Datacolor User to User Support Group.

David Miller said:
- There's supposed to be a blue cast on the white patch (meaning:
the white of the paper that you're profiling/measuring); because that's
the color of the paper. Most papers measure with a slight blue cast.

C. David Tobie said:
- Whites turning blue indicate that you are using an artificially
whitened paper. Checking the checkbox in Ref White should solve that.
Checking the checkbox in Ref Black should reduce your "grayveil"
effect as well, especially on matte papers.

- If the whitepatch (paperwhite) of your target measurements (after
calibrating on the white tile!) is showing up as a bright Cyan,
instead of white, that is hardware issue, and you will need to contact
support at ColorVision.com to get your spectro replaced. But first,
clean the whitetile with rubbing alcohol and a paper towel (don't use
cloth, it may have whiteners in it). See if white then measures as
white; if its still bright Cyan, then contact support for a replacement.

- Then you will need to contact colorvision support to get a
replacement unit.

What is usual and what is wrong?
Paper is Hahnemuhle Natural Art Duo. Matte paper.
 - Grammage g/m2 256
 - Thickness mm 0,38
 - Whiteness % 88,5
 - Opacity % 98,5
 - Media Colour natural white not bleached
 - pH-Value total 8,1
 - Cobb g/m2 74,0
 - Ink limit % 250
 - Water resistance very high
 - Special features coated on both sides
Chart is printed in Qimage (ICM off...), soft proof is realised in
Qimage (I haven't tried in S3P), prints are printed in Qimage (profile
with black ink compensation).

Calibration of the spectrocolorimeter on the base (white disk) gives a
light gray, 0.1 density.
Measure on a not printed piece of paper (natural art duo) gives a
light cyan. Which could be explained, as you said, by the artificial
white of the paper.
The prints are quite good (better when calibrated on the paper instead
of the white disk of the base).
Does the spectrocolorimeter need a replacement???

The issue is soft proof. Automatic ref white (with calibration on the
base), soft proof is yellow and psychedelic :D
Reading help of spyder3print make me understand that I must use ref
white in order to reach a white similar to the paper white. It was not
clear in my mind.

I will try ref black for the lack of contrast. The idea is also to
reach the black of the paper?

Black ink compensation in my printing software doesn't do anything to
my eyes, at least in soft proof. I'm not sure about prints. IS there a
link between black ink compensation and ref black in spyder3print?

I know that I repeated a lot of things and that I feel like an idiot,
but advice are a little bit opposite.

Anyway, thank you for your answers. My understanding of the process is
better now :D