[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Digital Negatives with enough Tones for Pt/Pd
Digital Negatives with enough Tones for Pt/Pd
It is becoming evident that a single scan will NOT produce enough tones
for a Pt/Pd print. If one recalls earlier, there was some discussion
about printing with three negatives (or masks) or printing multiple
exposures with appropriate negatives. My inclination has moved strongly
toward the use of three negatives made from three scans of the original
negative.
The reason becomes clear when one investigates what happens when a
12-bit scan is used to make an 8-bit image. Simply, no mater how you
slice it, the result is 256 tones.
With any scanner comes another artifact, that there is noise usually not
precisely identified by the manufacture, but definitely within the
density range of a standard 21-step.
The noise level is easy to determine. Scan the 21-step beginning with
the greatest density. Lines and texture from the noise will be seen.
The step at which the noise is not detected is within the noise level
limit and close to the greatest density that can be read without noise.
This is important as it will set a limit to the density range of the
negative. Each scanner model may be different. With the HP
ScanJet-6300C, it seems that a negative should be processed with about
half of the development that would be normally used for Pt/Pd. The
ideal is to get as large a density range in the negative that can be
read without the appearance of noise from the scan. (A later concern
may be how the range of the original negative may effect the outcome in
the print. For example: Is better discrimination of values in the print
related to a larger tonal range in the negative, NOT talking contrast.
Does Pt/PD discriminate tones better (or partly better) than g. silver
because of a negative with a larger range.)
Now one is left with a certain density range in the negative to convert
to densities in the print. If only one scan is used, the information
can only be made to transverse a 256 tone range. The number of unique
highlight tones produced are just not enough. In effect what is seen is
a posterization of tones to the number allotted, (the more number of
tones, the closer toward continuous tone). A slight posterization seems
to provide clarity, cleanness, and crispness, but leads to an
unphotographic look, a loss of subtle texture, and a loss of realism as
the number of tones are decreased. The minimum number of tones becomes
a choice or a restriction that the photographer must work within.
Also, do not be fooled into thinking that the 4095 tones of a 12-bit
scan can later be split into various segments (upper, mid, and lower)
because the three segments will only become something like 0-15, 16-63,
63-255 with a net of 256 tones, no mater how it is sliced. The 256
restriction is from the 8-bit software such as Photoshop that is
currently available. One of the greatest improvements to Photoshop
would be the increase of pixel depth in all functions, and should be a
serious consideration with today's faster computers.
In order to get more tones more scans are required. For example, a full
range image is set to scan with a range of 0-4095. Instead the upper
highlight portion of 3840-4095 is selected as full range and scanned,
then the highlight portion of 3329-3840 is selected as full range and
scanned, then the rest (mid and lower) 0-3328 is selected as full range
and scanned. This should result in 256 upper highlight tones, 256
highlight tones and 256 mid and lower tones. Note that for the upper
highlights no more tones can be generated unless a scanner with greater
than 12-bit depth is used. After some curve adjustment, this may
hypothetically result in a tone distribution as follows.
ZONE TONES with tri-scan TONES with one scan
Low-mid
0 42 24
I 42 24
II 43 24
III 43 24
IV 43 24
V 43 24
highlight
VI 86 24
VII 85 24
VIII 85 24
upper highlight
IX 64 24
X 64 12
XI 64 3
XII 64 1
Note that 13 Zones are used because this can be reasonably expected from
a palladium print.
These values represent the maximum number of tones evenly distributed.
Actually it is expected that the number of tones may be less and the
distribution not even (diminished in the upper values). This is likely
due to the logarithmic relationship causing an even distribution to have
greater discrimination between adjacent lower values and less
discrimination between adjacent upper values. But this is dependant on
how much discrimination the eye has at various levels. It is understood
that the eye has more discrimination at higher values, but not exactly
how much. Some of the upper tones may merge, indicating that there be
some limit to the maximum number of tones that can be discerned for a
given zone.
Next comes the concern of how many of the generated tones can be
produced with certain ink with a certain printer. The maximum amount of
ink density per dot area (dot area being determined by the resolution
used) is set for each negative so as to produce the maximum density of
that negative's range. The lower-mid negative may produce a maximum
density to give Zone V-VI in the print. The lower-mid negative plus the
highlight negative may produce a maximum density to give Zone VIII-IX in
the print. All three negatives produce a maximum density to give pure
white in the print. However it is still unclear how many unique shades
can be produced on each negative. It seems likely that it is not the
entire 256. It should be clear that three negatives from three scans
will assure much better results than a single scan with however many
negatives.
It would be informative if someone experienced with producing great
digital negatives for Pt/Pd printing would try the three scan three
negative approach and compare the resulting Pt/Pd print side by side
with a Pt/Pd print from the same original image form a single scan,
considering carefully any noticeable effects from posterization as
described above. I am working toward this, but do not have as much
experience with digital negatives as some others and would welcome their
efforts.
--
Jeffrey D. Mathias
http://home.att.net/~jeffrey.d.mathias/