Summarized from presentations prepared
for various regional planetarium conferences
June and October, 2001
© 2003, 2002 Loch Ness Productions.

The purpose of this presentation to show you how we at Loch Ness Productions make projectable, presentable planetarium-style slides. Clever planetarians should be able to see how they could apply these techniques in their everyday work.

I know, everyone thinks slides are a dying medium. Could be, but in the words of Monty Python, "I'm not dead yet." I say we should enjoy slides while we still have 'em. We're living in the heyday of analog/incandescent projection. These are the good old days. It doesn't get any better than this. Things may eventually move in a different, digital direction, but slides per se are the T. Rex of planetarium visuals -- call it the evolutionary pinnacle, or the end of the line.

The good news is that for those of us who learned the traditional planetarium methods of making slides, the old ways still work. And more good news, what we can do now is use computers in the slide making process. This may or may not make our lives easier, but properly used, it does allow us to achieve far more professional results more easily.

Way back in the 70s, when I worked at Fiske Planetarium, our plan (and every planetarium's plan back then) was simple: the artist would paint an artboard. We'd photograph it, process the film, mount it, and there would be our slide.

That was the concept then, and that's still the concept today. You take a picture of flat artwork, process the film conventionally, mount the slide, there you go.

In order to do that, you need slide-making equipment, and that's a calibrated camera/copystand.

The copystand doesn't have to be fancy, though it helps. You can achieve the same basic results using a camera on a tripod, with images tacked to the wall.

Here's the copystand we use, made by a company called Bencher, in Chicago, Illinois. (Actually, they don't make this particular model anymore, but they do have models that do basically the same thing.) You have a table to hold your art board, lights with polarizing filters to minimize glare, and the camera holder which slides up and down the vertical column to adjust the height.

The critical thing is consistency and repeatability. Your gear must be *calibrated*, so that you can set up and take pictures one day, then a month later, set up and take images that are identical, or as close to it as possible.

When I say calibration, in this case it means setting up a 1-to-1 correspondence between the camera and your artwork, so that when you shoot a picture, you know exactly how the image on the film is going to fit in the aperture of a Wess slide mount.

You note the distance from the camera to the printout that works (IOW, the rectangle image on the page fills the slide frame when mounted).

Copystand height marking is critical, so you can return to it consistently. Note the ruler built into the column of the copystand, which makes things much easier. You simply note the height, and you can even jot it down right on the artboard for convenience, along with any exposure adjustments.

There are a couple of "tricks" for calibrating the camera. First, it needs to have a pin registered camera back. This is a Nikon F3 that was modified by a company called Double M Industries, in Round Rock, Texas. (Actually, they don't make these kinds of camera modifications anymore, either. We really are dinosaurs on the pinnacle of extinction.)

When you're ready to click the shutter, pins drop through the sprocket holes of the film into carefully machined sockets beneath. The resulting exposed frame should be exactly centered between the sprocket holes (both vertically and horizontally), and every frame will be centered precisely, regardless of film winding tension, thickness, etc. Without pin registration, which physically forces the film into a fixed, standard centered position, there is nothing in non-modified conventional cameras to keep the film from otherwise sliding up and down, left and right within the camera as it winds. The picture ends up on the film, of course, and usually fairly consistently -- but almost always, "close but not close enough".

So you have a consistent height on the copystand, a consistent position of the film, now you need to position your artwork consistently.

Usually when taking pictures with a camera, you look through the viewfinder. Light from the subject travels into the lens, bounces off the mirror, through the viewfinder and into your eye. We reverse this process with our setup, and send light into the viewfinder, bounce it off the mirror, through the lens, and onto our subject. In essence, we use the camera as a projector.

With our our Double-M system came a fiber-optic lightpipe projector, the "hose" for which is seen in blue in these pictures. Riding up and down on the copystand column with the camera is basically a box with a bulb inside and a muffin fan to cool it. The light from the bulb travels down the light pipe, into the viewfinder of the camera (the eyepiece is fitted with a reticle to hold it).

Here's the next "trick" -- there's a chip of film with an alignment grid on it mounted in the viewfinder, and we actually project it onto our artwork. We line up the alignment marks on our printout with the grid, easy as cake.

We designed the grid to the exact rectangular aperture of a Wess #002 slide mount; we can see where the edges of the frame are (they're entirely within the viewfinder area), we know where the exact center is. We put into the grid indications for where the edges are for the different sizes of Wess circle and square mounts. So, in the illustrated example, we're aligning the image of Mars so it fits when mounted in a Wess 302 circle mount.

A word about that 3:2 ratio size. In theory, that's what most slides are -- 3 units wide by 2 high. Or 1.5:1, 300 x 200, 600 x 400, 1050x700, 900x600, 1200x800, etc. Wess 001 slide mounts are indeed exactly 3:2; the aperture is 34.3mm x 22.8m. However, the most commonly used Wess mount is the 002, which is just a smidge bigger all around: 34.8mm by 23.4mm. This is actually more like 1.49 to 1, or 2.98:2 -- in other words, the rectangle is a little taller than a perfectly-proportioned 3:2 ratio.

As mentioned above, in the olden days, you'd take the artboard or photo, make alignment marks on it, load color film in the camera and shoot it. But now it's the new millennium, and so the method we use today is... the same.

The only difference is what you place on the copystand. What used to be a painting from the artist, or a photo, or a page of a book is now a printout from your computer. The color artwork is a computer printout, the mask is a computer printout. Everything else remains the same.

The questions then become, how do we get images
   1) into the computer, and then
   2) out of the computer and onto film?

Our working plan is simple:

• You get every image into the computer.
• Then comes the work in Adobe Photoshop: you size it, crop it, clean it; adjust the color, contrast, brightness; make masks.
• When you've got the image the way you want it, you save and output the file -- to print, to disk, or to DigiDome or PolyDome for further image manipulation.

Now, here comes perhaps the most important point of this presentation. Everyone seems to think because you're dealing with digital and computers and such, in order to get slides made, you need to send your images to a service bureau with a fancy film recorder, or invest thousands of dollars in your own film recorder. You don't.

Instead of directing the computer output to a film recorder, you direct it to... a color inkjet printer.

Here's an Epson Stylus Color 740, cost $99 a couple of years ago. Now there are newer models that have double and triple the resolution, for basically the same price. The point is -- even the inexensive models produce gorgeous output, with quality that is eminently photographable on the copystand. To the audiences sitting 30, 40, 50 feet away from the projected image, the pictures look just fine.

Better, with your inkjet printouts, you're in control. If you don't like the way the printout looks, you tweak it and print out another one. If you spill your coffee on the printout, no problem, you print out another one. When you send the images to a service bureau, if you don't like the way the slide looks when you get it back, you're stuck. Your only solution is to pay to have them try again. We'll discuss costs later on.

A note about resolution. The question is always asked, "how much resolution do I need?" And the answer is, resolution needs to be only as good as the printer is capable of delivering. Forget about the "dots per inch" the printer manufacturer uses to rate it; what we're concerned with is a practical matter -- how good is good enough?

In my testing with the Epson Stylus Color 740, I determined that 150 pixels per inch was about all the printer resolution I required. I tried printing bigger pictures and higher-res pictures -- but after I photographed them, and projected the resulting slides side by side, I couldn't see any improvement. Here's where practicality comes in to play. The bigger you make your printouts, the more paper and ink you use up, and the printouts take longer to make. Higher resolution images take up more storage space on your disks. It takes longer to work with larger file sizes, because Photoshop has that many more pixels to crunch with each operation, and time spent staring at the hourglass cursor is worth something. So if you can't see the difference on the dome, it doesn't make much sense to deal with more pixels than you need. Again, your mileage may vary -- run your own tests and see for yourself.

Printed at a resolution of 150 pixels/inch, our 3:2 image rectangle is 1050 pixels wide x 700 high. It comes out of the printer sized at 7 inches by 4.5 inches. How convenient -- this allows for 2 images to fit on a standard 8.5" by 11" sheet of paper.

In Photoshop, make yourself a full-page template. Remember our grid from above? Why, here it is again. Using it, now we can see in the computer where images will fit in a Wess mount. Clever observers will note they could make their own alignment slides by simply shooting the grid.

So if you want to size Mars so it fits in a Wess 302 circle mount, you can. You can place it in a layer beneath your grid and size it up exactly.

Once you've settled on your standard image size, you're set. Simply make every image for every slide you make that size.

With every image the same size, every printout is the same size. You can use the same copystand height for every image. All of a sudden, everything becomes very simple, very easy. It may seem perhaps complex to set up, but you only go through it once; get the method down, and it becomes routine very quickly.

Okay, we've now got the computer, printer, camera, and copystand all operating with controlled, constant, fixed parameters. What remains is to get all our images in computer-ready, digital form.

Of course, if you downloaded the image from the Internet, it's already digital.

If your source is on paper -- a print, artboard, picture out of a book you have permission to copy, etc. -- then you can use a flatbed scanner like the one pictured here. It isn't essential that your work area be as cluttered as ours is.

Alternatively, you can also simply take a picture of the print with a digital camera, and download that into the computer. If you have a copystand, you've probably already figured that out!

If your original source is a slide, it's possible you can get passable results by simply taking a picture of it sitting on your light table, with your digital camera. Obviously, you'll want to get as close to the image as the camera's focus will allow (use the "macro" lens setting -- or an actual macro lens), and a good flat white source of illumination behind the slide. Experiment.

As always, "use the right tool for the right job" is sage advice. When we got our flatbed scanner, it came with a "transparency adapter". Great, we thought -- until we discovered the adapter was merely a mirror which reflects the image of the slide as the light passes beneath it. You end up with a little thumbnail-size image -- which is barely good enough for a thumbnail image.

You're likely to get better results scanning a slide with a dedicated slide scanner -- what a concept! We use the Minolta Dimáge Scan Dual.

As with all copy work involving slide film, you pick up a major amount of contrast increase -- brights get brighter, darks get darker. Compare the slide images with the "original" Mars Pathfinder scene (yes, we're mixing media here by making JPGs of slides for Web display; you'll just have to cope). There's not much you can do about it, it's the way film works. I've put the raw slide scan below, and then I got to work in Photoshop, tweaking the levels, the saturation and the color balance to try to match the original digital version. Still, Sojourner's wheels in the shadows, for example, tend to get lost in the the murk. Nothing is simple.

Also in slide scanning, you're essentially taking a picture of an illuminated slide, so the chances are good the resulting image will have the usual dirt and dust specks on it -- the perpetual, universal bane of planetarians. You can see it in the sky of the raw scan image, if you click on the image to select the enlargement. Meticulous cleaning is essential -- in the film medium, and again in the digital medium. I "rubber-stamped" out the major artifacts in the sky in the worked image.

Raw slide scan

Digital image

Worked slide scan

Costs

Let's compare and contrast some typical costs for the two methods of outputting computer to slide -- the service bureau/film recorder method, and our inkjet printout on copystand method. The figures I use here reflect real-world prices, the ones I deal with, as of 2001. YOUR MILEAGE MAY VARY, of course.

Both methods have two things in common: the cost of the film and its processing. (Well, three, if you factor in the cost of the Wess mount, but we'll leave that out in our comparisons here.) We're basically talking $.45 per frame to generate a color film chip.

Now let's price some service bureaus. The lowest price film lab we've found that does this sort of work is GammaTech -- they charge planetaria $1 per frame. Since we know that film and processing are constants, that means they're basically charging $.55/frame for the use of their film recorder and equipment. GammaTech is also located in New Mexico, so unless you live in Albuquerque, you'll have additional shipping and delivery charges involved too.

Locally here in New England, the lowest price we found for a service bureau with a pin-registered film recorder is $4 per frame; $5 to $10 is more typical of larger commercial film labs. So let's extrapolate using an average price of $5 per frame too.

Let's take a TIF format file in our computer, and say we're happy with the way it looks. It's the right 3:2 proportion, and we need a slide of it. To send it to the service bureau, you either copy the file onto a disk or disc for delivery, or transfer it electronically via the Internet. Using our method, we print it on paper, to photograph on our copystand. In any event, the picture comes out one of the computer's ports -- comm port, printer port, removable drive.

The cost to print out the image is basically a few cents. We use the recommended photo-quality paper (the matte-finish paper, not the glossy photo paper, which would cause reflection problems on our copystand). The paper costs about $10 for a 100 sheet package -- a dime a sheet. But we print our images 2-up on one sheet of paper, so it's really only a nickel per frame. The ink cost will vary based on what the image is, but let's ballpark it at a nickel per frame as well.

So using the LNP printout method, a frame costs us $.10 to print, and the same $.45 for film/processing. Duplicates are even cheaper; once the print out is made, you can click off as many additional exposures as you care to; you're only out the cost of the film stock and processing.

We've only been discussing single chips of film here, but of course planetarium slides usually have at least two chips of film in them -- one color, and a mask to make the background opaque.

Here at Loch Ness Productions, we use Kodalith, an orthographic film, to make our masks, and we process it in-house. Kodalith is roughly half the cost of Ektachrome per chip. Factoring in the costs of the time involved (15-20 minutes) for processing, as well as the chemistry (which we buy in bulk), call it $.25 per frame. The printouts cost much less too, since we can print on plain inkjet paper instead of the fancy matte-finish stock, and we only burn through black ink.

Kodalith works for us -- but service bureaus don't do Kodalith. They'll use the same Ektachrome for your masks as they do for the color, so it will cost the same $1-5 per frame. We should also point out that "black & white" Ektachrome isn't completely opaque in the black areas, or colorless in the clear areas. But Ektachrome masks can work well enough; just not as nicely, cheaply, or efficiently as Kodalith. (Actually, in yet more dinosaur-extinction fashion, it seems that Kodak has stopped making Kodalith film. We're searching for a ready alternative.)

The bottom line: a roll of 36 exposures, plus masks, going the service bureau route: $72 - $360. Using the inkjet printout method: about $25.

More bottom line

If you're just starting out and don't have any of the equipment above, it is true you will be out hundreds, if not thousands, of dollars when buying a slide making system from scratch. Our pin-registered camera did indeed cost nearly $3000 ten years ago, and the copystand was another grand. Together they're still less expensive than a film recorder, though. If you can afford one, go for it. We wouldn't mind having one ourselves. But film recorders are still output devices; you'll need input devices too, for getting images into the computer.