Back to Astrophotos

Slides Are Great
We astrophotographers use both color negative and slides films to shoot nebulae,
clusters, galaxies
and other deep sky objects. Even when we do use slide films like Kodak E200 in
the field, rarely are those slides suitable for optimal display by a 35mm projector.
The field slides may have significant sky glow, vignetting, and unaesthetic color
balance or luminosity profiles. Some astrophotographers use medium format films
in the field, which yield better detail, but medium format projectors tends to
be odd, primitive, and harder to find.
The serious astrophotographer scans raw images at about 2700 DPI, then digitally
post-processes them in Photoshop and other software
tools. Sometimes multiple images of an object are "stacked", yielding even higher
quality. As a result, post-processed images are generally vastly more presentable
than raw color images from the field.
Post-processed images are displayed on web pages -- that is, peoples'
monitors -- and hi-rez versions are printed on ink jet printers. Sometimes
we want to show our work in its full glory to our families and friends,
at local astronomy clubs, or at star parties for kids. 35mm slides have
an output dynamic range that exceeds prints', and it is practical to display
slides on 6 foot by 8 foot screens. Decent used 35mm projectors can be
had from eBay for under $50, which cannot be said for LCD projectors. Compared
with 35mm slide projectors, overhead (8"x10") slide projectors usually
have poorer optical quality; are positioned closer to to the screen and
off-axis of it, yielding distortion and uneven focus; cannot be remotely
controlled; don't have features like automatic timer advance; and are less
portable. These advantages make 35mm slides a better choice for presentations
than prints, LCD projectors, computer monitors, and overhead projectors.
How to Make Color 35mm Slides
There are a number of reasonable methods:
1. Write your post-processed digital image files
onto a CD-R or Zip disk. Bring them to your local film lab, and pay $8
per slide. They use a film writer.
2. Mail your CD-R or Zip disk to Tony Hallas for
$7.50 a slide. He uses a better film writer. This takes longer.
3. Make your own slides from a computer monitor
for $0.40 each (including mounting). I discuss this option in detail below.
If you don't have access to a CD recorder drive or a Zip drive, I guess
that eliminates options #1 and #2.
1. I brought a few of my post-processed astro images to a local film
lab on CD-R. Although it has proven to be an excellent lab for me in every
other respect, it turned out that their film writer, a Polaroid ProPalette
8000, was just not up to the task of producing pinpoint stars. In images
with rich star fields (like the North America Nebula above), the initially
resulting slides had bloated stars that dominated the field, and the nebulosity
was much less impressive. The operator did the best he could, but the ProPalette
simply was not designed to render astro images. In my opinion, the final
slides came out so-so. The operator's efforts to tighten the stars gave
some of the images a reddish cast and lost detail.
2. I talked with Tony Hallas, a well known astrophotographer and astrophoto
lab operator. He said he never got his ProPalette to work well either,
so he sold it in favor of a Lasergraphics Mark VI film writer. Although
I have not yet used Tony's film writing service, I believe that his astro
slide rendering is better than my local film lab's. Tony can be contacted
at http://www.astrophoto.com.
3. The slides I made from my monitor aren't as good overall as the film
writer's, but they are good enough for most purposes. This is discussed
in detail in the following section.
All of these methods will increase contrast, which is not always good.
You can counteract it by reducing the contrast of the source image or of
the monitor.
The pricing for options #1 and #2 assumes only one slide per image.
The second slide costs $4 or $5. For larger quantities, you may want to
use commercial slide reproduction, which costs about $0.45 per slide, with
collation, sleeving, numbering and imprinting.
Making 35mm Slides from a Monitor
Obviously, the better the quality of the monitor, the better the result.
Also, all else being equal, a large monitor is better. 17 inches is a good
size. If the monitor's screen is very curved, it can distort the image.
The color
temperature of the monitor is very important. In general you want the
color temperature of the monitor to match that of the film. Some monitors
are permanently set at 9300K, which yields a very blue cast. Your eye probably
does not notice it, but it will show up severely on film. 5500K
is an appropriate color temperature for daylight film. If the monitor has
no color temperature adjustment, you may be able to simulate warming by
reducing blue and increasing red levels, using the monitor controls. The
brightness and contrast settings of the monitor should be set according
to Chuck
Vaughn. You may also want to mess with gamma or ICC. For slide shooting,
I set my ViewSonic A75s monitor to 100% brightness and 50% contrast. To
learn more, read your monitor instruction guide and experiment with its
adjustments.
If you cannot make the color adjustments on your monitor, you can use
Photoshop's custom display settings. Or, in Photoshop you can either curve
each of the images specifically to be photographed on screen. Or, you can
use a filter
on your lens. Usually an appropriate filter is an FLD, which is magenta
colored.
The monitor screen should be clean.
Adjust your Windows (or Mac) screen resolution to the highest you can.
I set mine to 1280x1024 and 32 bits per pixel "True Color". If you have
to use a slower refresh rate, that's OK. You may also want to hide your
Windows Taskbar.
Open your high resolution image in Photoshop (or whatever tool you use).
Try to use most of the screen for image display. In Photoshop you can get
full screen display by typing the letter "f" once or twice, then "tab"
one or twice, then "control-zero". (Hit "tab" and/or "f" again to leave
that mode). If your zoom factor is more than 100%, lower it to 100%. Make
sure to move the mouse cursor into a corner. The image is ready to be photographed.
(Some monitors' pixels are not square. If you really need to have perfectly
calibrated X-Y scaling, you might need to proportionally size your images
in Photoshop.)
Set up a tripod such that the camera is on center and level with the
monitor. Make sure the tripod knobs are tight. I prefer an Olympus 135mm
f/4.5 lens for this work. Its long working distance minimizes screen distortion
and gives me enough room to get to the keyboard without knocking the tripod.
I use a lens hood.
My favorite film is E200, but any ISO 50-200 E-6 film should be fine.
You'll need to figure out your camera viewfinder's field of view (FOV).
My viewfinder FOV is 97% of the film's FOV. This means I need to crop the
frame a little tight. I move the tripod backwards or forwards to do this.
If I position too far out, after development I'll see a black strip on
one or more edges of the slide. Although zoom lenses allow for easy FOV
adjustment, zoom lenses simply do not produce images as good as fixed focal
length lenses. A focusing rail is the perfect tool for this job, but not
many folks have one. Some cameras viewfinders show even less than 90% of
the film's FOV. I have used cameras where the viewfinder FOV is actually
larger than the film's. You want both the monitor and camera in portrait
mode. Usually you will be more constrained in the vertical dimension.
You will want to use a minimum exposure time of 0.5 seconds, to avoid
monitor scan artifacts. Up to several seconds should be fine. One second
worked pretty well for all my shots. Use the camera's manual exposure mode.
What f-stop should you use? Well, that's going to take some experimentation.
I use a range of f/8 to f/22, depending on the image. A high f-stop is
good, because it obtains more depth of focus; the entire screen can be
in focus even if it is curved. However if you are at the lens's highest
f-stop, you may lose fidelity due to diffraction. You should bracket your
shots at least one full stop, or two stops by thirds if you're highly motivated.
You can use the camera's light meter or a hand held one to get a baseline.
Obviously you don't want anything moving during the exposure. You should
use mirror lock-up and aperture pre-fire if you have it. My OM-4T does
both if I set the self timer. That means I need to wait 12 seconds for
the shutter to fire, but I don't need to use a cable.
Before you open the shutter, close all the shades and turn off all the
lights. You don't want reflections on the slide. In fact, if you have windows
in the room, the shooting is best done at night. Even a glowing LED in
the room can reflect off the monitor surface.
You probably won't get the slides right on your first try. But you might
on your second try if you take good notes.
How Do They Compare?
| Aspect |
Screen Shot |
ProPalette Film Writer |
| Effort |
Get a roll of film, set up camera, display images on screen, shoot |
Get a blank CD-R, prepare images (they need to be cropped to a 2:3
aspect ratio), write onto CD-R |
| Grain |
The pixels (or phosphors) on the screen shots barely show in a 4x loupe,
but in a 20x loupe they are obvious (see scans below). If the pixels are
a problem, one option is to use a diffusion filter on the camera lens. |
The ProPalette writes at 8000 DPI, so the only graininess I see in
those slides is from my field film, and much of that grain got lost somehow. |
| Color Balance |
Totally under my control. Once I got the 5500K color temperature right,
the balance was excellent. |
Good on most slides. Some slides had a red cast from the curving that
the operator used to reduce star size. |
| Turnaround |
12 hours |
48 hours |
| Cost |
$0.40 per slide |
$8.00 per slide |
 |
Image A. This is a "magnified" 640x480 section from the master
post-processed 4000 DPI scan, centered on the "Pelican's eye" from the
image above. This is from the source post-processed 5000x3400 digital image,
which is the source for the display during the screen shots. However, the
camera does not "see" the image at this resolution, because the display
has only 1280x1024 pixels, much fewer than the source image has. |
 |
Image B. In Photoshop, I resized the entire image to 1280x1024
(my display resolution), then I resized it back up to the original size,
and cropped this section. This is only to illustrate what a display's
undersampling does to an image. I did not do this during the slide shooting. |
 |
Image C. This was from the resulting slide that was photographed
from the screen. As you can see, the process is lossy, both in resolution
and dynamic range. The screen shot was exposed for 1 second at f/16, using
6500K on E200.
Images C & D do not look as nicely here as the backlit slides do.
Both Images C and D are manually curved non-linearly to make them more
comparable. |
 |
Image D. This is from the slide written by the Polaroid ProPalette
8000. This is lossier than I would have expected, since it was made from
the hi-rez digital master (Image A). I think the operator used blurring
to reduce star size. |
Additional Info
http://thorin.adnc.com/~cymarron/monitor/index.htm
http://www.digidot.com/sean_homepage/crt/photocrt.html
http://www.artzonephoto.com/artzone/filterzone/temperat.htm
http://cybaea.com/photo/color-correction.html
Back to Astrophotos
|