First of all, for technical sake, let us stop the incorrect use of the terms
like enhanced highlight separation (bad), improved highlight separation
(worse), or increased highlight separation (worst, wrong). What we are seeing
here is the exact opposite: highlight compression.
Improved highlight separation means that each highlight zone will be nicely
recorded in the negative in a proportional (preferably linear) way. If you
imaging the straight line curve of TMax and imagine that the line can continue
on forever to the density of 5.0, 6.0,... that will be beautiful. That is
called beautiful highlight separation because you truly separate each
highlight into the proper zone.
The problem is that when you have that kind of truly improved highlight
separation, you cannot print it because your paper only can print a certain
negative range. The result is you will have textureless whites for the
highlights. In order to show some separation in the highlight, you need to
*COMPRESS* the highlights into printable zones. Just imagine you can pin down
the curve on the shadow and midtone, and then you bend the highlight portion.
You squeeze it down. You *compress* the highlights.
David Hutching understands this concept completely when he says that even 6 or
7 stops above zone IX are printable (you see, 6 or 7 stops more of the
highlights are compressed into the printable zone). However, he mistakenly use
the term "improved highlight separation", and unfortunately that term has been
copied again and again.
The reason I want to bring this out is when you want to further your research
or reading in the science of photography, you might get completely confused if
you don't get this term right.
How does one achieve this kind of highlight compression? By compensating
development. The basic idea is to use a weak developer. The highlights (densed
area) is depleted of developer sooner than the shadows (the film is willing
but the developer is weak). The effect is the highlight region gets compressed
(curve gets flatted in the highlight regions).
There are several ways of achieving compensating development:
1. by great dilution and low agitation (too much agitation and you continually
supply fresh developer to every region, so compensation cannot take place)
2. by divided development (the film soak up the developer, but there is only
so much it can soak up, so it is not quite enough for the highlight, but
enough for the shadows, thus highlights get compressed).
3. by waterbath development (similar principle to #2)
4. by special compensating developers (Anchell has 3 in his book, guess what?
all of them are staining developers). I am amazed that we have discussed the
magics for so long and nobody recognized the effect.
If one examines the curve of a negative developed in comp. developer, he will
see that the curve has a long toe. I have said this in the Tri-X / TMax
thread, but again, yes, the characteristic curve of a film can be altered. It
is SO EASY and has been done SO COMMONLY. LOTS OF PHOTOGRAPHERS ARE DOING
THAT. Just read up on some good photography books or even search the net. Some
photographers, however, do not recognize the change of curve themselves
although they like the effect and use it.
[By the way, there is a person by the name of John Douglas with his studio
called spectrum photography. I do not know him personally, but he is an active
user of compensating development. He sometimes use a highly diluted developer
and soak the film for overnight! I don't have his web site handy, but do a
search (it is a great site with developer formula and processing tips). You
will see all the effects that we have discussed in the pyro thread (highlight
separation, edge effects, *all*) mentioned in his site.