RGB Separation Negatives
(I offer a general explanation and you will have to adapt this to the
print process you use).
FILTERS:
There a large number of possible tricolor filter sets for tricolor and
bicolor use. Among these two of the most useful for in-camera tricolor
use have been the 25 red, 58 green and 47B blue set which I will call
25-58-47B hereafter and the 29 red, 61 green and 47B blue set or
29-61-47B. The 25-58-47B one might have advantages for many alternate
photo processes since it will be easier to find new or used, may be less
expensive and each of the three filters needs the same amount of
exposure adjustment +3 stops for daylight. The 29-61-47B is sharper
cutting so there is less color overlap, mainly between red and green.
Both sets have approximately the same overlap between blue and green
curves of around 1%. The 25-58-47B set has about 6% of overlap between
red and green curves and the 29-61-47B set has well less than 1% overlap
making it a slightly better choice if very critical color reproduction
is the goal.
Filters have gotten quite a bit more expensive over the years. It will
be worth checking photo shops and eBay for used ones that are in good
condition as a good deal of money can be saved. Here are some of the
manufacturers whose filters I have experience with:
Hoya - These are dyed glass filters. They vary in color from batch to
batch but seem to last a long time without fading and are thought by
some to have very parallel surfaces so be somewhat better than gels
sandwiched between glass with use of long lenses. Hoya does not
currently make a 47B blue filter but does make red and green ones.
LEE - These are available as 4x4 inch (100mm x 100mm) filters in an
optical grade of plastic. They will scratch easier than glass surfaced
filters. Some say you cannot stack many filters in an optical path
without degrading image quality (example polarizer + red _ neutral
density) but the manufacturer claims that you can - I have no experience
to offer on this. LEE does make a basic holder for 4x4 filters that
holds up to 4 filters and that uses a threaded adapter ring to attach to
your lens. They offer a Pro option for this holder that adds rotation
for uses such as a linear polarizer filter. LEE currently only makes the
25-58-47B tricolor set not the 29-61-47B set.
Tiffen - These are gels sandwiched between thin glass and surrounded
with an aluminum ring. Since the gels are true Kodak Wratten filters
they are quite accurate in color where dyed glass or plastic will
typically vary from batch to batch. I cannot say whether they are more
prone to fad as are gels alone. I have used these a lot and like them.
The company is forthcoming with useful information and has been a
supplier to the motion picture industry for many years. Over time I have
rarely seen filters separate from the glass at the edges. They
apparently make filters as custom operations. I currently have a
29-61-47B Series 9 set on order through BH Photo and it was stated that
the delivery time will be 2 to 4 weeks because this is a special order item.
Harrison & Harrison - They are out of business but you may still find
some of their filter in photo stores or used. They have also been a
supplier to the motion picture industry. I think their filters are
sandwiched glass with gel cemented in between but they may have made
filters in other ways too.
Various camera makers have offered dyed glass filters over the years
such as Asahi that makes Pentax cameras and Takumar lenses. These are
good if found used in good condition.
Kodak Wratten - These are thin gels that have been made in a wide range
of sizes over the years. As with many film related photo products demand
is decreasing and for business reasons Kodak has decreased the breadth
of their filter offerings. They are optically good and probably the most
accurate color available when new. Recently Kodak has begun offering a
new more rigid form of Wratten filter called Wratten 2. These are said
to be optically very good and of accurate color. I have not used them.
Either type can scratch and the gels will finger print or be moisture
damaged. Kodak also makes CP or color printing filters that are meant to
place in a light path but not use optically. You won't want CP filters
for in-camera tricolor.
There are a several other manufacturers of filters that are more recent
entries and that I have not used like B+W. These are typically more
expensive filters. They seem to have a good reputation among view camera
users.
FILTER EXPOSURE ADJUSTMENTS:
This is a hit or miss proposition. Aside from Wratten filters color and
optical density will vary slightly so that testing is always the best
choice. The following data provides a good starting point for testing.
The necessary exposure adjustment will be different for different films.
Older panchromatic films like Plus X and Tri-X will have slightly
greater blue sensitivity and less red sensitivity than TMAX films. Films
like Rollei Infrared have sensitivity generally similar to older
panchromatic films but extended to 820 nanometers in the infrared region.
Here are exposure adjustments from Tiffen for older panchromatic films
(Tri-X):
25 red = +3 stops in daylight (6500 degrees Kelvin) or +2 2/3 stops
in tungsten (3200 degrees Kelvin)
58 green = +3 stops in daylight and +3 in tungsten
47B blue = +3 stops in daylight or +4 stops in tungsten
29 red = +4 1/3 stops in daylight and +2 stops in tungsten (I
question whether a 25 takes more adjustment than a darker 29 in tungsten
light so test this one)
61 green = +3 1/3 stops in daylight and 3 1/3 stops in tungsten
Here are exposure adjustments from Kodak for older panchromatic films
(Tri-X):
25 red = +3 stops in daylight or +2 1/3 stops in tungsten
58 green = +2 2/3 stops in daylight and +2 2/3 stops in tungsten
Here are exposure adjustments for Kodak TMAX films:
25 red = +3 stops in daylight or +2 stops in tungsten
58 green = +2 2/3 in daylight and +2 2/3 in tungsten
Most other manufacturer's films will be similar to Kodak older
panchromatic films (Tri-X) I think. Always adjust shutter speed or use
neutral density to do exposure adjustments. You cannot use f-stop
adjustment because that will change depth of field and show up as color
fringing.
FILTER LIFETIMES:
I have a 61 green gel that dates from the 1970s that only needs +2 stops
of exposure adjustment in daylight so has lost over a stop of density.
Mostly filters change from exposure to light so you can lengthen their
lifetime by keeping them covered when not in use. My experience is that
dyed glass filter do not change as much and probably this would be true
of dyed plastic too but I cannot say for sure. I suspect that UV
degrades color so leaving filters in sunlight is probably worse than
leaving them in tungsten light. The tendency to fade varies by color and
Kodak has rated Wratten filters for this.
FILM DEVELOPMENT:
I have been using the figures Kodak and Rollei publish for roll and
sheet film without adjustment for in-camera tricolor separation
negatives and then tweaking the scanned images in photoshop. Just
adjusting density of an 18% gray card to be the same for all 3 RGB negs
brings the separation negatives into reasonable balance. With an 11 or
21 step gray step wedge you can finely adjust the color balance.
It was sometime customary to develop the red and green sheet film
negatives for normal development time and add 50 to 55 percent more time
to the blue negative development. I do not remember why this was done
but can find it in literature on dye transfers. I just follow the data
sheet recommendation for all three RGB negs now and balance scanned
images in Photoshop.
I use D-76 straight for the developer and hardening fixer with a water
bath in between. With tricolor separation negatives consistency of
processing is critical across the three separation negatives. Try to
keep the temperature consistent and agitate as recommended. Tray
development of sheet film produces very even results with gentle
agitation. Randomness in the agitation helps evenness so avoid lifting
the same edge of the tray every time. With roll film in Nikor type
stainless steel reels even processing gets more difficult. Agitate by
inversion of the tank and shaking every 30 seconds. Cutting a section of
a plastic bottle to act as a spacer and keep the reel from sliding in
the tank seems to help. A water bath at same temperature of 60 seconds
before pouring developer in seems to help. Tapping the tank solidly when
the film first gets wet will dislodge bubbles and help get even results.
Do not agitate too roughly with small tanks as this can make evenness worse.
A JOBO CPP2 processor could probably do very even and consistent film
processing. A dip and dunk line with nitrogen bubbles would be pretty
good. Small sheet film tanks with old film hangers will leave streaks
and not work as well as trays.
In earlier days of dye transfer (before Photoshop existed) it was common
to adjust a set of separation negatives after processing them by
bleaching them back. With Photoshop there is much flexibility to adjust
the scans after processing the negs so this is no longer needed.
FILM CHOICE:
Tri-X offers good exposure latitude so is a good choice to get started
with for sheet film and 120 roll film. I have not tried 35mm but it was
used a lot by students doing dye transfers in classes. If I were to use
35mm I would probably use TMAX 100 for the finer grain and greater
resolution. For doing things like separation negatives of paintings for
Giclee prints I'd use 4x5 or 8x10 TMAX 100 to get fairly grainless negs
and hold brush stroke detail.
The best choice used to be Super-XX but as with so many things it has
long been discontinued. It had a thick emulsion so could have gamma or
contrast adjustment over a wide range by varying development time. The
closest similar choice is probably Tri-X but maybe other manufacturers
make something.
SCANNING:
Unlike slides or transparencies where dynamic range is fairly extreme
separation negatives will have moderate dynamic range. So there is less
need for a drum scanner. I think any Epson or HP scanner should do
fairly a good job. If you use roll film it will be an advantage to have
a scan area that accommodates a strip of 3 RGB negs in one scan. Wet or
oil mounting will hold all three negs flat so aid registration. Mineral
oil used to be used for wet mounting. It was thick and sometimes bubbles
occurred. It was also messy to clean up. It is more common to use Kami
mounting fluid today. Here is an address for obtaining it.
Aztek, Inc.
13765-F Alton Parkway
Irvine, CA 92618
Phone: (949)770-8787
Fax: (949)770-4986
www.aztek.com
Their website has videos on doing wet mounting and Pam Stevens there is
quite helpful. Kami mounting fluid evaporates so requires little special
cleaning. It is also flammable and needs to be used in a ventilated
area. It may cause skin irritation. Get the hazards sheet on it from
Pam. Mineral oil is safer and less noxious but not as easy to use and
lots more work to clean up. Be careful not to get either mounting fluid
into the works of your scanner because it can destroy the scanner. When
you wet mount film use a small piece of Kami safe tape to secure the
film in only one place.
REGISTRATION:
I know Kodak films to have good dimensional stability which helps with
registration. I am not as familiar with other films. Rollei has a good
film base material but is thinner than Kodak so maybe not as good for
dimensional stability. Inconsistent storage can cause one of the sep
negs to differ in size from others. Store them in identical temperature
and humidity conditions. I have a Kodak paper that goes into far more
detail than makes sense to put here. Basically, moderate temperature and
reasonably low humidity are good choices. The basement and attic may not
be good places to store RGB sep negs.
Registering in X and Y is easy in Photoshop. Angular displacement is a
pain to correct. This makes roll film a better way to get started since
all three separation negatives can be in one strip of film at an
inherently same angle.
If shooting in-camera separation negatives ever got popular enough
someone would probably write software to automate registration or Adobe
might include this capability in Photoshop. What I am talking about is
the ability to compensate for size or angular mis-registration. It could
be done but don't hold your breath because the demand will remain
insignificant.
CAMERA CHOICES:
You can find pretty good deals on film cameras on eBay now. This seems
especially true for 120 roll film cameras. There are 6x6 cm (2"x2"), 6x7
cm and 6x9 choices. There are SLRs, twin lens reflex and 120 film backs
for Speed Graphics and view cameras. Any will work with SLRs being the
most convenient choice. A solid tripod is as important as the camera for
in-camera tricolor work. The difficulty with view cameras is that there
is slop in the camera back and between the film and its holder. By
tapping holders to align sheet film at the bottom of the holder and then
pushing the holder to the bottom of the camera back after inserting the
film holder I have gotten reasonably good angular registration. With a
view camera, camera movement is more of an issue. A real solid, heavy
tripod helps. Care in loading film holders helps. There used to be pin
register camera backs for in-camera separation work but finding one
today in usable shape would be quite difficult.
SUBJECT MOVEMENT DURING EXPOSURE:
Things like water and clouds can look interesting if they move during
the exposure. Movement of people or other subjects results in ghost
colored images with background showing through. It looks kind of 1960s
to me but you may like it.
PHOTOSHOP:
What makes in-camera separation negative shooting feasible today in my
opinion is Photoshop and the ease with which separation negatives can be
manipulated and re-registered or channel merged into an RGB image.
There are probably several ways to combine images in Photoshop. I scan
all three or four (+ infrared) sep negs as one pass. I crop out and copy
each sep neg to a layer to balance density and align them. Then I crop
to a common size. I save each later as a new TIFF image eliminating any
extra layers other than the sep neg scan. Then open only the three image
files of RGB images and in the channel menu select merge channels to
create an RGB image. There are several web topics on this.
SHOOTING ONE WAY:
By shooting in-camera RGB or RGBI (+ infrared) you gain near infinite
post production flexibility with one standardized exposure processes. In
the cold of winter this can be an advantage for landscape work. In post
production you can choose to make B&W or color prints. With B&W you can
use Photoshop combination modes and layer transparency to combine the
RGB sep negs and create any color to tone translation you like. It is
easier to make creative judgments back home out of the cold.
PAINTINGS:
If shooting paintings, especially oil paintings with brush stroke
texture dual polarizers can be useful. Use polarizing material over the
copy lights and a polarizing filter on the lens - both linear polarizers
and light polarizers oriented the same way. Adjust the lens polarizing
filter to control the degree of brush stroke reflection.
SEP NEGS FROM COLOR FILM:
You can also enlarge color film to create RGB separation negatives in an
enlarger. An 8x10 or 4x5 film holder can be used as the easel. Tungsten
exposure adjustments will be indicative of the relative exposure times
once you get one color's exposure determined. An incident meter can be
used with filter in place.
I am still experimenting and re learning how to do this. I wish I had
the camera back that the Russian used and could take RGB shots in 3 seconds.
Ed
