Datacolor User to User Support Group.

On Dec 19, 2009, at 11:22 AM, Bob wrote:

> So a few days ago I created a target for the first time in a while;  
> I'm using the latest SpyderPrint 4.1.1 with an older model 1005; it  
> has a serial number under 1500. To my surprise, an evaluation print  
> showed neutral colors coming out rather brown.
>

The entire profile will be thrown far off if your paper white  
measurement has a huge
color cast (see below). That's why.
> After repeated remeasurings, I started fooling around with it and  
> came to the conclusion that at a minimum the white calibration was  
> faulty; if I use a piece of white paper to calibrate rather than the  
> original target, the resulting print is much more neutral.
>

Shouldn't do that, though. If this "fudges" the calibration so that  
your paper
white measurement doesn't have the huge color cast (see below), that  
still doesn't
make your overall results correct.
> I'm not sure, however, if the problem lies in my spectrocolorimiter,  
> the calibration target, or elsewhere, and would like to know what my  
> options are beyond buying a whole new kit.
>

It's the calibration tile. (Better to call it the "tile", rather than  
a "target",
to avoid confusion... better to always, and only, refer to the actual  
target
print as the "target").
>
> A bit more detail: I'm trying to profile some Epson Premium RC Semi- 
> Gloss for use on a Canon ipf6100 (there was a great deal on this  
> paper at B&H). If I use the white target to calibrate, the Epson  
> Semigloss paper has a LAB around (97,1,-90) -- *really* blue; this  
> paper isn't exactly neutral but there is no way that it is that blue.
>
Yes, that's an incorrect measurement, and it's the reason for your  
problems. Paper white
is used as a scaling measurement for the entire profile and if that's  
wrong, the entire
profile gets color cast in the opposite direction.
> Some Canon Photo Paper Plus II, which is visibly whiter, measures  
> around (97,3,-5).
>

That's a good measurement, that's fine.
> Some Inkpress Matte 80 measures around (98,3,-98).
>
>

That's a very bad measurement, like what you're getting for Premium RC  
Semi-Gloss.
> If I calibrate using the Canon paper as the white target, the Epson  
> semigloss measures around (91,-2,-1), while the Inkpress Matte  
> measures around (93,0,0).
>

Those are obviously not the horrendous values that throw the entire  
profile off,
but they're not correct, either. The semigloss would end up measuring  
close to the
Canon Photo Paper, and the inkpress matte still wouldn't have a  
perfectly neutral
color cast.
> I'm really not sure what to do here; this does seems surprising  
> given how far apart these papers are when I use the official white  
> target to calibrate. And yes, this is repeatable; the numbers change  
> a little bit but within a few percent the character of the result is  
> the same, going back and forth between the two calibration targets.
>
> Thanks for any suggestions.
>
>

First suggestion is: try cleaning your calibration tile, even if it  
doesn't look
dirty. Use some soft cloth and Windex; then recalibrate on the tile,  
carefully,
and try measuring those paper whites again. When you calibrate on the  
tile, pay
close attention to getting the nose down "cleanly" on the tile, by  
putting
the spectro right in the base, with the feet lined up perfectly, so  
that there
aren't any gaps around the nose on the spectro.



David Miller
Senior Software Developer, Digital Color Solutions
Datacolor