The vm-sepia inkset uses the same "densities" as the standard MIS-VM inkset.
As such, I hoped that the curves that currently existed would be at least
close.  In fact, you can see the posted scan of the Epson 870 vm-s test
strips that used the standard MIS VM curves for that machine -- they are
close, with no significant flat spots or posterization.  They are a bit
darker than the standard MIS VM test strips, but well within a useable
range.  So, the first thing to check if the vm-sepia inkset produces bad
results with all the standard vm curves is the settings in the driver.

What I have noticed most is that the standard MIS VM "warm" curve -- which
withholds the toner -- often does not work as well as the other standard
curves.  By withholding the toner, the vm-s prints should be relatively
neutral.

The reality of these -- and no doubt most other -- inks is that while the
densities of the vm-sepia match those of the standard MIS VM inks at one
point, they may not match at another level of application.  That is, the
response curves of the two inksets are not completely linear.   I suspect
that because of this, the more radical the curves, the more likely that the
non-linear responses of the inks on the paper will show up.

Many of the MIS VM "warm" curves I've done have very radical "green" curves
that totally withhold the toner until absolutely necessary to "turn on" the
black ink.  While this often works for the MIS VM inkset (depending on
whether the particular driver can handle that steep a curve without
artifacts), it can result in less-than-optimal tone distributions in the
vm-sepia inkset (midtones that are too cool and make the shadows look
warm -- which they are), and sometimes -- but not always -- uneven density
distributions.

With my quad printers (1160 & 3000) I get very nice "neutral" curves by
putting some sepia toner in the mix all the way.  The resulting prints are
dead neutral (X-Rite spectrophotometer readings: c=m=y) in the midtones,
rolling off to the paper white in the highlights and black ink warm (y = c +
0.03) starting at about 75%.

The 870 is the first hextone machine I've tried to adjust, and I must say
that the neutral curve did not end up quite as nicely as I would have liked.
The shadows get slightly warmer than they do with the quadtone machines,
with a peak of about 0.04 units warm at 90%.  For example, the 870 neutral
curve I made has X-Rite readings at 90% as follows: c=1.38, m=1.41, y=1.42.
This is OK, but will get more so with warming.  For example, in the fade
test currently being done with this inkset, after the first 100 hours the
90% warmth was up to 0.05 units.  (70% was dead neutral.)

(A new "non-warming"  toner has knocked 0.01 units or warmth off the 90%
point at start and at 100 hours, but we'll just have to see how it looks
after much more testing.  The "non-warming" (really just more stable over
time) toner has a slightly lower gamut, but still has very ample tone to
make full sepia prints.)

So, what I have found, in general, is that making a curve that is between
the "warm" and "mw" curves usually makes for better "neutral" vm-s prints.
However, the hextone machines may not do quite as well at the neutral end of
the spectrum as do the quadtone machines.  See the 870 vm-s test strips that
are posted.  Go to the Files section of this forum:
http://groups.yahoo.com/group/DigitalBlackandWhiteThePrint/files/
Then follow the links to
Ink Sets > MIS > MIS Variable Mix > MIS VM Neutral-Sepia.  The image is
called, "VM-Sepia-Standard 870 v. 1 curves.jpg."

To make a better neutral curve for the vm-sepia inkset, start with the
better of the "warm" or "mw" curve and move the "green" (toner curve)
points -- very little if at all in the deep shadows, but so that they roll
off to close to horizontal by about the 50% point, but still not at the top
of the curves box.  For each green point that is moved, for example, up if
starting with the "mw" curve, move the blue point curve that many points
down to offset the density change.  The blue curve ink (yellow ink position)
in a hextone vm machine is the lightest gray ink.  It is a hair lighter than
the dark toner, and a bit darker than the light/photo toner (green ink
curve/magenta ink position).  So, offsetting moves keep the densities about
right.  Some final tweaking will be necessary.  Print 21-step test strips to
see the results of moves.

These Photoshop adjustment curves can be a bit tricky.  For those who have
not worked with curves, you might want to check out how the vm-s inkset is
working for others with the machine you have before diving in.

The vm-sepia is what I use as my regular inkset in the 3000, and I'm very
pleased with it.  However, with other machines, there may not be curves that
have been fine-tuned for this inkset.

Good luck.

Paul
http://www.PaulRoark.com