A Beginner's Guide to Astro-Videophotography: Part II
By Phil Hoyle
This is the second article on how to take images of the planets and the Moon using a video camera. The first article talked about how to get the images on videotape. If you haven't read it yet, you might want to get the last issue of the "Horizon" and read it now, because this might not make a whole lot of sense without reading that article first. As you read this article, keep one thing in mind; I am a beginner at this. I've read a few things about it, but I haven't been doing it myself for very long. In this article, I will cover how to get the images from your videotape into your computer, and how to process them once they get there.
I know from Dave Gill's E-mails that many of you have, or have access to, a computer. This is the single most expensive investment you'll need in order to process images from videotape. The entire list of what you'll need is this:
1 Video playback device (camera or VCR)
2 Frame grabber
4 Processing software
5 A video tape with the images you want to process
There are many devices on the market that will allow you to transfer video to computer and then process "moving" video on your computer. Most of these are relatively expensive and can do a whole lot more than what we are interested in right now. There is, however, a relatively inexpensive device on the market that will allow you to transfer just one video frame at a time into your computer. This device is called a "Snappy". It is basically an electronic widget that attaches to the parallel port of your computer. You then attach a cord between it and the output port of your video camera or VCR. I have only seen one obscure ad for a similar competing device. For all practical purposes, the "Snappy" is the only thing commercially available that can do what we want at a relatively cheap price. When the Snappy came on the market about five to ten years ago, the retail price was about $200.00. However, in the computer products market, prices tend to fall instead of rise. I got mine last fall from a mail order outfit for about $75.00 plus shipping and handling.
The Snappy is a very simple device and it uses very simple software. In fact, the software is all loaded into a single file on your computer. This is something that I haven't seen since the early days of DOS. I would still recommend that you read the manual though; or at least glance through it. I really don't want to have to repeat it all here. Basically, all I really need to say here is that by clicking on your mouse, the thing takes an image from your tape and makes a bitmap file out of it! It does need a computer that runs Windows 3.1 or higher and just because we'll be working with images, make sure you have plenty of RAM on your computer. My computer has 40 Meg of RAM and that seems to be enough.
The last thing you'll need other than your tape with the images, is software to process your images. Adobe PhotoShop is the most recognized software for processing images today, but compared to a lot of other programs, it's a bit pricey. That's why I don't have it. I'm sure it has all the features we would require, so if you already have it, try it. The two programs I have used are Paint-Shop-Pro version 5.0 and Picture Window. The reason I state version 5.0 for Paint-Shop-Pro is because earlier versions did not support layering, which is required for some of the processing techniques I will discuss. Picture Window also does not support layering, but it does allow us to create a composite image from two separate images. Picture Window is also a very good program for the money, less than $50.00. If you already have another graphics software package, make sure it will allow you to overlay one image on top of the other and then blend the two together. Usually, this is done with layers, but as implied above, Picture Window does it a slightly different way.
Downloading the Image
This might seem a bit too obvious, but the first thing you need to do before downloading images from your videotape to your computer is to watch your videotape! There are a few things you will need to do while watching it before downloading the images. Probably one of the most important things to do while watching the tape is to ask yourself "What would I do differently the next time I take video images of the solar system? Did I use the right eyepiece? Did I have the manual settings on the camera correct? Can I leave other things on the camera on auto?" The goal is to get better at it, and the best feedback is your videotape. However, we also need to find which segments of the video we are going to try to download.
To decide which segments to download, first rewind the tape, all the way to the start of the tape. This is because you need to reset the tape counter so you can queue the parts you want to download. Once the tape counter has been properly reset, you can fast forward back to the area you are interested in. As you watch the tape, record on a sheet of paper the observing notes you spoke into the microphone. Also record the tape count every time the image significantly changes. Here a significant change might be when you changed eyepieces. You'll probably have to stop, pause and rewind frequently while doing this. Watch the tape carefully for times when the focus, image brightness and atmospheric seeing all come together to give you a decent image on the tape. Record what the tape count is for the best images so you can easily queue them up later. You will need a minimum of about two seconds worth of good video to successfully create a good computer image. Keep in mind that these two seconds hold 120 video frames!
Once you have a good set of notes of your tape and you have decided which segments you want to work with, its time to actually download the images. As stated above, you need to connect your camera or VCR's output to the input of the Snappy. Then start up the software. There is a setup page in the software, but it is explained pretty well in the manual, so I won't repeat it here. To download an image, you need to play the tape back and then click on the "snap" button with your mouse. There is usually a one or two second delay between the time you click on the mouse and when the picture is actually taken. After the picture has been "snapped" the software will display it on the screen. The result is usually not quite as good as what you see while watching the tape, but better than what you see on the TV when you put the tape on pause. Snap a few images without saving anything at first to give yourself an idea of what the quality of the images are like. Different frames can vary greatly in the sharpness and over-all quality. Save at least 10 of the best images. These are the ones we will process. Don't be discouraged if the images don't look that great at this point! It's the processing that brings out the details and gets rid of most of the "noise". Noise in this context is electronic junk that was recorded on the tape, similar to "snow" on a TV screen. Save the files using some kind of logical naming convention. I use something like Mars01.bmp, Mars02.bmp etc.
Processing the Images
There are three basic procedures I use to process images. There are probably hundreds of ways to do them and each software package will do them in a slightly different way, but if you get the basics of these three procedures down, you can do a lot to improve your images. These procedures are:
1. Enhancing the contrast
2. Enhancing the sharpness--Unsharp mask
3. Stacking the images
Enhancing the Contrast
One thing is almost certain in any planetary photograph and that is the contrast in the raw image is not going to be what you want. The problem is basically this: if the image is even slightly over-exposed, the background is pure black and the planetary disk is nearly all white. If the image is slightly under-exposed, the background is still black, but the details in the planetary disk are all almost the same shade--dark. Even when the image is "properly" exposed, if there is such a thing, most of the shades of the disk will be close to each other making details hard to see. Enhancing the contrast is basically changing the image to keep the background black, but stretching out the shades in the planetary disk so that details are easier to see. In other words, if the disk is nearly all white from over-exposure, you need to get the parts that are not quite white to be darker while maintaining the brightest parts of the disk as white. In an underexposed image, you will need to brighten the lightest parts while keeping the darkest parts dark.
There are more ways to enhance the contrast in an image than there are graphics programs. Most programs have several ways to do it. Some even have "autofix" methods. The best advice I can give here is to try all of them to see what they do for your image. If your program has a "histogram" function, try that to, even if only to see what it does for your image. Some methods are more "manual" than others. These would involve changing the brightness and contrast by specific percentages. They work very similar to the brightness and contrast controls on your TV set. If your program has a "gamma" function, try that too. This is a very good way to darken or brighten the mid-tones of an image without changing blacks or whites into gray.
Enhancing the Sharpness
The best way to enhance the sharpness of an image is to use an "unsharp mask". This is a method that has been borrowed from darkroom photography. By subtracting an "unfocussed" image from the original image, the result has more sharpness. If you don't like that explanation, try this one; It's magic! At least, it looks like magic when you see what it can do for your images. One caution though; don't over do it! If you do, your images will look very grainy, or speckly.
Use caution with other types of "sharpness" functions. One function I have seen actually adds thin little black lines to the images between areas of high contrast! But please feel free to experiment. And please, if you come across something better than what I am describing here, let me know!
Stacking images is by far the most "magical" of the image processing techniques that I am describing here. It is amazing to see the difference between a "stacked" image and the raw images used to create the stack. If you properly stack ten individual video frames, each frame will contribute to 1/10th of the final image. The result is that features common to all ten frames are enhanced while features in only individual frames, like noise, are reduced. Stacking is also the most time consuming and most dependent on the skill of the person doing it of the three basic enhancement techniques. I will describe how I have done it using Paint-Shop-Pro version 5.0 and Picture Window.
Picture Window has a "composite" function. Using this function you can combine two images into one. This is like taking two overhead transparencies and laying one on top of the other. They can be blended together, one used as a filter for another, or subtracted for example. There is also a "transparency" slider that will allow you to change the transparency of the top image. At 100 percent, the top image is opaque, and at zero percent, the top image is totally clear. The procedures I use to stack two images are:
1. Load the two images you want to stack.
2. Under the Transformation menu, select composite.
3. Select the preview button. This gives a preview of what the stacked image looks like.
4. Now, here's the trick. Set the transparency to 100 percent and the operation to Negative. Under negative, when two pixels with the same color are aligned, the result is black.
5. Set the alignment to one point.
6. Now move the alignment cursors to some logical point on both images where it will be relatively easy to pick the same point for both.
7. Check the auto box in the composite window. This gives you a fresh preview after every change in alignment.
8. Now carefully move one of the alignment cursors on one of the images while watching the preview window. Make the preview window as black as possible. If the images are completely black, not only are the two images perfectly aligned, but they are also identical images, so you won't be able to make the images totally black. Try to make any bright areas symmetrical around the planetary disk.
9. Once you are happy with the alignment, go back to the composite window, set the operation to blend and the transparency to 50 percent.
10. Now click OK to create a new image which is a composite of the first two. Then save the new image.
This procedure can be repeated for several video frames. I usually use sixteen frames. First I create eight composites of two frames each. Then I create four composites from the eight I made previously. I continue this until I have made a single composite from 16 video frames.
The procedure for stacking images in Paint-Shop-Pro uses basically the same techniques, except we use different layers instead of two individual images. Also, it is possible to stack multiple layers on top of each other to create a single image of ten layers. To get the layers properly aligned, set the transparency to 100% and set the layer type to negative. Then, using the mouse, drag the top layer over the others until it is aligned. Once it is aligned, set the layer type back to normal and change the transparency. The transparency should be set to 50 percent for the first overlay (2nd layer), 33 percent for the second overlay (3rd layer) and so on. The transparency should be set to 1 divided by the layer number. When all the layers are in the image and aligned, you can condense them into a single layer.
As I said before, I'm new at this. At this point, I haven't decided if it is better to enhance the contrast and sharpness before or after stacking the images. However, if enhancing the contrast or sharpness is done before stacking the images, I feel that it is important to process every frame the same. You can also try to enhance the contrast and sharpness on just the resulting image after they are stacked. As you work with your images, experiment! Try different things to learn how your software works. Learn different techniques for manipulating your images.
These image processing techniques can be used with other types of images as well. CCD cameras do an excellent job of recording planetary as well as deep space objects and these techniques can be used for them. If you take astro-photos the traditional way, with film, you can still use these techniques. You just have to scan them into your computer first. There are a lot of people enhancing film photographs this way too.
Processing images can be time consuming, especially if you do a lot of experimenting. It is a good activity to do when the astronomy bug is itching and the weather is not cooperating. Also, with the advent of the internet, electronic images are easy to share with the rest of the world, so you might even get some recognition for your work! So when you can't get under the stars, get in front of your computer and get busy!
Click on the images below to see an enlarged view.
- Identification of the film, camera, lens, etc., that was used
Commercial 8 mm video camera hand held over eyepiece.
- Notation of special equipment used.
10" LX200, commercial 8 mm video camera
Back to Tutorials Index