Here is the latest chapter from my continuing digital negative saga. I
apologize for the rambling nature of the text, but this message is
actually a composite of excerpts from two pieces of private
correspondence. In any case, I hope that some will find it useful.
> Thanks again for the prints. I like the Platine paper. I tried it
> about a year ago but couldn't get it to clear properly. I tried EDTA and
> citric acid). What are you using as a clearing agent? How do you think
> it compares to Crane Parchmont and Crane Platinotype?
The Platine is nice paper. I did many weeks of experimentation to work
out my process so that it works the way I want. Here's a rough outline of
my process:
1. Mix sensitizer -- 14 drops Pd solution (hot, heated in a water bath on
a hot plate), 2 drops Pt solution (also hot), 1 drop of H2O2, and finally
14 drops of ferric oxalate solution, room temperature. This will cover an
8x10 using a coating rod (about 8 passes or so). It's important that the
Pt and Pd solutions be hot. This makes sure that there are no crystals in
the solution and makes the drop size smaller. This process prefers
slightly more ferric oxalate than Pt/Pd. Using hot Pd/Pd and room temp
ferric oxalate gives me that.
2. Put paper into a paper safe and let dry naturally overnight. The
Platine is not terribly absorbent, so the overnight drying was necessary
to allow the sensitizer to really work itself around the fibers. This
helped reduce grain and yields a better Dmax.
3. Repeat steps 1&2. For a killer Dmax, double-coating is absolutely
necessary. You'll lose significant paper speed by double-coating, but no
matter.
4. Print and then develop using room temperature potassium oxalate with
sodium dichromate added as a contrast agent (restrainer). Chlorate leads
to insurmountable grain problems.
5. Clear using Kodak Hypo Clearing Agent at normal strength.
Compared with Crane's Parchmont, Crane's Platinotype, Fabriano 90lb HP,
and Fabriano 140lb HP, the Platine has a better Dmax, though the Fabrianos
are really very close. The Platine holds up the best through all the wet
processing. However, the Platine seems slightly grainier than any of the
other papers. The Fabriano papers work well using the exact same
coating/drying/printing procedure and produces less grain, and is a less
expensive paper too. In my one test of the Fabriano it seems that the
140lb. is a bit less grainy than the 90lb. Crane's Parchmont seems to
print the smoothest of all, but it is a very absorbent paper which I find
difficult to coat, especially with a second coat -- you really have to
increase the amount of sensitizer for the second coat because it is just
sucked in to the paper. It is also much more fragile when wet.
Printing a Photoshop step tablet optimized for platinum printing by
Evercolor (i.e. some compensation curve has been applied) I note that the
difference between steps 100, 95, 90, 85, and 80 are greater on my paper
[a 1:7 Pt/Pd print on Arches Platine developed in potassium oxalate #2]
than on yours [a palladium print on Crane's Platinotype developed in
Ammonium Citrate], and even moreso on the Fabriano than on the Platine.
The Fabriano 80% looks like your 65%. Anyway, I'll have more deatiled
information tonight.
I went to the service bureau last night and I have some new data for you.
First of all, you mentioned that the film curve would be dependent upon
the imagesetter. This should not be the case, so long as both
imagesetters are using the same type of linescreen, and they are both
set-up and calibrated so that the output to film is linear. Some people
prefer to spend extra money to buy Kodak's more linear film, others just
build a correction curve into their process. The former is the more
desireable meathod. My friend says he spends about $10,000 a year extra
just to buy the better film. In any case, there is no reason why you
couldn't use the same curve on different imagesetters.
Next, Evercolor is using a 45 degree, elliptical dot linescreen. I
rechecked the densities with what you reported, and they seem correct. I
have derived the Photoshop correction curve from this and it is as follows.
Working with a positive image open in Photoshop, you should apply the
following corrections to duplicate your Evercolor curve.
Input % Output %
(desired %) (set actual film output to this)
___________ ________________________________
100 90
95 70
90 47
85 38
80 31
75 25
70 21
65 18
60 15
55 12
50 10
45 9
40 7
35 5
30 4
25 3
20 2
15 2
10 1
5 1
There are a couple of things worth noting.
1. The curve is pretty radical. This is necessary, and is just the way
that linescreens work, so don't panic.
2. When you take 256 shades of grey and try to express them in 100
percentage points, there are some cases where two adjacent tones will have
to be expressed as the same percentage. In the above chart you will note
that 20% and 15% are both set to 2%, and the same with 10% and 5%. My
recommendation is either set the values of 5% through 50% equal to 1%
through 10% respectively (to ensure seperation of all tones), or to just
set the 0%, 5%, 15%, and 25% points, skipping 10%, and 20%, thereby
letting the Photoshop curve define the intermediate values.
Actually, what I usually do is enter the curve using the Photoshop
"Curves" dialog box and then I smooth out any odd ripples until I get a
nice, continuous, sweeping curve. Make sure that 0% input = 1% output,
and 100% input = 90% output. I am assuming that my paper does not have
any bizzare, narrow-band phenomena, so this procedure makes sense.
3. The 100% step uses a 10% dot, which coincides perfectly with what I
reported before, i.e. I have to make 90% equal true black because all
values above print the same. It seems Evercolor came to the same conclusion.
Printing your tablet (an Evercolor digital step tablet optimized for
palladium prints) on the Platine, and using my #2 developer, I was able
to obtain a result which is not that different from your printed result
on Crane's Platinotype. The Fabriano printed differently. I think I
would have to use the following output percentages with the Fabriano:
Output values for 140lb Fabriano HP
___________________________________
90
75
60
47
43
36
31
25
22
19
16
15
12
10
7
6
4
3
2
1
[note: since writing this I have entered these curves in Photoshop.
There are a couple of small wrinkles which needed to be ironed out. In
the next day or so, I'll let you know what my final values are, though as
I explained, it's still basically the same curve.]
I like the smooth tones of the Fabriano, so I may give it a try.
Finally, we (my friend and I) came to the realization that the desitometer
is fine for reading film, but is not really to be trusted when reading
prints. I think the fiberousness of the paper confuses the densitomer and
throws off the readings. For example, the 50% step from your Evercolor
tablet, printed on the Fabriano, measured 53% dot on the print, however
if you looked at it with your eye you would guess that it was *much*
lighter than middle grey. In fact, the 65% step "looked" to be middle
grey and when measured with a densitometer it came up as 73% dot (on the
print). The Fabriano printed the step tablet very linearly (at least in
the 1% - 50% range), as measured with a densitomer; but if you look at it
you can tell that it is too light.
The practical upshot of this is that I won't be able to come up with the
magic curve just by making measurements, as the measurements of the print
cannot be trusted. Therefore, final tweaking will have to be done by
eye, for every image. I was really hoping to get into a situation where
I knew that what was on the screen would print just the way I wanted it
to print, but I guess that was just wishful thinking (as if I didn't
really know that in the back of my mind anyway!). I was, at the very
least, hoping that I wouldn't have to make multiple negatives -- making a
new one, printing it, deciding what needs to be changed, making another
neg, etc. -- but, it looks like I'll have to do at least some of that
too.
Happy printing,
David Fokos