Why should color print film be so hard to print?
Dave Bernard
When processing transparencies, there is a strict set of
controls- time, temp, dilutions- and what you get is what
you get. The exposure + the processing = the image. A lot
of photographers like this, since they know what to expect
when they take the picture, without lab folks, or computers,
to interpret what the picture should look like. So a lot
of people like transparency film for this reason.
Color print film has a lot going for it- latitude, easy to find &
get processed cheaply, etc. My question is, why can't color print
film be processed and printed in the same way that transparency film
is? ie, with strict adherence to time, temp, etc., all according to
manufacturing specs, with the actual printing being done also according
to manufacturer specs? No trying to "correct" a backlit scene, or alter
colors, or anything? In other words, no lab or tech to impose their own
interpretation of the scene as the photographer imagined it. If this
were possible, it seems color print film would be even more attractive?
Dave
donl mathis
In article <1kogkq...@cronkite.Central.Sun.COM>,
dber...@clesun.Central.Sun.COM (Dave Bernard) writes:
Color print film is processed to more or less the same standards as
transparency film, assuming the lab cares about quality standards.
Color print paper is also processed to strict standards, again assuming
that the lab cares about doing it right. Both can (and should) be
verified with control strips -- standard reference samples of the film
or paper, exposed under very tightly controlled conditions, and kept
refrigerated from that moment until the lab worker is ready to take it
out of the box and run it through the machine. The densities of
different colors are measured and plotted, and reveal whether or not
the process is doing the right thing.
The remaining question is still "Why does it seem to be so difficult to
just get it right and stop fussing with it?" The answer is a
combination of several things.
First, when you start looking at color reproduction as a whole, you
begin to see the complexity and compromises that are taking place. It
is not as easy as it may seem to get three emulsion layers to respond
to three bands of light. Their natural tendency is to see blue, and
it's tricky business to get them to see red or green while ignoring
blue, in a smooth and predictable fashion. The bottom line is that it
more or less works, but not with anything close to what you might call
"precision". The layers each record colors they shouldn't, and even
within their own band, say blue for example, the definition of blue as
a spectral energy distribution is not really great.
So now we more or less get the three colors recorded on film, and then
the chemistry converts these records into dye layers. Note that at
this stage, the layer that saw blue light is going to be converted to
yellow dye. The yellow dye, in theory, is the exact opposite of the
blue that the emulsion recorded, but in practice, it's not so precise.
The responses of the emulsion and the dyes are only loosely related, as
they operate within a third of the spectrum in some approximate way.
The same goes for green/magenta, and red/cyan.
Note also that the dyes are not really constrained to their own
layers. The magenta dye is not perfect by any stretch of the
imagination, and has a bit of yellow mixed into it. Those are the
breaks. It's not really possible to get perfect dyes to begin with,
but even if it were, the problem is that they have to be made out of
half of a molecule that's floating around in the emulsion, and the
other half of a molecule that is the garbage byproduct of the silver
development process. This severely restricts the available dyes, and
it's amazing that they are as good as they are.
So now we have a continuous spectrum that has been inaccurately broken
up into three sections, inaccurately recorded in the emulsion, and
inaccurately converted to innacurate dyes. So far, so good.
In color negatives, or theoretically, in any film that we're only going
to print, we can tinker a bit with the original. So in color
negatives, we find a way to compensate for the excess yellow and
magenta dyes in the layers they don't belong in, and that's called
masking, and it's the reason negatives have the orange base. It helps
a bit. You can mask transparencies, too, but they don't look right
visually. They will print more accurately if they are masked, because
transparency dyes have the same problems as all the other dyes, and
they appreciate the extra help.
All of this, with the exception of masking, also applies to the paper.
The primary difference is that we're theoretically starting with three
confined bands of color instead of a continuous spectrum, and that
relaxes the situation a bit. It's important to note that the
emulsion's spectral response is not the same as the film's. It doesn't
necessarily even match the dye all that well, so that the definition of
"blue" presented by the light coming through the negative is not the
same as the definition of "blue" that the paper emulsion layer is made
with. Also note that the dye in the paper has a different spectral
response than the emulsion layer, so even those don't match up.
If you're counting, you'll not that we now have four completely
independent definitions of the three bands that the original,
continuous spectrum was broken up into. The fact that they even come
close to giving the right color is a miracle, in my estimation.
So, fine. Transparencies have all of these problems, too, but they
always look fine to me! Well, transparencies look fine, but not
because they are accurate. Transparencies look good because they have
a wide dynamic range. The dye saturation is not fundamentally
different than negatives or prints, but the exposure scale and contrast
curves are such that most colors are artificially enhanced. The
dark/light ratio in a transparency is on the order of 1000:1, and for
prints the ratio is more like 200:1, so a sense of brilliance and depth
is lost.
When I design my negatives, though, I have the freedom to depart from a
representation that will look good visually, and concentrate on one
that will (at least potentially) print well. So the sensitivity curves
may be a little different, and the dyes may be a little different. I
can add masking layers, and tinker with them as appropriate to match
the dyes. I can tinker with the basic structure of the film and add a
fourth layer if I want to record the spectrum in more than the three
usual bands, and make things a bit more accurate (i.e. Reala). And on
it goes.
For many of these changes, the appearance of the film visually gets
worse, or at least different, and the accuracy of the recorded
information is improved. Since I'm printing it, and have no intention
of looking at it visually, that's not a problem. It would certainly be
a problem in a transparency intended for direct viewing or projection.
The bottom line (so far) is that color negatives record a more accurate
rendering, but there is more variation between different kinds of
film.
On top of those variations, we give the user the freedom to cover a
wider range of exposures, because we can compensate to some degree when
the film is printed. Most people take advantage of this freedom,
knowingly or unknowingly, and that's why people who aren't going to be
able to pay as much attention to precise exposure control should use
negative film.
We also offer the freedom to expose under light with different color
balances, because we have a chance to change the color balance when the
print is made. People expose indoors and outdoors, where one is orange
one is blue, and late, blue shadows, and early morning orange light.
The effect on film is often exaggerated, because of the spectral
sensitivities and the way the spectrum is broken into three parts and
record, so it's not necessarily desireable to leave these variations
uncompensated.
Another simple fact of life is that film's color balance changes as it
ages, and is exposed to heat and other influences. Two identical rolls
of film will produce different results if one of them sits in the trunk
of the car for six months, even if all else is equal, including
strictly controlled processing and handling.
So now we have variations in the "look" of the negatives -- the dyes,
the densities, the masking layers, and so on -- and the exposure and
color variations, and aging variations, and who knows what else. When
it's time to print, these are all (theoretically) compensated for. As
it happens, in real life, you really have to be prepared to take each
frame of each roll of film as a separate entity, and make the
appropriate adjustments to make it look good. If you do a good job
with those adjustments, and use good chemistry and paper, and keep
strict controls, you can produce color that is technically superior to
that produced by transparencies.
A very real problem with this is that it gets expensive to do that. So
a few things start to slip. We invent a machine that can examine the
overall color distribution of the negative, and in a sense teach it how
the paper responds to color distributions, and let it set the color
balance. We let the chemistry slip a bit this way and that in color
balance, contrast, and cleanliness, because a batch of chemistry is
very expensive, and our customers don't want their film processing
costs to double. We use a cheaper paper, which might not have as a
high a quality in the dyes as we would like, but again, it saves us and
(our customers) money. And so on.
Each frame comes along in the machine, and the exposure fluctuations
are random. This one is dense, the next is terribly underexposed.
This one was done in bright, white sunlight, and this one was done
indoors with blue-deficient tungsten lighting. This first exposure was
made six months ago, and this last one was made yesterday.
When I tune the various parameters of the new negative film I'm
designing, I tune them to help get the best results under each of these
variations. When I tune the new transparency film I'm designing, I try
to make sure it will give consistent, more or less realistic color
under all conditions. Accuracy of color is less important than
consistency, and the slides are going to be treated to a "go, no-go"
decision, whereas the job with negatives is to make a reasonable print
from each one, no matter what.
- donl mathis at Silicon Graphics Computer Systems, Mountain View, CA
do...@sgi.com
% make CFLAGS=-Dnotdef all
Alan Sanders
| printed in the same way that transparency film is? ie, with
| strict adherence to time, temp, etc.
-Dave
For one thing, transparencies are not printed. They're only
developed; and color negatives are usually properly developed
also, even by high-volume snapshot operations. But printing
is another matter entirely. The average snapshooter wants
an average 4x5 print, and that's what the neighborhood photo
finisher gives him.
For someone interested in *photography*, a better alternative
is to ask a pro lab to develop and print a color contact sheet.
My local lab charges $8 for this service. I know my negs have
been properly developed, I save money and I see my actual
exposures instead of machine-averaged snapshots. I also have
a one-page record of each foll of film instead of a fat
packet of crummy prints. It works for me.
-Alan
G. Hugh Song
|>
|> film be processed and printed in the same way that transparency film
|> is? ie, with strict adherence to time, temp, etc., all according to
|> manufacturing specs, with the actual printing being done also according
|> to manufacturer specs? No trying to "correct" a backlit scene, or alter
|> colors, or anything?
I know that you are not asking a question. You are complaining!
|> In other words, no lab or tech to impose their own
|> interpretation of the scene as the photographer imagined it. If this
|> were possible, it seems color print film would be even more attractive?
|>
|> Dave
I always put a special instruction that says "Do not attempt to color
correct. Use the lab standard color balance for a specific film I use"
However, I found that several Kodalux labs do have their own standard setups
for Kodak GA100, and that they do not change no matter what film I send.
Fuji Anaheim Color lab always ignores my instruction and changes color
balance even in the middle of a roll. Arrrrg... Their handling is better than
Kodalux though.
I tried several Kodaluxes and found that
1. The enlarger of Fair Lawn, NJ is not focussed at all for a year.
So I decided to switch.
2. Pensylvania Kodalux is always dark. Also the enlarger is not well
focussed.
3. Iowa Kodalux is very well focussed. Very sharp. Alas! they make
a lot of scratch in the negatives.
I will continue to try other Kodaluxes. You know, Kodalux charges more
than K-marts. BTW, I also tried Mystic, Clark, Flair, ... etc.
The conclusion from my experience is that I need my own enlarger, but
I can't afford a space for it.
Hugh
Steve Holzworth
>| My question is, why can't color print film be processed and
>| printed in the same way that transparency film is? ie, with
>| strict adherence to time, temp, etc.
> -Dave
stuff deleted
Color FILM *IS* processed with a strict adherence to time/temp./etc.
Color PRINTS can vary dramatically for the same negative, depending on
who does the printing, and how well they print.
>For someone interested in *photography*, a better alternative
>is to ask a pro lab to develop and print a color contact sheet.
>My local lab charges $8 for this service. I know my negs have
>been properly developed, I save money and I see my actual
>exposures instead of machine-averaged snapshots. I also have
>a one-page record of each foll of film instead of a fat
>packet of crummy prints. It works for me.
> -Alan
Actually, what you get is a sheet of negatives printed at the same exposure &
color balance, which is an *average* color balance, possibly tweaked
for your subject matter. If you're really lucky, they'll balance against
a gray card shot in the roll (you do shoot a gray card?). This is still
better than the balance changing for each shot, which your local quickprint
place MIGHT do. The "solution" to this problem is to shoot a gray card
every time you substantially change the lighting (intensity, type, number,
direction,...), then use a densitometer to balance the card to gray (a
reflection densitometer will compensate for paper/developing factors
also).
Unless you're shooting for reproduction, just balance until you get a
print you like. You don't HAVE to faithfully reproduce the exact scene you
shot (that's why it's an art :-) ). Save a copy of this "reference" print
to balance against in the future, if you print this negative again (assuming
you want the same balance).
If you ARE doing this by hand (without benefit of analyzers, et al). Expose
and print your own personal reference negative at the beginning of each
session. Balance this print to equal your "reference" print of same, then
consistently use THIS SESSION's balance to print your contact sheets. Now
you have a consistent base to work from, even if you want to dramatically
change the balance for a particular print w.r.t. the standard. You will
develop a feel for what a +20CC, etc. change will do relative to your
standard print, making it easier to compensate for any weirdness in the
negative's exposure or lighting.
You can get a Macbeth Color Checker card to use as a reference. This is
a standard color reference used in the printing and video industries.
It has known densities for a gray scale ramp, and assorted color patches.
Take a picture of the card under your typical lighting conditions, using
your normal film and exposure. Balance the print to match the card as
closely as possible (you probably can't get EVERYTHING to match, due to
response characteristics of your film and paper). Different types of film
WILL require different color compensations (Ektar .vs. Gold, for example).
You want a reference negative/print that includes colors across the spectrum.
It's amazing how a good balance for a skin tone, shot against a solid, warm
backdrop, can look hideous when trying to balance something with green in it
(like a lot of plants).
--
Steve Holzworth
s...@unx.sas.com "Do not attribute to poor spelling
x6872 That which is actually poor typing..."
SAS Institute - Open Systems R & D - me
Cary, N.C.
Kevin Davis
Hugh Song writes:
|
|Dave Bernard writes:
|> In other words, no lab or tech to impose their own
|> interpretation of the scene as the photographer imagined it. If this
|> were possible, it seems color print film would be even more attractive?
|>
|> Dave
|
|I always put a special instruction that says "Do not attempt to color
|correct. Use the lab standard color balance for a specific film I use"
|
|However, I found that several Kodalux labs do have their own standard setups
|for Kodak GA100, and that they do not change no matter what film I send.
I have heard a professional photographer refer to something called
a "shirley" (spelling? surely?). Reportedly, you can ask the lab to
print according to the shirley (?), which is a standard possibly
provided by the manufacturer (of the film? of the equipment?).
This is the same as what Hugh is suggesting.
With regard to the rest of the replies to Dave's original question
(or "complaint" as someone else called it). What I read is a lot
of defenses why printing is complicated. IMHO, I do not think
that answers Dave's question. Assuming the negative is new
(very little aging), and assuming the negative was processed with
rigid controls (as suggested by follow-up posts), then consistent
good printing should be achievable without paying a premium to a
custom lab.
There is one solution no one seems to mention very much and I am
seriously considering myself: simultaneously shoot both print and
slide films. Use two camera bodies or interchangable backs (med.
format cameras). Yes, clearly this is not practical for many
applications of photography, but for some applications (landscape)
it is very feasible. The transparency being valuable in itself
would also serve as a guide for the printer of the negative film.
This suggestion is mentioned only in acknowledgment of what has
been said before: negative film and slide film are different
animals; if you want prints, shoot negatives; if you want slides
(or want to publish) shoot slides.
'''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
ke...@pictel.com Only the love can make you a player
Kevin Davis Have you got the love?
PictureTel Corp
steve hix
>| My question is, why can't color print film be processed and
>| printed in the same way that transparency film is? ie, with
>| strict adherence to time, temp, etc.
It usually is...
>For one thing, transparencies are not printed.
This is going to come as a rude shock to the makers of Cibachrome
and various Type R papers.
While *most* transparencies aren't printed, a fairly large number
are. As in many (if not most) magazine illustrations.
>They're only
>developed; and color negatives are usually properly developed
>also, even by high-volume snapshot operations. But printing
>is another matter entirely. The average snapshooter wants
>an average 4x5 print, and that's what the neighborhood photo
>finisher gives him.
>
>For someone interested in *photography*, a better alternative
>is to ask a pro lab to develop and print a color contact sheet.
It's probably cheaper, unless you're doing your own color printing,
to just stick with transparencies.
Except for professional (read: expensive) labs, anything really
interesting is going to be screwed up by the printer.
At least with slides, you have something to point to as a
reference for the finished print.
>My local lab charges $8 for this service. I know my negs have
>been properly developed, I save money and I see my actual
>exposures instead of machine-averaged snapshots. I also have
>a one-page record of each foll of film instead of a fat
>packet of crummy prints. It works for me.
Next best thing to rigorously sorting through your slides...
assuming you have no darkroom to work magic in.
...?...oh, all right: "in which to work magic."
Satisfied?
--
-------------------------------------------------------
| Some things are too important not to give away |
| to everybody else and have none left for yourself. |
|------------------------ Dieter the car salesman-----|
Ron Speirs
>asan...@adobe.com (Alan Sanders) writes:
>
>>| My question is, why can't color print film be processed and
>>| printed in the same way that transparency film is? ie, with
>>| strict adherence to time, temp, etc.
>
>
>Color FILM *IS* processed with a strict adherence to time/temp./etc.
I must take exception to this. For a recent activity in the Photography
Club where I work, we took 8 identically-exposed rolls of Kodak Gold 100
and farmed them out to 8 different local photofinishers. As you might
guess, the prints were all over the place color-wise, but that's another
story. I placed the first 5 frames of each roll (which were bracketed
exposures of a gray step card) in a neg sheet and made a contact proof
sheet. The differences were SHOCKING. I can't tell which photofinishers
were out of tolerance, but I sure could tell they were different from
each other, at least 1.5 stops and 20 units of color. If this is
typical, it totally blows away any hopes of accurate film/processing/
prints where you could depend on the color.
I printed the corresponding frame from each of rolls myself, balancing
the print with an analyzer. They all printed successfully to neutral
gray, but each required a DIFFERENT color and exposure setting.
>Color PRINTS can vary dramatically for the same negative, depending on
>who does the printing, and how well they print.
Amen to everything else Steve said in his posting; lots of good info.
Ron Speirs, Evans & Sutherland Computer Corp., Salt Lake City, Utah
Ludwig van Halen Rocks RIoT
>
Dave, I am glad you asked, because a few people have mailed me with very
similar questions, so I will take this opportunity to explain this.
Color print film, process C-41, is a carefully maintained process with very
tight parameters for time, temperature, specific gravity, agitation, and
replenishment. It is just a hard to keep a C-41 process in line as it is to
keep an E-6 process right on. The only difference is that the E-6 yields a
finished product while there will be another step in the color negative film.
That said, lets talk a bit about color printing. When you expose a color
slide, you are doing so exactly as you want it to appear and you get back the
actual film as your final product. If you get that slide printed, the printer
knows exactly what you want, since your slide is something to match.
Taking a look at the film, you will see that the film can either be very clear
(a density of nearly 0.0) or very dark (a density of around 3.2 or so). Now if
you were to examine a color print, you will see that the lightest density will
be close to 0.0 as well, but the darkest the print will get (called D-Max) will
only be about 2.1. This holds true for both negative and reversal (chrome,
slide) films. So, now you have to make a decision and a compromise. You have
to select which 2.0 of the 3.2 density range on the negative you wish to print.
Certianly, you could set up a standard by which everyone exposes their film the
same way, and then print it the same way. I am afraid, however, that if you do
this, there will be a whole lot of people out there complaining about how
bright their flash shots look and how dim thier church photos look.
So, I guess that the long and short of it is yes, you could make a standard to
have everyone expose film the same way, but noone wants it. You really do want
the flexability that negative film provides.
Please post requests for clarification here to benefit the whole of netland.
Chris Kohanek
RIT Photo Systems Management
Eastman Kodak Intern from Hell
p.s. Would anyone like to hire an Intern from Hell? Mail me.
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
+Christopher J Kohanek \\ CJK...@RITVAX.isc.rit.edu +
+Photo Systems Management \\ 164 R. Quigley Dr; Scottsville NY 14546 +
+Rochester Institute of Technology\\% Eastman Kodak, Rochester, NY 14650-1202 +
+ Disclaimer: If my employers have opinions, it's news to me +
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Steve Gombosi
Attractive to *whom*? To serious photographers, maybe - although I would
think this would require tighter manufacturing tolerances in the
production of negative films. However, "serious" photographers are
a pretty small share of the market (if we weren't, Kodachrome 25 would
be available in 120 ;-)). The "average" photographer is someone with
a point-and-shoot snapping pictures of Aunt Edna at the family reunion,
sitting in the shade, backlit by the sun. Try explaining to *him* that
Edna's going to look a bit cyanotic because "open shade" is bluer than
"standard daylight". If you *really* want your negatives printed uncorrected,
I'm sure you can get a custom lab to do it. Your local one-hour *won't*,
because your request is atypical.
Steve
William Tyler
>I have heard a professional photographer refer to something called
>a "shirley" (spelling? surely?). Reportedly, you can ask the lab to
>print according to the shirley (?), which is a standard possibly
>provided by the manufacturer (of the film? of the equipment?).
Shirley is a standard negative that can be and is used by labs to
calibrate their equipment/processing. Unless you are also shooting
standard negatives, your mileage will vary, and the lab will need to
make adjustments. You don't want to do the things you would have to do
to get completely standard negatives (trust me).
Bill
--
Bill Tyler wty...@adobe.com
Steve Wall
(Kevin Davis) wrote:
>
>
> With regard to the rest of the replies to Dave's original question
> (or "complaint" as someone else called it). What I read is a lot
> of defenses why printing is complicated. IMHO, I do not think
> that answers Dave's question. Assuming the negative is new
> (very little aging), and assuming the negative was processed with
> rigid controls (as suggested by follow-up posts), then consistent
> good printing should be achievable without paying a premium to a
> custom lab.
>
consistent prints. Assume the photographer exposes his negs as carefully
as a slide shooter does. Assume the photographer uses color filters
on his lens to correct for light sources other than 5600K daylight.
Assume the film has been handled like fresh pro slide film all the way
from the factory. Assume all shots were made at about the same time.
The sad truth is that commercial photofinishers can't make these
assumptions. They see a lot more rolls of snapshots taken with point
& shoot cameras than they ever see from pros. The classic roll
of film seen in their labs has a Christmas tree at the beginning,
beach scenes in the middle, and another Christmas tree near the end.
Exposure and color balance vary wildly within a single roll. Thus the
popularity of the automated processor that examines the overall color
& exposure level for each print and adjusts to a preconceived average.
These processors have something very similar to the matrix metering in
the latest Nikons, and adjust for exposure pretty well. Color is harder
to get right, and involves so many variables it's hard to explain them
all, let alone correct for them at mass production rates. The best
machines use a video reversal as a preview, and someone with a trained eye
(not the teenager at one-hour-photo) to make corrections.
Steve Holzworth
>In article <sch.728774090@gargoyle> s...@unx.sas.com (Steve Holzworth) writes:
>>asan...@adobe.com (Alan Sanders) writes:
>>
>>>| My question is, why can't color print film be processed and
>>>| printed in the same way that transparency film is? ie, with
>>>| strict adherence to time, temp, etc.
>>
>>stuff deleted
>>
>>Color FILM *IS* processed with a strict adherence to time/temp./etc.
>I must take exception to this. For a recent activity in the Photography
O.K., Theoretically, film is .... :-)
Counter-example deleted.
>typical, it totally blows away any hopes of accurate film/processing/
>prints where you could depend on the color.
Let's try this again :-)
There is no INHERENT reason that film processing should yield different results
for standard C41 processing. The realities are that lab quality-control
will vary all over the place. The trick is to find a good lab that runs
quality checks often, uses new chemistry when the tolerances shift noticably,
and takes care in the handling of your film. Another thing to consider is
how consistent is the SAME lab day-to-day. If they consistently end up
out 10CC relative to some other lab, they're both probably self-consistent,
but may be running hotter/cooler processes relative to each other, or running
different manufacturers' equipment, which may have different ideas of what
the "correct" process is. Find a lab (or two) you can trust, and stick with
them. At least then, your negatives will be consistently processed.
>I printed the corresponding frame from each of rolls myself, balancing
>the print with an analyzer. They all printed successfully to neutral
>gray, but each required a DIFFERENT color and exposure setting.
This is the important part. If the negative differences are minor enough
that they can be compensated for at the print stage, it doesn't matter a
whole lot (ignoring gross differences in density, etc.)
>Ron Speirs, Evans & Sutherland Computer Corp., Salt Lake City, Utah
--
Dave Bernard
I realize all this. I understand that there are compromises made at the local
hypo-in-the-box. My question was, is it possible to get around those
compromises, and use them to print X-color film so that I can pre-visualize
color, contrast, exposure, etc., the way I can pre-visualize results with
X-chrome film? In practice it would seem easy enough, as some posters have
suggested- just tell the "lab" to process the film without trying any color
or exposure compensation.
Dave
donl mathis
In article <1kuiuo...@cronkite.Central.Sun.COM>,
dber...@clesun.Central.Sun.COM (Dave Bernard) writes:
|> I realize all this. I understand that there are compromises made at
the local
|> hypo-in-the-box. My question was, is it possible to get around those
|> compromises, and use them to print X-color film so that I can pre-visualize
|> color, contrast, exposure, etc., the way I can pre-visualize results with
|> X-chrome film? In practice it would seem easy enough, as some posters have
|> suggested- just tell the "lab" to process the film without trying any color
|> or exposure compensation.
I am quite certain that everyone involved in the discussion is
intelligent, but it still seems to be taking forever to get this thing
resolved! I suspect it's because there are some basic premises,
impressions, and foundational concepts that need to be resolved first.
The question was "Why should color print film be so hard to print?" A
lot of discussion has taken place to actually answer that question,
which I have taken to be a genuine question, and not just rhetorical.
It *is* hard to print color. It is harder to print color from
transparencies, due to their particular characteristics, but printing
negatives is not so easy, either.
There is no real reason that you can't achieve the same level of
consistency in negatives as in transparencies. It is distinctly
possible that transparencies are nowhere near as consistent as they
seem. Once objective measurements are made of different films
processed in different labs and so forth, the overall level of
consistency will be less than any of us would like, in both negatives
and positives.
That's not where it stops, though, for negatives. They have to be
printed.
Not knowing what your background is makes it hard to come up with an
appropriate formulated answer. All I can do is make some assumptions
based on how you worded your question. It's that word "compensation"
that I've picked up on.
Pop quiz: When you're out making photographs, and the exposure is
determined to be f/16 at 1/60, what is the exposure compensation?
Answer: There is no exposure compensation -- just the correct exposure
for this particular subject, and even then, the definition of "correct"
is subjective.
The same is true in the real world of printing color negatives. It's
not a matter of the lab "compensating" for the color in some way, it's
just a matter of trying to find the right exposure and color balance.
(You can think of color balancing as simply trying to find *three*
exposures instead of one. That's all it really is.) Every subject,
i.e. every frame of film that rolls by, is different. Some are
lighter, some are darker, some were exposed in bluish light, and some
in yellowish light. There is no "correct" color balance or exposure.
If you carry your gray card around with you to various places at
various times and photograph it, all on the same roll of film,
processed in absolutely fresh and proper chemistry with the utmost in
strictest controls, when it is time to print it, you will find that it
requires different exposure and filtration on each frame to produce the
same gray. If you are extremely careful with your original exposures,
you will find that the exposure variations are lessened, but still not
totally eliminated.
Think of the film as having three speeds, not just one. Hopefully, for
your ISO 100 film, all three of the layers (red, green, blue) will have
the same speed. In practice, they're close, but not identical. The
color responsiveness, i.e. the balance of these three layers, is
determined by the particular characteristics of the emulsion batch that
the film came from (and they're all just a bit different), the type of
film, its handling, its age, and so on. Color balance is also affected
by flare in the camera -- try photographing a gray card on both green
and magenta backgrounds, and you'll no doubt find that each requires a
different filtration to get gray.
Now, when you make the exposure in the camera, you give the same
exposure to all three layers. So the red layer, at the equivalent of
ISO 97, is pretty happy. The green feels a bit deprived, because it's
speed is ISO 86, and the blue is overexposed (as usual) because it's
speed is ISO 123.
In your next batch of film, these numbers will be different. In fact,
if you go shoot this same roll of film next month, these numbers will
be different. The important thing in this case is that they each drop
by about the same amount, so they stay consistent with respect to each
other.
How do you pick the exposure for this film? Easy -- just average them
all out to about 100, and hope for the best.
The printing paper is similar -- three layers, each with a slightly
unique speed. In addition, there are spectral response considerations,
having to do not only with the speed of the blue layer, for example,
but also the actual *definition* of blue -- the balance of energy at
various wavelengths. The dyes in the film determine this balance, and
each kind of film has its own peculiar dyes.
How do you set the exposure for the paper under these circumstances?
What's the "correct" exposure? Answer -- there is none. It will be
different for each kind of dye that comes rolling by, presenting its
unique balance and spectral characteristics, and the film color balance
will be off a bit because the film's layers have slightly different
speeds, and on it goes.
The best you can do when is attempt to learn the characteristics of the
paper, teach them to a machine, and then let the machine pretend to be
the paper, to give you a chance to see how it might come out. If it
doesn't look right, you fiddle with the balance of red, green, and blue
light until it looks better. Then you make a print. And it might be
close to what you were after. In real life, what you do then is make
another print.
These are not just compromises made at your local hypo-in-the-box;
they're the normal (and only) way to print color.
Now, since the lab knows that it is going to have determine color
balance for each and every frame anyway, it knows that if it lets the
C-41 process go just a tad off into the yellow, for example, nothing
bad will happen, because the printer will use a bit less yellow, and
everything will come out fine (according to the analyzer). If it's
*cheaper* to let the process to a tad off into the yellow, it is
very likely that it will happen. To prevent it from doing so would
mean freshening the chemistry, which is really rather expensive. That
would mean that they would have to raise their prices. They know
nobody wants to pay higher prices, so they let the process go a bit
yellow, but not too far, and let the printer find the right exposure
for this yellow film, and keep the prices low. And everybody's
happy.
Well, almost.
If you *really* want control, you're going to have to do it yourself,
or pay someone to act directly on your behalf, like a professional
lab.
Pierre Renault
speak on their behalf, but I'll try to explain.
C-41 is one of the least tolerant film processes their is. Film goes through
once, and that's it. No retries on the fix, nothing. You get one chance to getit perfect. C-41 film, on the other hand is a very tolerant film. People
get acceptable results over a wide range of exposure/colour balance situations.
The reason Dave Bernard is having problems with different labs resides within
a few situations.
The average photofinisher provides service for people who don't wish to pay
more than $20 for a roll of thirty. Some of these don't even want to pay
more than $15. (That's Canadian. Translate the currency.) If you think
about it, that's quite cheap. I have people asking me to process and print
some black and white for them and to charge them for it. Assuming I used
a print processor, it woult still take me an hour to print 36 good prints.
If you substract soft and hard cost, at $15 for a roll of film, I'm left
with $3 or $4 dollars for my effort. It would be very difficult for
anyone to understand why they should pay me $40 or $50 for a roll of film.
The average photofinisher is faced with the same problems. The market
wants cheap prints. The market wants results in an hour. The market wants
a nice white-tiled store in a shopping centre. These things all cost money.
The compromise will be made at the film/paper process stage.
There are currently one-hour machines available that run without running
water. The print are washed with a chemical solution and recycled water.
The prints are not archival. The rating for their lives is "good to average"
but nowhere near what people who do their onw B&W would find acceptable.
Who's fault is it? No one party's really. The average user can't afford
film processing at $50, nor do they need it. The average amateur (by
"amateur", I'm refering to the hoobyist/obssessed artist type. ) cannot
spend the amount of money he would like for this level of work. The
average lab cannot affort to pay the hourly rates properly trained
individuals command. People are not willing to wait for their film when
a process problem comes up, they want their film.
There are good labs and there are bad labs. I've seen people reccomend
pro labs that I personnaly knew had uneven runs managed by incompetents
while other labs that ran spotless processes died. The reason for this?
Most people don't know good processing until they sit down and compare
films, and people rarely do this. To properly evaluate film processing,
whether positive or negative, demands a certain critical eye, a detached
reasoning/evaluation, the right sort of equipment.
I remember a respected pro photographer around here who dropped off a few
of the test rolls he had shot for a national newsmagazine. He had a habit
of spreading test rolls out to various labs each year to see what was up.
After picking up the rolls, he returned a few days later with more film.
After a few weeks of this, I dropped by his studio to pay him a visit. He
asked me in to talk about his processing. I imagined I would be going in
for a verbal beating. He had been examining the results of the shooting
for the first rolls he had dropped off. My main competitor's film was
their next to mine. He, the photographer, siad something like this:
"Look at them! Can't you see it? I used to think my lab gave me good
processing until I saw yours. How do you do it? What's different about
your process?" This went along these lines for 30 minutes or so. I had
to explain to him that his problem (blocking up of the whites, high blue
contrast, etc.) was quite common. The process plots for my competitors
line were probably very close to mine, the difference was in the subtlties.
Point is, that he hadn't realised the differences until, after two weeks,
he decided to compare film to film and examined them. Every point he raised
about the differences where the result of analysis, not "judgment".
That was for E-6 film. The problem is compounded in C-41: there are two
steps involved. If people were to examine carefully the prints in any
well-printed magazine, they would see the same sort of problems cropping
up. Even magazines that use Photo Shop corrected 24-bit scans still suffer
from certain problems. Take any flyer printed on newsprint and look at it
closely. What you are examining is photography done in a studio with the
very best equipment there is, shot on same-batch film, colour-corrected,
scanned, analysed, separated. There are still problems. If you look at
the shadows of white objects (the pale shadows), you'll see a pinkish/
magenta tone usually. This shift is unnoticeable in the film. Why did it
appear in the print? Beats me, but I do know that the problem is in the
original E-6 process and no amount of correcting will fix this. This is
actually quite interesting to do, analysing magazine printing, it reminds
one to constantly question accepted values.
The same sort of problems exist in the C-41 process. First let's look at
colour negative film. Colour negative film is balanced for specific colour
balances of ambient light. Nexty time you're at a camera shop, ask to see
a roll of Vericolor film. Exmaine the box and you'll see that Vericolor
Type-S is balanced for Electronic Flash, not daylight. To compound
things. Vericolour S is designed for use by the wedding photography trade.
Expect good skin tones (for white people) and nicely graded whites (for the
wedding dress), but only expect this with flash, not with daylight.
You can experiment with this and see for yourself: set up a situation using
reflectors and such and shoot outdoors. If you really want to play it super
scientific, shoot in ANSI standard daylight conditions (between 10AM and 2PM
on a lightly overcast day in the sunshine). Then shoot the same set-up using
electronic flash, same exposure. You'll be shocked. The flash pictures will
jump out at you when you see them.
Same goes for shooting with the amateur films. The problems involved in
obtaining proper colour balance are many. I won't list all the reasons,
Donald Mathis had already done a good job explaining why.
I'll stick with colour negs. Colour neg films have quite a wide range
of latitude, but a very poor range of neutral densities. A very high density
on a colour negative film hangs around the .250 range. The highest density
on a well-processed slide film hangs around the .320 to .350 range. That's
a full three to four stops extra. The colour paper had a marginally better
density range, but that only as a result of double-opacity (by this, I refer
to the fact that light enters through the emulsion on a print, gets filtered
on the way through, bounced off the back, and gets filtered on the way through
again, doubling the density). The paper itself really has a very low range,
but that's OK because the negative has a very low contrast. The result is
a nice acceptable print.
The problem is this: any little error in the process results in the sort of
problem I was describing earlier about E-6 and colour separations. Any
little problem in the paper process compounds this. Another factor is
the client's misconceptions about what makes a nice print. For example,
I frequently see rpofessional photographers request "warm and deep" on
their work when they get rpints. Few pros have ever asked for "cold
and neutral" in people pictures. THe ones that do discover just how
pleasing the results can be. Same goes for slide film. Few photographers
will deny their liking warmish slides. Few of the same photographers have
ever sat down and examined the range of results from reddish to warm to cold
to bluish. (If you want to know, I prefer an extremely neutral balance with
maybe 1CC or 2CC at the most towards blue, it's surpisingly stunning.)
Now, the kernel of the question, why don't these problems happen with slide
films? Well, they do. Quite often as a matter of fact, but most people, and
this includes too many pros and experts, can't tell. They are too busy
being stunned by saturated colours, the wide dynamic range of slide films,
the richness apparent in such a small picture.
If ever a lab had a problem with it's E-6 line, it had two or three ways to
tell that there is a problem.
The best process monitoring involves the collective smapling of many factors.
Time, temperature, chemical concentration, replenishment rates are all
factors that should be involved in the consideration of whether there are
problems in a process. An E-6 line giving "good" results when running
6F too high, with double the replenshment rate is giving inferior results,
ragardless of what the control strips say and what a quick eyeball of the film
says.
Second method: the control strip. (The control strip is a piece of film
that is purchased from the process manufacturer, Kodak or Fuji, and processed
along with the line. Results are compared with a previously developped strip
with corrections applied for variances in the manufacturer's process line.)
Many many labs ignore manufacturer's reccomendations and follow the control
strips as if they were godly statements. All manufacturers emphasise that all
control strip evaluations should be confirmed by visual inspection and other
means. Control strips are affected by the same sort of problems film is, I've
seen control strips not reveal colour imbalances when the process was running
with non-standard chemicals.
Third method is eyeballing the client's work. This is easy to do with E-6
and K-14. Quite hard to do with a negative. I do not know anyone who can
tell me within 10CC whether a colour negative film is well balanced or not.
Now, this may seem like not such a big deal whether the film is properly
precessed or not, but it is. The paper is made for a specific film. There
are many more than three colour layers in any film and the same goes for the
paper. The problem is matching curves.
When you vary B&W process, the shape of the curve changes: the length of the
toe and the shoulder varies, the "overall contrast" changes, many things
happen. You can examine this yourself. Shoot five or six rolls of film
of the same scene/lighting/exposure-bracketing situation. Soup in two or
three different developpers (to create the effects of chemical imbalances
in the C-41 process) and tthen try to print each and every negative on
different papers (to simulate variances in the paper process). Go for a
standard shade when you do this, not for a "best" print. Why standard
shade? Because the problem you are having with the colour neg is when the
colour is off, not that the colour looks nicer. Now, compare all the prints.
Look at the shadow contrast, the medium contrast, the highlight contrast.
Examine the gradation of the various shades.
What you should see is this: some combinations will yield low contrast in the
shadows, high contrast in the highlights all the while having coarse gradation
in the shadows, and fine gradations in the highlights, and any permutations
of factors.
Now, consider the colour negative. You have to do this for 14 layers! If you
want to consider this as three layers, then fine. You have all these problems
compounded to three different colours. And the lab that processes the negs
cannot tell if the control strip is revealing a problem or not.
Why? Well, different film have different masks, different manufacturers
film look different when properly preocessed/exposed. By the time
a careful study and analysisi of the prints is done, the client has
already picked up his 36 prints and paid his $13.95 and gone. What's a lab
supposed to do?
INSTRUCTIONS PART: What to do. Pick a good lab. To figure out who is
running a good line, send them negs for a long while. If you live in
North America, send them Kodak film (why Kodak? because most labs use
Kodak chemicals, Kodak Control Strips. Fuji's C-41 equivalent is not
quite the same as C-41.). Examine the prints you get back and look at
the colours and contrast carefully. Look for colour imbalances in the
highlights and in the shadows, not just in the dark-to-light midtones.
Do this over a period of time to see if the process varies. If the lab
seems to be having problems and you are a reasonably big enough client,
then discuss this with the owner/manager. Whatever you do, do not use
an accusatory tone. Just ask for information, specific information. If
you seem to be getting yellowish highlights, then ask about yellowish
highlights. (in the case of imbalanced highlights, the lab may just be
experiencing the problems associated with under-used print runs due to
oxygenated developer. The problem usually varies from non-existent to
obvious from day to day depending on the production. Not a big problem,
really.) If your concerns are legitimate, then the lab should be
forthcoming. If you are experiencing a problem, try to be as analytical
as you can. Ask yourself: Is the entire print imbalanced or just the
contrast of one colour (the opposite imbalance in darks and highs)? Does
the problem only occur under specific shooting curcumstances? Etc., etc.
Once you have found a lab that offers what appears to be a good line, and
who seems to maintain it fairly constantly, then stick with that lab. Now
you can start to correct the problems you are causing. Whenever you
can, include a picture of a grey card in the scene (for 35mm the gray
card should be shot slightly out of focuse and should fill at least
one quarter of the frame.) and make sure that the card is list with the sme
sort of light that the scene will be. Then ask for a contact sheet based
on the grey picture or prints with the printer set to manual all based
on the grey card scene. Repeat this over many weeks, correcting problems
as you go. If you are getting familiar with the lab people, then let them
know what your results are. Don'
Don't try to be snarky when you do this. They probably are mcuh more
knowledgeable than you are and won't take kindly to misplaced or
misinfirmed criticism. If you notice and improvement, like better
gradations in the highlights, then fergawdsake, tell em! This will help
make their day, and will help to guide the process monitor towards better
results. And pay your bill promptly if you have an account. I've seen
too many good labs suffer due to poor receivables. The people that spend
the extra effort to give better results deserve to be paid. If the prices
are a little high, don't complain. If they are too low, then tell your
friends to bring their stuff there. Whatever you do, be loyal. There is
nothing worse for a lab than to spend the time and money to get the best possible results only to have the clients say "Hey! How come E-Z Kolor down the
street gives me 3 extra 4x6 prints with my film for only $10, everything
included?"
Otherwise, you are only encouraging poor work.
Brian Godfrey
>(or "complaint" as someone else called it). What I read is a lot
>of defenses why printing is complicated. IMHO, I do not think
>that answers Dave's question. Assuming the negative is new
>(very little aging), and assuming the negative was processed with
>rigid controls (as suggested by follow-up posts), then consistent
>good printing should be achievable without paying a premium to a
>custom lab.
Spoken like a true computer professional. :-)
I don't mean to make fun of you, but to point out that our business
(computers) deals in bits and logic and absolute rules and things that
always work exactly the same every time unless they are broken, while the
rest of the world is much less precise.
It is possible to state assumptions which would make photographic
printing flawlessly reproducible, but they would be unrealistic assumptions.
I have seen statements, in this thread, that slide film is always accurate
and always "right" so why can't negative film be. Well the fact is that
slide film is *not* always accurate. Many photographers are careful to buy
their slide film in large lots from the same emulsion batch and freeze
it. Then they calibrate to determine necessary filtration to render the
color characteristics they desire. And they buy expensive Kodak color
correcting filters and do other manic depressive things to be sure fo their
color accuracy - and to their eyes, they are successful. (How accurate
are *your* eyes?). And every now and then we see people complain that their
Kodachrome was too green or their Velvia was too magenta or something. So
what if you hate magenta Velvia? Will you be happier with green Kodachrome?
There are variables in silver and dye that we cannot just wish away or assume
away.
So what about printing photos? Well, I have a couple of packages of
color printing paper in my fridge with labels on them recommending starting
filtration, but reminding me that differences in the color temperature of
the bulb in my enlarger (color changes with age of the bulb) and many other
variables may necessitate variation from the recommended filtration. And
the two packages start with different recommendations.
I could go on, but anyone who hasn't got the point by now, won't.
>There is one solution no one seems to mention very much and I am
>seriously considering myself: simultaneously shoot both print and
>slide films. Use two camera bodies or interchangable backs (med.
>format cameras). Yes, clearly this is not practical for many
>applications of photography, but for some applications (landscape)
>it is very feasible. The transparency being valuable in itself
>would also serve as a guide for the printer of the negative film.
You could do this. It sounds expensive to do all the time. I frequently
shoot transparency and B&W negatives of the same composition, but for
different reasons. Another option is to buy large lots of film from the
same emulsion batch and freeze them and find a reliable lab (not a cheapo)
and you should get pretty consistent results - assuming you make pretty
consistent exposures. But there's another of those pesky assumptions
again...
>This suggestion is mentioned only in acknowledgment of what has
>been said before: negative film and slide film are different
>animals; if you want prints, shoot negatives; if you want slides
>(or want to publish) shoot slides.
Sounds like pretty good advice to me.
--
--Brian M. Godfrey
atlastele.com
Russell Williams
viewed in different conditions. Slides are viewed singly, in a
darkened room, while prints are viewed with other objects (and other
prints) visible.
Your eye isn't very sensitive to the fact
that incandescent light is yellowish compared to sunlight when they're
viewed by themselves. But turn on a reading lamp near a window, and
the yellow cast is obvious. Similarly, when watching slides, you
won't notice minor color differences from one slide to another. Put
those same colors on paper, next to the white table and your hand
and other prints as references, and the color differences are
immediately obvious.
If there were something magical about the processing of slides that
gave more precise results than negatives, you might expect that making
reversal prints from slides, which uses basically the same process
and controls as the slide developing, would give consistent
results without manipulation. But it doesn't -- you have the same
filtration problems and variations when printing slides as you
do when printing negatives.
--
Russell Williams
Not speaking for Apple Computer
Michael Covington
color slides. You must adjust the density of each of the 3 color
layers to fit it into the available range, trying to mimic the way the
human eye adapts to scenes with different kinds of illumination.
--
:- Michael A. Covington internet mcov...@uga.cc.uga.edu : *****
:- Artificial Intelligence Programs phone 706 542-0358 : *********
:- The University of Georgia fax 706 542-0349 : * * *
:- Athens, Georgia 30602-7415 U.S.A. amateur radio N4TMI : ** *** **