CMOS vs CCD

[Al McAdam]

I am thinking of purchasing a Sony@550 which has a CMOS sensor and my current Sony has a CCD. What are the pros and cons in DSLR astro? I want the 550 for its live view and it is a 14 megapixel while mine is 10. Comments please Thanks Al

[Greg Marshall]

Well, the thing that most people focus on is the pixel count, but more important than that is the pixel SIZE. Not only is the pixel size closely related to sensitivity and signal to noise ratio, it should also be considered in comparison to the spot size of your telescope. That is, if the smallest spot the optics can produce is 10 microns a camera with 5x5 micron pixels would have around 4 pixels responding to a well focused star. That is called "over sampling" and is much preferable to the opposite. Or put another way, a camera with 20 x 20 micron pixels might have terrific specs, but you won't actually achieve good results if the star image covers only 10 microns - it would be under-sampled and since only part of the photo site is illuminated, the output would be weak. My QSI 583 has 5.4 micron square pixels and under the best conditions I have seen the FWHM get close to 1 pixel, but most of the time seeing makes it more closer to 2 pixels. I think that is about ideal, but suspect that some people would argue for more over-sampling.

[Eric Hughes]

Other than noise levels and dynamic range, are there any other chief "stats" that I should be aware of when buying a camera (CCD or CMOS)?

[Greg Marshall]

>Is any company making a CMOS for astrophotography?

Most of the low-end cameras are CMOS. That is, the lunar/planetary cameras and some of the guide cameras, such as Orion's SSAG.

Except for the Canon 20Da, no DSLR is designed specifically for AP, but they sure get used a lot for AP. More and more DSLR cameras are CMOS-based, including all the popular models from Canon. But, AFAIK, all the better astro cameras are CCD.

[Charles Dunlop]

Is any company making a CMOS for astrophotography?

[Greg Marshall]

It used to be true that CMOS sensors were inferior to CCDs in image quality. For conventional use, I don't think there's much truth in that anymore because CMOS sensors have evolved to make use of their big advantage: They can have image processing functions right on the same chip to compensate for defects. The story is a little different for astro-photography. Perhaps the biggest reason is simply that the processing done for a CMOS sensor is not optimal for astro-photography. Now a CCD-based DSLR is also not optimized for AP, but the "raw" signal doesn't need as much processing.

But that's all theoretical. In practice, differences in performance from one model to another and from one generation to another are probably more important than the difference between CMOS and CCD. Look at each camera's noise levels and dynamic range (dpreview.com is a good source). As long as a camera is functional for AP (has "bulb" shutter setting, USB control, etc.) noise level and dynamic range are almost the only thing that matters.

As for resolution, at this level, higher resolution doesn't buy you anything because you're never going to be able to either focus or track a star well enough that you'd even get close to a single pixel with either of these cameras.