Home / Astrophotography / Cameras, Observatory and Processing

I use several different cameras:

A Connectix QuickCam - original model, bought in 1997

A Creative Webcam II - color camera, bought in 1999

A QuickCam VC - colour camera_bought in 2001

A Starlight MX516 - no adaptation needed for this one! 

A QuickCam Pro 4000 - colour camera bough in 2002

QCPro4000 & Mx516                               QCVC, QCPro400, Mx516 & Skysensor

---

 

My images are processed like this: 

Before going into image processing one must address the subject of image acquisition. Processing an image can only make visible the data contained in the original. The trick is to record the maximum amount of data in each raw image, and not the usage of software and filtering algorithms, although they are of great importance.

Basically you must deal with tree factors: Atmospheric turbulence, the telescope’s maximum resolution and light gathering power and the camera’s image sampling, light sensibility and noise level.

You cannot easily change the factors themselves, only their interplay. This is achieved by finding the right f/ratio to obtain your images, generally by means of using a Barlow lens. The objective isn't getting the aesthetically best image, but the one that records the maximum details possible.

Small magnifications will allow shorter exposure values and noise free images, but the amount of spatial information obtained is low, as the object will be only a few pixels across.

High magnifications demand longer exposures, making the effects of turbulence more visible (a way to shorten the exposures is to use a higher electronic gain level, but the image’s noise level will then rise). Nevertheless, the planet will cover more pixels, and all the spatial information will be recorded.

These two situations will determine the path to follow regarding image processing.
In the case of undersampling, meaning small planetary images that fail to record all the spatial detail possible, a special procedure named superresolution should be employed.
The second case is oversampling, where the planet is large enough to allow all of the details to be recorded. Special image processing algorithms such as Maximum Entropy deconvolution or Wavelet Filters should be used in this case.

Regardless of the situation your first work will be to average the best images obtained in each session together. All images blurred by atmospheric turbulence or out of focus should be discarded. This will remove the random noise present, allowing you to perform subsequent image processing on a "clean" image. Software such as Registax, K3CCD Tools or Iris are able to choose, align and stack the best frames automatically.

This is the basic procedure to follow in processing planetary images. I’ll just add some common operations that are crucial and performed many times. There is no recipe to their use, so experimentation and practice are advised.