One - you could try building a device link with higher resolution in the tables. In my testing I found that the resolution of the tables in the link played a significant role in the accuracy of the link and its ability to mimic the conventional transform.
Two - you could check the link vs the conventional transform using ColorThink Pro.
Open a CMYK color file (there’s one in the stunt files folder) into the worksheet
apply the link to the file to generate the new CMYK
apply your destination profile to the end of this workflow to get correct Lab values for comparison. **
click ‘new’ in the workflow region and Pro will open the same CMYK file in a new workflow (same sheet) for comparison
open your source & destination profiles into this workflow
turn on delta-E and see if it shows any difference between them
**(Pro will use a SWOP profile to calculate the Lab values for the CMYK numbers as a default)
In the 3rd step, when you “apply your destination profile…to get correct Lab values…” The Lab values would be derived from the AtoB tag of the destination profile, right?
