Film
Slide Film Versus Print Film
I like slide film. The main reason is that a slide immediately shows me
the real result of my work. The film after all is the place where
the light got converted to an image so it contains the most unadulterated
information. Any subsequent steps such as making a print or digitizing
can only lose data and degrade that image. With slides, there is only one
step: Develop the film. You can then see the best possible results immediately.
(Little footnote here. There are image processing techniques that involve
"deconvolution" schemes that can counteract the effect of the diffusion
of light within the photographic emulsion thereby leading to an even better
image than the one recorded by the film. I'd still want to start by digitizing
a slide instead of a print however. A negative would be ok for that purpose
too of course.)
The problem with print film is that after the film is developed, "somebody"
has to make a print from the negative. In my experience so far,
that "somebody" has no idea what an astrophotograph is supposed
to look like but somehow deduces that what it needs is three stops of brightening!
That
washes out all the detail and you don't know whether you took a good shot
or not. Unless you have your own darkroom and all of the equipment and
chemicals to make prints, I think slides are a good idea.
The other thing is that prints, in my opinion just look duller than
slides. I have never seen a print that I would prefer to a slide. I have
a small portable slide viewer that I got from Edmund Scientific that I
use to show other people my slides. It consists of a light that shines
on a diffusion screen that the slide is placed in front of then a magnifying
lens so that you are looking directly at the illuminated slide through
a magnifier.
BUT, if you have a good slide/negative scanner (I am very happy with
my $500 HP PhotoSmart Scanner), and plan to view your images on a computer
screen or publish them on Internet, then you should consider print
film. The prints you will get from the one-hour photo processing lab
will probably still look terrible but with the scanner and an inexpensive
photo processing system such as PaintShopPro, you can get rather spectacular
results on your computer screen that are as good or better than you can
get with slide film.
Film Speed
One of the things that I found surprising is that "film speed" does not
correlate well with "true film speed" for long-exposure low-light work.
(See Reciprocity Failure.) "Slow" films
can be faster than "fast" films for astrophotography. These things must
be tested either in the field or with procedures discussed in "A Manual
of Advanced Celestial Photography" by Wallis and Provin.
Which Slide Film?
Before I started doing astrophotography, I had done a lot of photography
with Kodak Ektachrome film. I compared some Ektachrome and Kodachrome shots
of the Grand Canyon and concluded that the Ektachrome film was giving me
more accurate colors. At least they were esthetically more pleasing. So
it was natural to begin using Ektachrome film.
Ektachrome 400
Since I knew the objects I would be photographing would be, for the most
part, rather dim, I concluded that the fastest film would be best (later
I discovered that this is not necessarily true) so I started using Ektachrome
400. I found that it produced very good color-balanced results even when
hypered.
I felt I was quite successful using hypered Ektachrome 400 film. Then Kodak
stopped making it.
Ektachrome Elite II 100
This film has amazing reciprocity failure
characteristics that would make it an ideal astrophotography film BUT
for
long-exposure, low-light photographs, which astrophotographs tend to be,
its color balance, in my experience, is deplorable. I used it for quite
a few tests and red objects seemed to show up rather well against a decidedly
green background. Blue objects however did not record well enough to be
seen against all the green. Since in nature, there are red astronomical
objects and blue astronomical objects but few if any green astronomical
objects, this film did not seem to help me.
Ektachrome Professional E100S
So I thought, why not try the "professional" version. Presumably it would
have better characteristics. In actual practice, it seemed to be very similar
to the "unprofessional" version. I have a hypered unexposed frame
that is decidedly green. I took the Horsehead Nebula image using E100S.
It came out decidedly green as you can see in the image at the top of this
home page. The digital image of the Horsehead Nebula looks very much like
the slide with respect to color so I would say the digitization process
was rather accurate although that is a subjective judgement.
Ektachrome Elite II 400
This film may still be ok for astrophotography. I have an unexposed frame
that is not green. I have not experimented with it because I have become
interested in using films with finer grain. Fast films tend to have larger
grains and that granularity puts a limit on the detail that can be resolved.
Since my autoguider is performing so well, it seems logical to try for
all the detail possible.
Agfachrome RSX100 Professional
Became interested in this film by comparing data in "Peterson's 1997 Photographic
Buyers Guide". Downloaded further information from AGFA's WEB site. I was
of course looking for information regarding reciprocity failure.
Recent experiments with this film have given superb results! I hypered
it "in the can" 48 hours at 15 psi 8% Hydrogen, 92% Nitrogen at room temperature.
Took it out to the desert and shot a three hour exposure of the Cone Nebula.
Got the Cone Nebula and a rather bright blue reflection nebula centered
approximately at 06 40 38, +09 74 36. I have not yet had time to research
the reflection nebula. I am very happy with this film however! No green!
Sensitive enough to shoot the cone which, in my opinion, is a rather difficult
object.
Was able to add a shot of the Crab Nebula, M1 that I just got back from
the Digitization Lab. An unexposed frame that I also had digitized showed
average color densities of 35, 41, and 40 for RGB values indicating that
red light intensity is captured a tiny bit less on this film but overall,
it appears to be pretty good. Measurements of dark areas in the M1 shot
showed color densities of 33, 38 and 42. I made corrections to the M1 image
to compensate the red and green and adjusted the overall brightness and
contrast of the image but did no other image processing operations on it.
The original slide shows somewhat more detail in the red filaments and
I wondered if I should have had it digitized at a density greater than
1200 DPI. So I sent it back to the lab to have it digitized at 2400 DPI.
Could not coerce any more detail out of it. Looks like 1200 DPI is good
enough at the moment. Of course I'm not sure that we were talking "optical"
or "interpolated" 2400 DPI or not. Some of the scanner sellers are doing
"specsmanship" at the moment and advertising "interpolated" resolution
instead of "optical" resolution. The only thing that counts is "optical"
resolution. You do not get any more "information" when you interpolate,
you merely break up the big pixels into little pixels to increase the number
of pixels in the image and then smear that information around a bit.
Which Print Film? PJM-36 Ektapress Multispeed
640!
I had not seriously considered print film until 1 November 97 when Jerry
Lodriguss appeared at our observing site as a guest of Tom Polakis. Before
he left early Saturday morning, Jerry gave me a roll of unhypered Kodak
PJM-36 Ektapress Multispeed 640 print film and urged me to try it. I tried
it Saturday night and was astounded by the results! Thanks Jerry! I shot
several images that night including the Horsehead and Cone nebulae. In
the past, I had trouble getting any image at all of these objects so I
had been forced to shoot two to three hour shots at f/6.3 using the Meade
Focal Reducer. I decided to put Jerry's film to a no-holds-barred test
so I reshot these objects at f/10 using the times I had used for my f/6.3
shots. I expected to get blank film back from the one-hour photo lab I
used. Instead, I got the best shots yet of the Horsehead and Cone nebulae!
Jerry says he doesn't bother with hypering this film anymore. It works
too good as-is. (I can't resist the temptation to see what hypering will
do though so I might try it anyway...) The film comes in a 5-pack that
cost me $25.00. Pretty good price I think. I have the film in the refrigerator
now awaiting my next photo expedition. I will probably use my film loader
to load half the film onto an empty reel since there is not enough time
in one night to shoot 36 deep-sky photos. 10 would be stretching it. That
way, each 36 exposure roll of film becomes two short rolls of film. The
other interesting thing is that the film required no color compensation!
Reds and blues appear to be correctly balanced even at the long-exposure
low-light conditions characteristic of deep-sky photography. Note that
this film is referred to frequently as PJM-2. The only place that designation
appears is along the edge of the film after it has been processed. There
apparently was a PJM-1 film that was not as good. Philip Perkins has reported
better blue response if this film is hypered. Extrapolating his hypering
times led me to conclude that 24 hours, in the can, 15 PSI, 30 Degrees
Centigrade would produce hypering similar to his. My shots taken on 20
April 98 convinced me that at least no harm resulted and I did get reds
and blues within M101 and M51 as expected. No way for me to tell yet if
it is better, worse, or no different than unhypered results would have
been.
Rudimentary Color Correction
In color compensating the digital image of the Ektachrome Elite II 100
Horsehead photo, I determined that the average pixel values for red, green,
and blue in "dark" regions were 31.0, 41.75, and 23.75. If the film had
maintained color balance, all three of the numbers would be the same or
nearly the same. To color compensate the film, I multiplied all of the
red values in the image by 1.37 and multiplied all of the blue values in
the image by 1.76. That helped as is seen in the second Horsehead image.
I do not believe that the reciprocity failures were linear however so,
actually, to get it precisely correct, one would have to know what the
failure curves look like with respect to brightness and compensate each
pixel of the image with values specifically related to the intrinsic brightness
of that pixel.
Film Digitization
One of the most difficult aspects of getting these slide images into a
form suitable for posting on internet was getting them digitized. I tried
Snappy but that didn't work too well although it works wonderfully for
a lot of other things. The dynamic range and AGC circuits of my video camera
were definitely contributing factors. Finally discovered "Arizona Photographic
Imaging" that is convenient to me. They in turn worked with "Arizona Microimaging"
which is apparently now defunct. Scanning was done with a Nikon LS-1000.
They could scan up to 2700 DPI. I had been requesting 1200 DPI since it
cost less and required less space than a high density scan. The images
were placed on a CDROM in TIFF, BMP, GIF, EPS, JPEG, or PSD. I requested
TIFF. Very convenient, reasonable cost, and they could add images to an
existing CDROM. I also requested that they not do any "image processing".
I want the scanned image with nothing done to it so that I can, among other
things, evaluate the color balance of the film. (Assuming that their scanner
was color balanced.) This essentially gives you a "poor man's densitometer"
for color images.
After Arizona Microimaging folded, I again tried to figure out where
to get slides digitized. I sent some to Kodak to have them put on a Kodak
"PhotoCD". This also required me to buy a $20 Kodak software package to
decode the Kodak proprietary format that they store the data in. Main drawback
is the 14 working-day turn-around time, i.e., three weeks!
Then I found *the* answer, the new HP
PhotoSmart Scanner! This has solved my slide scanning problems completely!
Film Hypersensitization
I have been using Lumicon's gas hypering apparatus to hypersensitize film
that I have been using for astrophotography. This consists of a pressure
chamber into which new film is placed. The pressure chamber is then evacuated,
then pressurized with gas which is 8% Hydrogen and 92% nitrogen. The film
is left in the pressure chamber for some length of time. I have been using
two days at room temperature. Film which is treated in this fashion becomes
more sensitive to the low-light, long exposure conditions required for
deep-sky astrophotography.
I recently bought the temperature controller for the heater that is
built into the pressure chamber. The controller I got is set to 86 degrees
F. The downside is that this is Arizona where I might have to put the whole
apparatus in a refrigerator so that I can maintain the 86 Degree temperature!
(Just kidding - Unless my air conditioner fails...). I plan to use 48 hours
at 86 Degrees for AgfaChrome RSX 100 film.
I initially concentrated on Kodak Ektachrome film primarily because
I found that film to provide the most accurate (in my opinion) colors for
general landscape photography here in Arizona. When I compared the results
of Ektachrome with Kodachrome, I was much happier with the Ektachrome rendition
of colors in shots I took of the Grand Ganyon for example. (Of course this
is somewhat subjective. I read long ago that eastern people in the USA
set their TV sets to display pale flesh tones and that western people in
the USA set their TV sets to display darker flesh tones presumably because
western people are more used to seeing people with suntans. Of course that
was before "Bay Watch".)
I used Ektachrome 400 film almost exclusively until Kodak stopped making
it. I did use Konika 3200 for shooting Comet Hyakutake. For Hyakutake,
I used my Canon AE-1 Program camera with a 50 mm lens. I did not have a
camera mount on the LX-200 at that time so I just used a tripod and took
various unguided exposures. The 3200 speed film was necessary because,
according to calculations I made in advance, the longest exposure time
I could expect without star trailing was 18 seconds when Hyakutake was
at its closest approach. Further calculations based on that indicated that
the 3200 speed film would be essential. In actual practice, exposures of
at least 2 minutes were required in order to begin to show the tail very
well. For Comet Hale-Bopp, I mounted the Canon camera on the LX-200 scope
and guided the photos. 2 to 10 minutes seemed to be best but it varied
depending on conditions. Most books on the subject say to shoot film like
its going out of style and try all sorts of exposure times since it is
often a long time between comets. I will put my comet images up if I can
ever get them digitized...
Film Reciprocity Failure
An object half as bright should take exactly twice as long an exposure
time to result in the same film opaqueness or "density". "Reciprocity failure"
refers to the fact that it takes LONGER than twice as long and gets worse
and worse as the object gets dimmer and dimmer. This is why two and three
hour exposures are necessary. If there were no reciprocity failure, we
could manage with far shorter exposure times for deep-sky photography.
We look for film with as little reciprocity failure as possible. Hypering
the film reduces the reciprocity failure somewhat.
Exposure Charts
One of the best books I have found for getting some sort of handle on exposure
times is "Astrophotography" by Barry Gordon. The system described in that
book works relatively well and at least gives you a starting point. I used
his book to determine exposure times for Jupiter, Mars, the moon, and various
deep space objects. I was able to devise exposure charts using his methodology
that allowed me to quickly determine a nominal exposure time about which
to "bracket" my photos. I have made numerous exposure charts. I have charts
for my telescope at f/10 using 400, 1600 and 3200 speed film; at f/6.3
using 400, 1600 and 3200 speed films, etc.
For deep space objects other than the Orion Nebula, my observation so
far is that for the hypered Ektachrome 400 film I was using, you can hardly
go wrong using two hour exposures. The Horsehead and Cone Nebulae need
two to three hour exposures and, in the case of the Cone Nebula, a little
more wouldn't hurt.
Copyright 1997, 1998, 1999, 2000
Howard C. Anderson
RETURN TO
HOME PAGE