Lunar Imaging Basics - Part 2
In part I we addressed some of the biggest issues that make your raw images look poor compared to the experts' finished images: (a) digital noise from the camera, (b) seeing issues, and (c) optical flaws like dust. Now we'll explore a short practical example of going from raw images to a finished moon photo.
Let me describe the setup I use. I have a Philips Toucam 740 connected to a Celestron NexStar 5. The camera is attached to a Mogg 1.25" adapter and with an IR filter on the adapter. The neighborhood homes and driveways tend to produce a lot of wobbliness that shows up dramatically in solar system objects. There are moments of good seeing, but not a lot.
The software I use to capture the images is a low-priced product called K3CCDTools. This product is one of the most popular for webcam imaging. K3CCDTools stores sets of images in Windows AVI file format - normally a video viewing format.
As I prepare to capture some images, I go through the normal process of getting good focus and adjusting the camera settings for the best image. Different cameras have different settings, but my Toucam software lets me adjust the brightness, contrast, gain, frame rate and shutter speed. Getting these settings right is an entire topic itself so I can't go into all the decision choices. But for most lunar images, I used normal brightness and contrast. I set the shutter speed to 1/60th of a second at 10-15 frames per second. The gain is usually around 25%. This yields an image of the moon with no saturation and a nice range from bright to dark. The graininess due to digital noise is fairly low (if you get a lot of grain, try less "gain" and a longer shutter speed). I use the video capture feature of K3CCDTools to create a 2.5 second long video of the moon. I usually find that I had some serious dust donuts (out of focus dust pieces on the filter or camera). These flaws are obvious when you watch the video because the moon will move due to tracking mistakes or seeing, but the flaws remain constant in the image.
This is the point that frequently discourages novices: They have images, but the images look poor individually. Some are wobbly due to seeing. All might have some digital noise. And there might be dust or other optical flaws. But the experts' raw images might have some of the same problems you see in your raw images. The key is to correct, select and compensate. I take flat field images that I can use later during post-processing to fix the flaws. I simply put a white cloth towel over the objective and aim at the moon at the brightest section I can locate. The white cloth diffused the light to make for (more) even lighting. I take another 2.5 second sequence.
So my session is complete with a set of raw images of the moon and a set of flat field images of a white towel illuminated by the moon. To put this together, I want to use a more powerful computer than my laptop, so I switch to my desktop system. These usually have faster disks, more memory and more CPU power than the typical laptop.
K3CCDTools is a good capture tool and it also has some nice stacking and fair image processing features. I use K3CCDTools to create my master flat field image by stacking all the frames from the flat field video I captured. When you stack flat field images (or dark images or bias images if you're familiar with those concepts) you don't worry about aligning the photos on a specific feature - you just pile up the data to reduce digital noise. So it only takes me a few seconds to build a master flat image using K3CCDTools.
While K3CCDTools has some nice stacking features, especially for planets, I usually prefer to do my stacking with another tool - Registax (free). Registax does a good job with stacking images and it has some other selection and processing features that make it a little more appealing to me.
In Registax, I load my avi file and I tell Registax to use my master flat field to correct the final image. I review all the frames of the avi using Registax's Frame List feature. It lets me scroll through the image and select the images I want to stack. (You can also let Registax do the selections automatically.) Then by clicking on an alignment feature - a mountain or crater rim, the program will carefully align and stack the images. Depending on the power of your computer and the number of frames (and the size of the frames) this can take a while. For small numbers of images (I use about 15 on average for the moon) it will take minute or two on my laptop and much less on the desktop. But larger images and larger numbers of images (100's or 1000's for planetary work) it can take a very long time.
At the end of the stacking, the program allows you to make adjustments to the image (contrast, histogram and wavelet adjustments). I generally use these adjustments lightly and save the picture. Further image processing gets done with PaintShop Pro.
Before I do anything significant in PSP, I resize the image to 200%. This gives all the other image filters a little breathing room for processing. Then, I usually use the unsharp filter to give clearer shadows and edges. The PSP histogram feature lets me adjust the darkness, lightness and other brightness controls of the image. For lunar images those are the primary image processing features I use. For planetary images, I frequently need to adjust the colors, too. At the end of the process, I resize by 50% to get back to the original image size.
This process - from capture to final processing - isn't trivial and it usually requires multiple programs. (Some programs claim to be able to do it all, but most folks end up using multiple programs still.) So don't be surprised if you find that you need to use a couple or more programs to capture your image, align and stack the results and refine the image to its best appearance. If you apply the same techniques used by the experts (flat fields, aligning and stacking, and final processing) you can get similar results. And the good news about lunar imaging is that some of the cheapest cameras ($200 and less) can produce some of the best results.
The moon is a great target to cut your astroimaging teeth on. Once you can reliably capture and process individual lunar images, you're ready to move onto planetary imaging or lunar mosaics. Have fun and get started!Published in the October 2005 issue of the NightTimes