QTR-Quadtone RIP

Paul says ...

>>While not used with ABW, my experience with the 51-step test strips was
not
very positive. I'm not sure why, but the results were usually not as smooth
as with the 21-step. I ended up often getting a "see-saw" pattern in the
final. I thought part of the problem may have been that with so many small
steps the effects of rollers and other causes of minor fluctuations had too
much influence.

Yes paper coating variations on a sheet of matte paper in my experience can
start to approach or exceed the 51 step differences.
To use 51 step or greater successfully I have to print several targets at
different orientations (coating changes vary across the paper) and average
them.
I also compare target readings and discard any sets where coating varies too
great.

Of course this means you should really read and average multiple 21 step
targets as well ... its just the deviation isn't so visible.

Mike



On 25 July 2011 00:03, Paul Roark <roark.paul@gmail.com> wrote:

> **
>
>
> Hi Tom,
>
> Is there any advantage using the 51-step wedge over the 21-step when
> > creating profiles for the Epson R2400 ABW mode? I'm using the Spyder3
> > spectro.
> >
>
> While not used with ABW, my experience with the 51-step test strips was not
> very positive. I'm not sure why, but the results were usually not as smooth
> as with the 12-step. I ended up often getting a "see-saw" pattern in the
> final. I thought part of the problem may have been that with so many small
> steps the effects of rollers and other causes of minor fluctuations had too
> much influence. So, I tried vertical 51-step test strips. I ended up
> giving up on these also. The bottom line is that I have found a good
> 21-step in sufficient and the 51-step was not an improvement.
>
> And what about random ...
> >
>
> A good, averaged random might be very good. It would average out some of
> the inevitable fluctuations. Offsetting this is that I think most use strip
> reading for these, and my tests indicate that strip reading is not as
> accurate as individual, manual readings. Additionally, I find the smooth
> ramp on top of the 21-step I use to be very helpful (Jpeg version at
> http://www.paulroark.com/BW-Info/21-Step.jpg). If you see what looks like
> a
> bump or vertical line in that ramp, you've got a problem, and it may be one
> that a 21-step misses.
>
> The main issues I've had with "plain vanilla" 21-step test strips can
> usually be detected. The first is that if there is a sharp turn in a ramp,
> it is often a problem with a cross-over than can't be cured with
> linearization. You probably will not have this with ABW, as it uses Epson
> cross-overs, which are generally outstanding. Second, you're more likely to
> get good deep shadow transitions if you have a compressed shadow region
> going into the linearization. That is, going in the black part of the curve
> ought to look somewhat like -- but milder than -- gray gamma 2.2. This
> gives the program real information about what is going on down there, as
> opposed to interpolated information. In the extreme case, if there is a
> sharp drop between 95% and 100%, you're asking for trouble.
>
> Needless to say, to see these potential trouble areas I strongly recommend
> graphing the pre-linearization results. Linearization can easily handle
> smooth curves, but sharp turns and kinks are trouble. These are easy to see
> when the pre-linearization curve is graphed. The Spyder output text file
> can be opened with Excel for graphing (inserting a line chart).
>
> Paul
> www.PaulRoark.com
>
> [Non-text portions of this message have been removed]
>
>  
>


[Non-text portions of this message have been removed]