Digital BW, The Print

"Jacob" <flashinet@...> wrote:

> I'm trying to figure out QTR, getting ready to print target.
> eye-one-readme.pdf says:
> 
> The 21x4 target is recommended for the best accuracy. 
> It has 4 patches for each step value and the QTR software 
> will automatically average the 4 measurements
> which gives a more accurate result.
> 
> My question is: what will be more accurate with this target
> - measurements or printer profile? I understand that every 
> patch measured 4 times and averaged gives more accurate 
> measurement but will profile be better than with 51-step
> target?

I think multiple sampling will just give a better read of the target.  I don't use strip reading at this point and find a single read of the 21-step test file with my Print Fix Pro spectro is fine most of the time.  

> 
> And for that matter - is 51-steps enough for quality 
> printer profile?

If the curves going into the linearization step were well done, 21 steps is enough. 

I've recently been profiling Arches un-coated watercolor paper where I hold the ink limits of the midtones down to avoid bleed.  This keeps the midtone inks in their relatively more linear areas.  When I set the relative densities based on these midtone reads, the pre-linearized QTR profile is nearly linear already. 

> Bob Atkins for example uses 4096-step target for RGB profile.

RGB is 3 dimensional, whereas the grayscale is a single dimension.  21-steps to the third power may give you a better idea of what would be equal.

I found the QTR 51-step profile target did not produce a better profile.  In fact, it was rather saw-toothed.  I'm not sure where that is coming from.  A modified, vertical target helped, but it was not worth the effort.  

A properly made QTR profile using just 21-step test strips should work very well.

Paul
www.PaulRoark.com