QTR-Quadtone RIP

I decided to recommence the linearisation process.  In doing so I
realised that part of my problem with darker than expected shadows may
have been at what point in the process I set the "black boost"
parameter in the curve definition.  I find the "Calibration.pdf" and
"Eye-One-ReadMe.pdf" clearer and easier to understand than "User
Guide.pdf" (sorry Tom), but the user guide is a little more detailed
on issues like black boost.

I gather that I turn it off for setting the ink limits and
partitioning the black and grey inks (UC on R2400).  But if I decide
to use it, when do I set it?  

I now gather that when I first print out the 21*4 step wedge, I need
to decide at that stage whether to use the black boost, and only when
I have an unlinearised step wedge with a sufficient dMax do I
linearise the curve definition with these measurements.  The ICC
profile is also therefore based on measurements from a linearised step
wedge using black boost.  Is this correct?  

In starting again I came across a couple of odd measurements.  I
compared the unlinearised dMax numbers from my first attempt with
those from my second, where I was much more careful in measuring in
patch mode.  Here is a chart of the two:

http://www.ferdinand-paris.com/v1-v2.png

I assume that this sort of variation comes from the printer.  Or is it
from ambient light when the measurements were being taken?  Is this
sort of variation in measurements taken a couple of days apart normal?
 I had more confidence in my second set of measurements until I looked
at this chart.  I wonder if they signify that I need to repartition
the inks.

Finally, in examining the variation between the four samples in the
4*21 step wedge, I noticed that there was a fair bit of difference in
the samples for "30".  All the others are pretty close, which I assume
indicates good technique, but not so for the "30" samples.  The ones
at A1 and A3 at the start of a row consistently measured brighter than
those at U2 and U4 at the end.  Again, I assume that this reflects
printer behaviour, but it's odd.

Ferdinand


--- In QuadtoneRIP@yahoogroups.com, "ferdinand_paris"
<ferdinand_paris@...> wrote:
>
> Lou
> 
> I have calibrated the screen using the i1 and linearised the QTR using
> the i1 so I was hoping for some consistency.   The X-Rite XR Pro
> software that I was using for screen calibration gave a range for
> screen brightness, whereas i1 Match specifies a precise value.  I may
> opt to set it below this level, but first, given the issues with the
> i1 reading the step wedge in another thread, I think I will start
> again and proceed carefully.
> 
> Thanks
> 
> Ferdinand
> 
> --- In QuadtoneRIP@yahoogroups.com, "Louis Dina" <lou@> wrote:
> >
> > Ferdinand,
> > 
> > Blocked up shadows in prints (compared to your monitor) is often due 
> > to setting the monitor luminance too high.  Some software vendors 
> > recommend very high monitor luminance, which I consider too high.  
> > 
> > If you have an accurate printer profile, I would generally trust it 
> > to reproduce fairly accurately.  Your Eye One will read the luminance 
> > levels accurately and translate those into a profile, with correct 
> > density and color (not perfect, but very close).  There are really 
> > no "user settings" that can skew your results.  
> > 
> > With monitors, however, the user can choose both the luminance level 
> > and the color temperature, which can influence the final display.  If 
> > your monitor is set to a luminance above 100 cd/m2, this could be 
> > your problem.  Your image will appear very bright, and during 
> > editing, you may be inclined to darken the tones in Photoshop.  These 
> > edits end up in your final numbers, and ultimately in your print.  
> > You would need to view your print under bright lighting to get a good 
> > match.  When viewed under less brilliant lighting, shadows can look 
> > very dark, blocked up, and with little separation.  
> > 
> > Try calibrating your monitor to about 90 cd/m2 and then compare your 
> > monitor to your print and see if this improves your match.  
> > 
> > Lou
>