Help:Tutor imaging

From Lynkeos
Jump to: navigation, search
Top ↑Tutorial Tutor_list→

Shooting the planets with a webcam

Languages English - Français


The quality of your final image will be limited by the weakest link in the chain of events that starts with the photons leaving the subject you are imaging, and ends with a print on your wall.

The weakest link is the atmosphere. Unless you bought a space shuttle at an auction you unfortunately have no control over it. The best you can do to limit atmospheric turbulence is to avoid the proximity of buildings and roads, and observe from your lawn.

The second weakest link is your telescope, it hosts a lot of gremlins. To stress the most important points :

  • Your telescope needs to be at ambient temperature. It is important to realize that for large telescopes it may take several hours for these to cool down and reach equilibrium. This is very important for "open tube" reflectors.
  • Your telescope must be nearly perfectly collimated. If you move a reflector, the collimation process should be redone.
  • Refractors, even "apos", are not corrected from chromatic aberration in the infrared. To use these for webcam (CCD) imaging, an IR blocking filter must be used. Otherwise your webcam images will have a strong defocused contribution from the IR part of the light spectrum - webcams see pretty well in the infrared.

The next link is the magnification device you use to form the image (nothing, barlow lens, eyepiece projection, ...). To get all the information you expect from your telescope, the detector shall have a sampling frequency at least twice the highest frequency in the image (Shannon's theorem/Nyquist theorem in the spatial domain). That means that the pixel size must be less than half that of the Airy's disk on the resulting focal plane. In other words, you must use a focal/diameter ratio higher than f/D = 3.44 . p where p is the pixel size of your webcam in micrometers, f is the focal length of the telescope, and D the telescope diameter. Ex : for a ToUcam with 5.6 µm pixels your optical setup must be at least at f/D = 19.

This is not merely theory : I was reluctant to use eyepiece projection for my first tries with a webcam and used a barlow lens with which I attained f/D=9, and the results where encouraging... The day (no ! the night) I used eyepiece projection to obtain f/D=20, I regreted I had wasted so much time gathering bad images ! This must be combined with the fact that the exposure times must remain short enough to freeze the turbulence (1/20 sec at most).

Having the right image scale is in itself not sufficient. The image must also be very precisely focused. The internet is full of techniques to achieve a "perfect focus". I have no preferred one. All I have to say is that it is not easy, and of prime importance

The next link is the webcam. To send large images at high frame rates across a USB® link to a computer, a webcam compresses the images. Most compression algorithms are lossy. These losses are not visible in the form of random noise. Instead they tend to form artifacts (such as darker bands next to bright objects) that are eventually enhanced by processing. To avoid these artifacts, use the lowest frame rates possible (I use 5 Hz) and, if you can afford to increase your focal length, lower frame sizes. With some devices it is possible to select uncompressed image formats - use these if possible. Another option, I'm considering, is to use a FireWire® webcam.

When you master all the issues described above, you will have no troubles capturing great pictures of the planets. On the other hand, if you don't pay attention to these details it is likely that no software will be able to help you make your "perfect" image.

Next