QTR-Quadtone RIP

On Mon, Mar 29, 2010 at 9:02 AM, Michael King <drmrking@...> wrote:
>>>Is there any benefit in re-creating custom curves for this new version? I
> have just finished profiling two papers for QTR and wonder if they would be
> a) compatible, b) have greater accuracy with this new version?
>
> As I read in Roy's answer to my earlier question - there has been no change
> to profiling. Its still 8 bit based. BUT he is interpolating between the 8
> bit values to generate intermediate values to support 16bit.

There have always been 256 points in the QTR curves but the values at each
of the points have always been 16-bit.   The 256 points are already way overkill
compared to, for instance, color ICC profiles.  ICC color management typically
uses interpolation with just 25 points so QTR curves have 10 times that.
Interpolation is a perfectly good technique.

>
> To be honest paper coatings are barely consistent enough across the page to
> let you profile at 7 bits (128 values) never mind more than 8 bits. So we
> would be able to gain nothing from Roy directly increasing the curve
> resolution beyond 8 bits.
>
> You might ask what's the benefit of 16 bit printing ? Well I expect the main
> benefit is in helping smooth tonal areas avoid posterization, by creating
> more ink values in the tonal graduation.

If you have posterization issues it's almost certainly due to other issues.
More bits and more overlap are pretty much the very last things to be
concerned about.  Posterization is usually in the image or you have bad inks.

Roy

>
> I've been wrestling with this issue for the last few months and have been
> experimenting with Cone style overlapping curves and QTR less overlapping
> style. There is no doubt more overlap reduces the problem BUT also for some
> papers overlapping curves reduces dmax. So as usual there is no free lunch.
> I am excited about the potential of 16 bit printing to address this issue -
> just got to Hackintosh my PC first :)
>
> Mike
>