misleading or "fiction" is it not?

[Doug Gray]

Generally, I do trust it. I was referring to the fact that M3 profiles in ColorThink show such a huge increase in Dmax vs. the M0,M1 and M2 profiles and gamuts that are 200,000 plus colours larger. For example L14 to L2 when the actual corresponding density readings for matte papers are in the area of 1.8 vs. 2.6 for gloss / semi-gloss papers. So, in itself, the Dmax L* value of the M3 profile is misleading or "fiction" is it not?

The colors a paper can print are maxed out at the gamut edges. These are where one or more RGB values is either 0 or 256. M3 profiles simply read the colors differently from M0/1/2  but one sees the same colors.

For instance if you print a saturated cyan using RGB 0,255,255 it can measure something like Lab 45,-40,-20 with M2 but 42,-50,-30 with M3.  But they are the same exact printed color. And M2 (absent OBAs) will give a reading much closer to what one sees under daylight.

Same with black patches like 0,0,0. These might read L* of 3 with M3 or 18 with M2. But if you view the patch in daylight it will match the L*=18.

OTOH, if you set up a special lamp with polarizers and view it with another polarizer, properly aligned, you will see the more saturated cyan and the darker black.

[Mick Sang]

The colors a paper can print are maxed out at the gamut edges. These are where one or more RGB values is either 0 or 256. M3 profiles simply read the colors differently from M0/1/2  but one sees the same colors.

For instance if you print a saturated cyan using RGB 0,255,255 it can measure something like Lab 45,-40,-20 with M2 but 42,-50,-30 with M3.  But they are the same exact printed color. And M2 (absent OBAs) will give a reading much closer to what one sees under daylight.

Same with black patches like 0,0,0. These might read L* of 3 with M3 or 18 with M2. But if you view the patch in daylight it will match the L*=18.

OTOH, if you set up a special lamp with polarizers and view it with another polarizer, properly aligned, you will see the more saturated cyan and the darker black.

Fascinating! Thank you for this detail. Hence the reason for my distrust of the M3 values as being more accurate than the other M reading results.

I cleaned the Barbieri thoroughly, ensured all the charts are dust and lint free, reread them all and reprofiled the Hahnemuhle Museum Etching. Next, I will print my test image with the new profile to see if the cleaning got rid of any interference from lint which may have played a part in the previous result with that paper which showed poorer shadow rendering than the M0, M1 and M2 profiles of the same paper. This was the only paper to present such a result. All others which I have profiled to date, aside from the Museum Etching have shown slight to significant improvements.

[Doug Gray]

Aside from the "accuracy" issues, it's worth pointing out again that M3 with matte paper can provide more shadow smoothness. Especially in B&W printing. This is because the low L* values are spread out over more LUTs in the profile. A side effect is that whether you print Rel. Col. or Perc. the results are nearly identical. In effect you have a profile that will print perceptual or Rel. Col. with BPC. Soft proofing Rel. Col. w/o BPC won't match physical prints and show paper black won't match any print intent. But one can achieve smoother dark shadow rendition.

Another use is with metallic papers. Profiling these is problematic with M0/1/2 as well as M3 but M3 may produce more pleasing results. Colorimetric accuracy is not the goal with metallic papers. Rather the specular effect is used for artistic purposes.

[Mick Sang]

A side effect is that whether you print Rel. Col. or Perc. the results are nearly identical.

It's interesting that you mention this. With respect to the result in print which caused me to post my question regarding M3 profiling vs M0, M1 and M2 profiles in the first place i.e. an M3 profile of Hahnemuhle Museum Etching where the test print showed a loss in shadow rendering versus prints made through M0, M1 and M2 profile, the first M3 print was made using the perceptual intent. Subsequently, I tried a print using rel. col. While the latter still showed weaker shadow rendering versus the M0,1 and 2 profiles the result was clearly superior to print using the perceptual intent. Go figure!

My 7900 suddenly has a encountered a main board issue which should be repaired tomorrow (let us pray). Once it's back up, I will run that same test again through the new M3 profile and report the results.

[Mick Sang]

So, it has been just over 5 weeks and our 7900 on which the tests were made has finally been fixed. The five weeks were wasted waiting for a Decision One tech;  when he arrived, various tests determined that the main board needed to be replaced; Epson sent the wrong board and on & on. Finally, I was able to reprint on the Hahnemuhle Museum Etching using a new M3 profile which was created from readings made after a thorough cleaning of the Barbieri LFP and every chart. The result, I'm sorry to say, was exactly the same as before. The print from the M3 profile using Red Col or Perceptual showed less shadow separation that the print run through the M0 profile.

All other M3 profiles made so far on matte papers with and without OBAs have resulted in prints with noticeably improved shadow rendering with the exception of Hahnemuhle Museum Etching. So, once again, for this particular paper at least, I'm stumped.

[Doug Gray]

If you post the CGATs files for M2 and M3 Hahnemuhle Museum Etching paper I would be interesting in analyzing the differences. Quite a curious phenomina that's outside the norm.

[Mick Sang]

Here is a link to the readings from the i1Isis and the M3 readings form the Barbieri LFP with polarizing filter: https://www.dropbox.com/s/2wyspf4ghcy79we/Hahnemuhle%20Museum%20Etching.zip?dl=0.

I have never uploaded a file or a link or the like to this site. So, please let me know if there is a problem.

Thanks very much,
Mick

[Doug Gray]

Here is a link to the readings from the i1Isis and the M3 readings form the Barbieri LFP with polarizing filter: https://www.dropbox.com/s/2wyspf4ghcy79we/Hahnemuhle%20Museum%20Etching.zip?dl=0.

I have never uploaded a file or a link or the like to this site. So, please let me know if there is a problem.

Thanks very much,
Mick

Not a problem. I was able to download the files. However, it's not necessary to post them to dropbox. If you zip the files you can attach the zip file to a post here. When you reply, just select attach files, browse to the zip file and add it to a list. You can attach up to 4 files. zip or image. The image files show up at the bottom of your post and the zip files just show as a one line attachment note.

[Doug Gray]

Mike,

I think I see the problem. There are fairly large deviations in a*/b* at very low luminance on the M3 v the M2 measurements. I'm running some comparative tests but the differences are large enough you should see some of the effects soft proofing dark shadows. Ignore the luminance differences since those are intrinsic to M2/3 but the a/b* differences will be noticeable. Note huge, but noticeable.

I'll write up what I see and post some graphs in a bit.

Added:
Upon closer look, the large b*/a* deviations that occur at low L* on the M3 data are exaggerated by the low L*. When the data is looked at through the M2 data set, the a*/b* deviations become much smaller. OTOH, the L* slope changes are more problematic. Turns out the major difference is a lumpy L* instead. There is a significant slope change in L* between about 5 and 15 on the M3 profile. The profile tracks this and alters the smooth transition BPC affords the M2 profile which doesn't have that slope change. Interesting stuff.

Here's an interesting graph. It's the L* of the M3 and M2 perceptual profiles (BtoA0) that is reported by the M2 RelCol profile (AtoB1), which more closely approximates what we actually see, when plotting requested L* (what would be in an image).

The blue line is the M3 Perc. response, the orange line is the M2 Perc. response. The reason is that the M3 measurements show a large increase in L* of from 3.5 to 8 between rgb 3,3,3 and 10,10,10. This is interpreted as a real change, not an artifact of polarization rejecting specular reflections, which our eyes can't do. The result is a lowering of the actual print response.


Another problem is apparent gamut increase. Also attached is a graph of the a*/b* gamut at L*=30. The outer circle is the M3 profile and it shows that it can print colors about twice as saturated as the M2 gamut. It can't. In fact the gamut boundaries are limited by the device RGB values when they clip at either 255 and or 0 and those clipping points are the same regardless of whether the spectra is measured with polarized filters or not.

[Mick Sang]

When you reply, just select attach files, browse to the zip file and add it to a list. You can attach up to 4 files. zip or image.

Thanks for this, Doug. I have removed the files from DropBox and attached them herewith in case anyone wants to view them in future. Stupidly, I was searching for an "Attachment" function in the top list while it was in bold beneath the post area all the time. Dah!