Digital BW, The Print

I am using a softproofing method described by Dan Culbertson 
in "RGB_Softproof_Quad.pdf" (in the Files / Image Processing 
and Workflows section of this group) fairly successfully in 
combination with a Roark warm curve tweaked to yield darker 
shadows and smoother midtones with MIS Variable Mix hextone 
inks. I tweaked the Roark curve (softproofing on) to yield a good 
onscreen 21-step grayscale, and am getting prints that are a very 
close match. However, the RGB softproof and print are not a 
terribly close match to the original grayscale image onscreen 
(both open in separate windows). I'm finding it hard to tweak the 
Roark curve properly to make the softproof 21-step match the 21-
step in the unmodified window (even factoring in the neutral --> 
warm shift). As a result, a final edit must be made to the image 
after softproofing and just prior to printing. I wish I could avoid 
having to target the image this way for printing. Grayscale --> 
softproof --> print consistency would be nice.

Still, I'm really surprised that the softproof so closely matches the 
print, particularly considering that I'm using ColorVision's 
ProfilerPLUS to create the printer profile, and this product is 
made for profiling color, not b&w, let alone the MIS Variable Mix 
inks.

I can't get my head around exactly how ProfilerPLUS is 
translating the VM tones into a profile, or exactly how the red, 
green, and blue curves in Roark's curves work. How, for 
example, does one modify the R, G, & B curves to force the 
Epson driver to use more black ink?

Anyway, this workflow seems promising. I'll keep posting 
questions and report on progress. If this turns out to be a good 
scan-to-print workflow, I'll post the details in this forum. Thanks 
for your support to date.

Jacques Cornell