Amateur Astrophotography

Skywatcher Skymax Maksutov-Cassegrain f = 1500mm / D=127mm F12 SynScan AZ GOTO
Skywatcher Skyhawk Newtonian f= 500mm (450mm) / D=114mm F4
Skywatcher Explorer 130PDS Newtonian, f=650mm / D=130mm F5 with Skywatcher Mount EQ5 Pro SynScan GoTo


2. Choosing a Camera

Maksutov

The idea is to replace the eyepiece by a sensor. This means that the image that we get at the focus of the scope will get projected directly on a sensor. Of course to see the image you need a computer and a USB cable usually provided together with the camera. This sounds simple and easy to do, just replace the eyepiece with the gadget you can see in the left image. And the results will be outstanding.

If you are able to get a cheap HD webcam, get rid of the optics, and attach it to the diagonal prism of the telescope, for instance with rubber bands, you will have something to start playing with. You can take the first impressive pictures of small regions of the moon that way.

When getting rid of the optics you will realize that right behind the optics, close to the sensor, there is a glass. That glass is a near infrared filter. All sensors are very sensitive to near infrared, but the optics is not designed for infrared so the glass absorbs it to avoid images becoming blurry. This means that whatever camera you use you are going to need an infrared filter. There are exceptions. People working in deep space photography, however, may love the near infrared into the sensor, because hot hydrogen emits a lot of it, and this makes it possible to take pictures of some nebulae and other objects otherwise invisible.

There are many cameras in the market. Some are cheap and not so reliable, others are expensive but overpriced. And there many sensors... Choosing the right camera is something that has to be done taking into account the characteristics of the telescope. And also we must take into account the characteristics of the available sensors. Based on those both factors we will choose the right sensor. Then we choose among the cameras having the right sensor. This took me some time, because there is more than once possible choice.

The first thing I decided is to buy a camera with a SONY sensor. Reliability, high quality, small pixels and highest sensitivity are the main reasons for such choice. Then, what pixel size do we need?

Looking at the characteristics of the telescope, the maker says that the maximum angular resolution of the telescope is 0.91". This means that two stars can be separated if they have such minimum angular distance. In a way this defines the maximum size a pixel could have. Taking into account a focal length of 1500mm, using simple trigonometry we get a pixel size of 6.6 microns. Using this pixel size two stars separated by 0.91" seconds of arc will project their light in two consecutive pixels. That's a very close and very critical separation, difficult to work with, since the image from every star is a diffraction pattern made of rings.

Perhaps easier to see the separation if we get a pixel size about half that number, or even smaller. With 3.3 microns the two stars are now separated by one pixel. This is still quite critical but better. Clearly, we want a sensor with a pixel size of 3.3 microns or even better smaller. There is another way to make things less critical. If use a Barlow lens, with a 2X, separation of the objects become double. Then, using a pixel size of about 3.3 microns or smaller, together with a Barlow lens, we can get an optimum system.

Using technical words, the image will be oversampled, but this will be a good way to make sure we are able to see by eye, watching at the image, every possible detail. We must take into account that we will have so many different sources of "noise" that will add unwanted information to the image. Electronic noise from the camera itself, atmosphere turbulances, no perfect focusing, etc. Redundance originated from oversampling is also a way to make sure we will be able to eliminate to a good extent the noise (by averaging surrounding pixels for instance) and will make it possible to recover the original image by a complicated process of image processing. There is simple enough free software that carry out the complicated processing in quite a straightforward way, so we do not have to worry about that now.

If we check Sony sensors with a pixel size of about 3 microns, and take into account camera prices containing them, I finally got two choices:

IMX290 2.9 x 2.9 microns 1945×1097 2.13M 5.64mm x 3.18mm;
final image of 1920x1080 at a maximum speed 126FPS

IMX224 3.75 x 3.75 microns 1305x977 1.27M 4.89mm x 3.66mm;
final image of 1280x960 at 150FPS (1x1) or 640x480: 240FPS (2x2)

The sensor Sony IMX290 is used in the cameras Svbony SV305 and SV305 pro. The SV305 is quite inexpensive but not implemented at full speed for the full resolution but I did consider it seriously. It has full Linux support. I was about to buy it since I am a Linux supporter. If I had to buy a camera for a young relative the Svbony305 would be my choice. You can get it easily in Amazon or Aliexpress.

There is a quite unknown but good maker of cameras called Risingtech. They have the Sony IMX290 in the G3M290 model implemented at full speed. The sensor IMX224 is used in the RisingTech camera G3M224C. My final decision, based on many factors, was the RisingTech camera model G3M290. They advertise Linux support but it's not so good. In windows support is excellent but I hate using Windows. Anyway, it is a good camera and I am enjoying it very much.

Taking into account that everything is made in China, you may get better prices in Aliexpress. But you may have to take into account taxes when crossing your country's border.

Let us review the final decision (G3M290 or Svbony 305, same sensor). Pixels of 2.9 x 2.9 microns and a telescope with a maximum resolution of 6.6 microns on the focal plane. This means little more than one pixel of separation between two stars separated by an angular distance of 0.91". Using a 2X Barlow lens, the separation becomes 6.6 x 2 = 13.2 microns. This means a separation of little more than 3 pixels between the stars. Not much but enough. The problem of going further with a 3X Barlow lens is that the image gets darker (same light energy spread in a bigger area), and that's really bad for the kind of processing we are interested in. We want to be able to record many frames per secons, as much as possible. So there is a trade off between camera speed (better if smaller image) and resolution (better if bigger image). The 2X Barlow is a good compromise.

Let's check which field can we cover with the sensor. Sensor size is 5.64mm x 3.18mm, which means a field of about 12.9' X 7.2' without Barlow. If we use a 2X Barlow lens, the field will be half: 6.5' X 3.6'. As you can see we are talking about small fields, but planets are very small.

To start working, it is a good decision to use the stock Barlow lens, 2X, provided by Skywatcher. Again, even being light and cheap and simple performs amazingly well. Later on, after some experience and observing the images, we may decide, or not, to replace it with an expensive (about 100 euro) 3 lens Barlow (again, my personal opinion, waste of money!).

How big are the planets going to look like with a 2X Barlow (which diameter in pixels)? Using some simple trigonometry:

Mars    has a size of   25"    ----->   125 pixels
Jupiter   has a size of   45"    ----->   225 pixels
Saturn   has a size of   21"   ----->   105 pixels

How accurate are those numbers? I tried two different 2X Barlows. The stock one, provided an image of Jupiter with a diameter of 205 pixels. With a branded Barlow (Celestron), heavier (3 lenses) and more expensive (did not provide far more resolution) Jupiter had a size of 233 pixels. Small variations from different eyepieces or Barlows from different brands are normal. The average of 205 and 233 is about 220, so this probably means the calculations, so far, are accurate.

Before next section it could be good for you to check or carefully read the following link https://www.astropix.com/html/i_astrop/Planetary_Imaging.html.