Digital BW, The Print

I've occasionally made references to some tricks that might be able to
expand the digital dynamic range.  I think it is time that this group focus
on one of these in particular.

 

A small group of off-list B&W shooters has been pursuing an idea that was
originally proposed by Charlie Bustamante and Lloyd Baggs.  Tom Baker, Lloyd
and I have been trying to follow up on Charlie's idea with both promising
and mixed results.  Based on some outstanding prints shown to me that Llyod
made, and the elegance of the concept, I think there is potential, and I
think this group has the expertise to push the idea - forward or into the
dumpster if that is where it belongs.

 

The basic idea is very similar to what we did to solve the problems of the
early 4-ink printers.  Rather than have 3 color inks fire at once, we
partitioned the inks.

 

So, what I've called the "digital zone system" (admittedly in part just to
get peoples' attention) is the use of filters to partition the scene's
dynamic range among the color sensors.  The interaction of filtration over
the lens and the existing filtration over the individual RGB sensors results
in differential neutral densities that spreads the sensor responses.

 

Basically, the further one wants to push down the blue sensitivity to
capture highlights, the stronger the yellow filtration needs to be.  The
more one want to raise the red sensor sensitivity, the more magenta one adds
to hold the green back.

 

Just using the filters I have on hand, I think I've increased the dynamic
range of the Canon by 2 stops.  I have posted a quick and dirty sample shot
comparison at http://home1.gte.net/res09aij/8-30-Sample.jpg .

 

As I also explain at the bottom of my Index at
http://home1.gte.net/res09aij/index.htm, these shots were taken with the
Digital Rebel on auto exposure, and they were processed in CS2 at the
default settings.  More could probably have been seen if the processing had
been tweaked.

 

Needless to say, there is a lot of refining, testing, and calibration to be
done.  It may be that there is nothing here, but Charlie and Lloyd's basic
idea of filtration as a way to get more dynamic range seems to have some
merit.

 

The threads and issues we've been discussion on the forum relate to this and
have already been very helpful in understanding how this might work and the
limits of it.  

 

As I've found in other contexts, the green pixels seem to be the most
important, probably due to their larger total area and capacity.  The R & B
are the secondary ones that might pick up the more marginal parts of the
image.  When the big green "buckets" overflow, the smaller ones seem to be
more impacted than when one of the small ones overflows.  So, keeping the
green range from clipping too badly seems like a good idea.  I've ordered an
85B (Heliopan KR15) that may be the first level of partitioning.

 

The Kodak book/pamphlet on filters is most helpful.

 

To avoid overkill in this post, I'll stop for now.

 

Have at it.

 

Paul

www.PaulRoark.com <http://www.paulroark.com/>  

 



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