geckzilla wrote:Cosmic rays are transient so it is very easy to take several exposures and eliminate them completely. They only limit visibility if there's not enough exposures to eliminate them.
Exactly. So the ability to eliminate them depends on other decisions, which typically involve compromise. And the S/N is reduced in every part of the image where a cosmic ray has been removed.
They're not like noise at all.
Actually, they do represent an actual noise source, in the true mathematical sense. They are very like sky background noise. Consider some extreme example, where you need a very long exposure to detect a faint object. If the exposure requirement were so long that every pixel the object occupies were saturated by cosmic rays, that would certainly represent the loss of information to noise.
I don't think things really get lost in the noise.
That's exactly what they get lost in. The ultimate limit of detection is determined by the S/N.
Bigger telescopes would definitely have an advantage.
Ignoring the matter of optical resolution, bigger telescopes mean more photons in a given exposure time, which means improved S/N in that same time. In principle, there's no difference between using a larger aperture and using a longer exposure time. In practice, however, there are some noise sources that are a function of time- dark current noise and cosmic rays being two of them.
[quote="geckzilla"]Cosmic rays are transient so it is very easy to take several exposures and eliminate them completely. They only limit visibility if there's not enough exposures to eliminate them.[/quote]
Exactly. So the ability to eliminate them depends on other decisions, which typically involve compromise. And the S/N is reduced in every part of the image where a cosmic ray has been removed.
[quote]They're not like noise at all.[/quote]
Actually, they do represent an actual noise source, in the true mathematical sense. They are very like sky background noise. Consider some extreme example, where you need a very long exposure to detect a faint object. If the exposure requirement were so long that every pixel the object occupies were saturated by cosmic rays, that would certainly represent the loss of information to noise.
[quote]I don't think things really get lost in the noise.[/quote]
That's exactly what they get lost in. The ultimate limit of detection is determined by the S/N.
[quote]Bigger telescopes would definitely have an advantage.[/quote]
Ignoring the matter of optical resolution, bigger telescopes mean more photons in a given exposure time, which means improved S/N in that same time. In principle, there's no difference between using a larger aperture and using a longer exposure time. In practice, however, there are some noise sources that are a function of time- dark current noise and cosmic rays being two of them.