The main conclusion I come to is that all this measuring is inconclusive. One needs to make prints and use what one thinks looks best. For example, in the Northlight B&W test page, I found a slight margin of preference for ABW Normal given how well it opens shadow detail, but the ICC profile is not far behind. Based on these tests, previous tests, and previous test prints using both an ICC and ABW-based workflow (but no QTR), I would normally opt for the ICC-based workflow because I really do like the ability to soft-proof my actual photos before sending them to print.
many thanks for submitting your test data. it is really appreciated.
I have some small questions:
- the ICC used was the Ilford provided one or a custom one?
- It was used in relative colorimetric + black point compensation or in different way?
- regarding the a* and b* values of the white patch (paper white) I see a strange inconsistence for the ABW-normal case. Usually I never see this kind of difference in different instances of a paper white measurement. Are you sure do you have not accidentally swapped the a* with the b* values for this single wedge one (L* 100 of the ABW-normal set)? Because the other two are good matching and if you swap the a* and b* values of the not matching one (ABW-normal) it seems again a good match with the other two.
However I think that your measurements are not inconclusive at all.
I have taken your data and I made the same Lab graph I usually do for my linearization test for all the 3 cases you showed (ICC, ABW-normal and ABW-dark), you can find it in the first attachment.
The main difference is that I have highlighted the linear straight line (visible as dotted line) between measured black and measured white, because this line is my usual adopted reference target for my L* linearization.
- As visible, in the first (left) graph the ICC L* shape is a typical "S" shape, with slightly compressed shadows/highlights and slightly expanded midtones
- in the center graph the ABW-normal L* shape is always a little bit too high
- the right graph clearly show a quite good match of the ABW-dark L* shape with the theoretical linear target, this is clearly the best of the three if a true linear approach is desired.
Please, note this is not the first case I find this kind of coincidence.
In my second attachment you can find a similar comparison I have done using Ilford Gold Fibre Silk.
I used my Epson R3000 and Epson K3 VM inks. ICC was the Ilford provided one used in relative colorimetric and black point compensation on.
The strip was made of only 18 B&W wedges (not 33 like yours), the measurements were done with my humble ColorMunki.
- As you can see the ICC L* graph is not so good as yours, remaining always under the L* target line
- ABW-normal is very similar to yours, remaining always over the target line
- ABW-dark is the best of the 3, with good match to the L* target line and very similar to yours too
In ABW the gray tone is more neutral than the ICC one, and not so slightly bluish as the one you found (ColorMunki is UV-cut).
Overall in ABW the gray tone is very neutral, becoming a little bit warm only in proximity of the paper white, which is a bit warm for this paper.
In addition note that the L*, a* and b* values of the paper white is near perfectly matching in all the 3 cases.
In my humble opinion these data are very interesting because they show that the ABW modes provide results that are similar for a 4900 + Gold Mono silk and R3000 + Gold Fibre silk. (There is a little difference in the overall b* shape, but in different paper/coating/inks it could be expected).
Your ICC is clearly better performing that the one used in my case, ABW-dark is the best choice for L* linearity in my opinion and gives deeper blacks than ICC, but I agree with you that if you want to do some tone grading and you will be able to see on screen the results before printing, the ICC workflow is your solution.
I hope it will be useful.
Any comment is welcome.