I am looking for articles/information that will allow astrographers to choose a matching CCD chip size for different types of astrophotography. (Deep sky, planetary, solar, etc)
I've seen articles that has difficult to understand calculations that will indicate a good chip/pixel size based on the type telescopes you use.
Also good to have information that explains why the need to have different CCD/pixel size. Basically to answer questions why CCD camera manufacturers have big range of cameras for astrophography.
Thanks.
CCD chip size, pixel size articles for Dummies
CCD chip size, pixel size articles for Dummies
Wee Nghee the Pooh
yes. have come across many articles describing this as well, but very often, i end up getting more confused at the end. nevertheless this is what i understand from the concept of matching chip size (this can extend to film size as well, but we will discuss digital here).
However, first and fore most, the factors that influence imaging includes the following:
1) Pixel size of the detector (in the order of micrometers,um)
2) Focal Length of the optical instrument (mm)
3) "Seeing" resolution (in arc seconds)
Let's elaborate on the above factors before going into the equations that derive useful entities.
1) Pixel size refers to the dimension of its pixel on the detector surface. Normally, it is of the order 6.45um, 7.4um or even larger, like on the SBIG ST8 which is 9um. Nikon D70 has a pixel size of 7.8um. The larger the pixel size or area of each pixel, the more photons it can collect per given time - and hence the more sensitive it is. Here we define sensitivity as photon collection rate. We would choose large detector pixel size if we want to collect data from dim objects like galaxies. We would choose small pixel size detectors if we are imaging bright objects like Jupiter, but yearn for better resolution.
2) Focal length of the optical instrument is a factor here as well, but it is bounded by the local seeing conditions. Lots of beginners would think that it does not matter, and probably already have one before starting out on choosing a suitable imaging detector, but in practice it does. We will see why later.
3) Seeing conditions is actually the most important factor here. If we have poor seeing resolution, no matter how high resolution your instrument setup is capable of, there will be a bottleneck in getting high res results if you are imaging from a location that yields poor seeing resolution. Here is Singapore, the seeing resolution is typically 4-5 arc seconds. Mersing can sometimes be better, up to 2.5arc seconds. Basically the lower the number, the better it is, in the sense you get more resolution out of your pictures.
Now how do we put everything together?
Let's introduce the sampling rate of the detector system which is dependent on the focal length and the pixel size. This will determine the relation between a pixel and the diameter of sky coverage in arc-seconds.
Sampling in arc-seconds = 206.265 x PixelSize / Focal Length
Let us take a detector pixel size of 7.4um which is a "general" size good for shooting nebulae, and yet sensitive enough for galaxies. We will also assume a focal length of 500mm for now.
Hence Sampling would equate to 206.265 X 7.4 / 500 = 3.05 arc-seconds per pixel
Which means each pixel of this 7.4um detector will cover 3.05" of sky or rounded off to 3".
From Nyquist Sampling Theorem, under theoretical conditions a star is required to be fitted into a 2x2 pixel square configuration. This means each star in the above configuration would be 6" in diameter. This is pretty generous as even in Singapore, since this would mean a seeing resolution of 6". If we are imaging using this setup, and if the seeing is (say) 4", we can say that the detector is not "high res" enough to capture all the detail of a given object that the seeing allows. This condition is termed under-sampling. For a seeing resolution of 4", we should get an empirical sampling rate of 2" instead - which means either we get a smaller detector pixel size or we increase the focal length (get another instrument, or employ a barlow/optical extender/tele-converter).
Lert's say we use a 750mm focal length instrument now.
The sampling rate (assuming we use the same detector) is now 2.03" arc-seconds.
For a seeing resolution of 4", this is now ideal and you can get optimum resolution according to your imaging location with your imaging setup.
Personally, my current imaging setup is yielding a sampling rate of 3.05" (see my website), so you can say it is undersampling. Well, it is, looks like i got to look for something in the 700-800mm FL range now!
Hope this helps!
However, first and fore most, the factors that influence imaging includes the following:
1) Pixel size of the detector (in the order of micrometers,um)
2) Focal Length of the optical instrument (mm)
3) "Seeing" resolution (in arc seconds)
Let's elaborate on the above factors before going into the equations that derive useful entities.
1) Pixel size refers to the dimension of its pixel on the detector surface. Normally, it is of the order 6.45um, 7.4um or even larger, like on the SBIG ST8 which is 9um. Nikon D70 has a pixel size of 7.8um. The larger the pixel size or area of each pixel, the more photons it can collect per given time - and hence the more sensitive it is. Here we define sensitivity as photon collection rate. We would choose large detector pixel size if we want to collect data from dim objects like galaxies. We would choose small pixel size detectors if we are imaging bright objects like Jupiter, but yearn for better resolution.
2) Focal length of the optical instrument is a factor here as well, but it is bounded by the local seeing conditions. Lots of beginners would think that it does not matter, and probably already have one before starting out on choosing a suitable imaging detector, but in practice it does. We will see why later.
3) Seeing conditions is actually the most important factor here. If we have poor seeing resolution, no matter how high resolution your instrument setup is capable of, there will be a bottleneck in getting high res results if you are imaging from a location that yields poor seeing resolution. Here is Singapore, the seeing resolution is typically 4-5 arc seconds. Mersing can sometimes be better, up to 2.5arc seconds. Basically the lower the number, the better it is, in the sense you get more resolution out of your pictures.
Now how do we put everything together?
Let's introduce the sampling rate of the detector system which is dependent on the focal length and the pixel size. This will determine the relation between a pixel and the diameter of sky coverage in arc-seconds.
Sampling in arc-seconds = 206.265 x PixelSize / Focal Length
Let us take a detector pixel size of 7.4um which is a "general" size good for shooting nebulae, and yet sensitive enough for galaxies. We will also assume a focal length of 500mm for now.
Hence Sampling would equate to 206.265 X 7.4 / 500 = 3.05 arc-seconds per pixel
Which means each pixel of this 7.4um detector will cover 3.05" of sky or rounded off to 3".
From Nyquist Sampling Theorem, under theoretical conditions a star is required to be fitted into a 2x2 pixel square configuration. This means each star in the above configuration would be 6" in diameter. This is pretty generous as even in Singapore, since this would mean a seeing resolution of 6". If we are imaging using this setup, and if the seeing is (say) 4", we can say that the detector is not "high res" enough to capture all the detail of a given object that the seeing allows. This condition is termed under-sampling. For a seeing resolution of 4", we should get an empirical sampling rate of 2" instead - which means either we get a smaller detector pixel size or we increase the focal length (get another instrument, or employ a barlow/optical extender/tele-converter).
Lert's say we use a 750mm focal length instrument now.
The sampling rate (assuming we use the same detector) is now 2.03" arc-seconds.
For a seeing resolution of 4", this is now ideal and you can get optimum resolution according to your imaging location with your imaging setup.
Personally, my current imaging setup is yielding a sampling rate of 3.05" (see my website), so you can say it is undersampling. Well, it is, looks like i got to look for something in the 700-800mm FL range now!
Hope this helps!
