Hopefully other novice amateur astronomers, like me, will find something of interest in reading about my first light experiment at using a Webcam with my LX90 scope.
I had recently finished reading Martin Mobberley’s article on using a Webcam in the February issue of ‘Astronomy Now’ so have been itching give it a go.
For ages my night skies have been blighted with cloud, save one evening, the 12-02-05, of unstable seeing but which at least gave me visibility of Saturn. I was not able to visually focus on the boiling image of Saturn, but at least it presented me with a chance to capture some frames of the ringed planet with which to experiment.
I had bought a ToUcam Pro II and adaptors from Nick Hudson at ‘True Technology’ some months earlier which I plugged into a 2x Barlow and then into the visual back of the LX90. This arrangement gives an f/20 system and reasonable image scale on the little CCD chip in the Webcam. Within the adaptor I also had an IR blocking filter to replace the camera’s own.
I spent 20 odd minutes getting Saturn to appear on the screen of my laptop! This was due to, a) viewing the laptop screen through my patio doors, i.e. focus knob on scope outside and the laptop inside in the warm, meaning I’m running back and forth to improve already bad focus till happy with the image; b) forgetting to at least approximately par focus with an eyepiece in 2x Barlow first! An out of focus doughnut image can be far too big or faint to see on the little CCD chip especially through patio doors.
Anyway, I now had a picture of Saturn dancing around the preview window of Philips Vrecord. The image controls had been set to manual, 640 x 480 and 10 frames per second, as suggested by Martin. I entered a file name, adjusted the gain, so as not to saturate the image then captured some 306 frames of avi footage. Be aware that using a Webcam results large files; my file was 131MB!
Now for my next mistake! I then forgot to enter a new file name before shooting my back-up sequence. Fortunately, I had thought to check file size and view the captured frames via the default viewer before packing things away; otherwise I would have kept on overwriting the original file and been none the wiser!
Okay, I finally made and thus had two sequences to play with in the free download program Registax. Please keep in mind at this point that each frame in these sequences is of really poor quality, as can be seen below, since this was just an experiment.
I did make a dark frame sequence, but was not at that time able to use it correctly within Registax, so proceeded without dark frame subtraction. Also, the magnification of the Barlow lens showed up two dust specks on the resulting images, as you can see. Thus far I’ve not scored too highly on technique, me thinks.
This was my first go with Registax so I left all the settings at their default and went on through the frames selection and limiting sequence. This checks the quality of each frame against the one you choose as the reference image. In all just 282 frames were used to align and stack; which I thought was not bad considering many were just fuzzy blobs with a splash of colour added for good measure.
The last part of the Registax process gives you wavelet filters to play with. These, as Martin said in his article, are selective versions of the Unsharp Mask feature common to Photo processing software packages. I just applied a small amount of the layer-4 slider then saved the result as a tif file, ready for the next experiment, but the stacked image is already looking much better.
My next step in this game was to use AstroArt 3 (AA3) to try and improve the image a bit. Some months ago I tentatively played with a technique called Multiple Unsharp Masking (MUM) or Linked Unsharp Masking (LUM) depending on where you read about it. This is a method for pulling out planetary detail that might otherwise be lost.
To do this, I duplicate the original image twice and then close the original, in case things go wrong. Note here that making a duplicate of a tif file in AA3 results in it making them FITS files, thus the next sequence is done using this format.
The first duplicate, I’ve called Copy1, will be used to process. To the second duplicate apply a Gauss filter with Sigma of 2.0 and call it Mask1. Now duplicate Mask1 and apply a Gauss filter with Sigma of 2.0 to this duplicate and call it Mask2. Next I merge two of images using (2 x Copy1) - (1.1 x Mask1), the resulting effect on Copy1 should show signs of being sharpened. Repeat the previous merge but this time using (2 x Copy1) - (1.1 x Mask2) and yet more of an improvement should be visible and should be saved as the MUM result. You can make new duplicates of the processed image and go round again if it will stand it. How much this improves your image will depend on the original image quality and the values you experiment with for both making the masks and the merge processes. There are various web sites that present the MUM technique, although each use slightly different sequences and values.
Two web sites I’ve utilised are: -
http://www.licha.de/astro_article_multiple_unsharps_mask.ptp and
AstroArt’s home site http://www.msb-astroart.com on their Astroart in LAB #3 page.
After the MUM sequence I had an improved image that now showed off the rings better but sadly also the dust specks!
Time for Photoshop I reckon … so I export the sharpened image as an 8-bit grey-tones tif file, because that’s the only high resolution tif format choice you get in AA3 that also matches the image manipulation requirements of Photoshop 7; at least at my entry level of experience with this package. This is fine for what I want to do next anyway.
The job now is to clean off the dust specks (obviously not good practice but done here as part of the experiment) and brighten the image up a bit and save it ready for re-use in AA3.
This done, I open the clean/bright grey-tone image (now known my Luminance image) in AA3 and also have a copy of the original colour image open at the same time. In AA3 you can perform what’s know as a LRGB Synthesis on the copied colour image, where it combines the colour information from the not so good image with the better quality luminance image. The result is a correctly coloured version of the Luminance image. Now all I need do is crop the image to the size I want and save it as a jpeg or bmp file, as required.
Not too bad considering the messy avi footage I started with in this little experiment? Hopefully, you will give both a Webcam and Registax a try. Even if you don’t go the whole hog with other programs, it’s surprising just what you can achieve with this inexpensive PC add-on. … Clear skies …