I think I might have confused you by what I started out doing. I initially wanted to compare the prints between color and piezo. When i used the profiles from each of different methods, they did indeed seem to match when I soft proofed them. When I went to print them, I found that the print on the piezo did not match the soft proof image, whereas the color printer did match. The way I had generated the piezo profile was from Walker’s suggestion (from a post) that said to use a 21 step target. I thought that might have been the problem. I wanted use the 256 step profile, but the method in the icc generation did not work for me (or the other person on the forum, as documented). You then asked me to generate a 256 target for both the color and piezo, and do a comparison between them. I did, and found that there were (what I thought were major) differences between the target in one particular area. The fact that I wasn’t using a 256 step target for the profile, had me concerned that using a 21 step target was missing that problem found in the 256 target. I still believe that is the problem. The other person I pointed you to, also wanted to use the 256 target, put had the same issue that I had.
I believe that a 21 step target is not enough steps to capture the nuances of the curve. The image I’m attempting to print, I believe, falls within this area that we found the variance. I feel that if we could get a profile generated using the 256 step target, the problem would be ultimately solved.
P.S. It is not the shadow detail that is wrong… it is tone of the building itself…
Note: I can send the details of the comparison of each of the 256 points in a pdf file, which I am unable to post. That file seems to contain some data for the piezo that is suspect and the reason we can’t use it to generate a profile from it. For example, the color target for 256 has one negative value for the b* value where the piezo 256 target has over 20 (I stopped counting) negative numbers. That seems to indicate that piezo target is not linear. I’m new to this and may be wrong… but I can’t use the 256 measured results. I think that tells us something…