by Chris Peterson » Fri Jul 17, 2009 1:59 pm
rstevenson wrote:I recall reading years ago that you need roughly 4.5 megapixels in a digital camera to duplicate the number of grains of silver in the emulsion of a reasonably good film.
You need to distinguish between resolution and information content. When you make an astronomical image, there is a maximum resolution which is determined by your optics and the effects of the atmosphere. As long as your pixels are close enough together that their spacing isn't what limits resolution, that's all you need. This condition is generally easy to satisfy. The number of pixels is then equivalent to the size of film you might have once used, and matters in terms of how big an object you can image. With a few exceptions, astronomical targets are small. There are many cases where a few thousand pixels is all you need to get the highest possible resolution image of a target. In general, a 3.1 MP camera will be more that satisfactory for astroimaging.
As a previous poster mentioned, you're going to get false pixels in the dark areas. This is because the camera simply makes up something when it can't get enough light on the CCD to record any image data.
All sensors, including film, add noise. Some of it comes from simply reading out the data, and is found in all cameras. Some comes from thermal effects, and is also found in all cameras. Cameras designed for astroimaging are cooled to minimize this; other cameras are restricted to short exposures to limit how much this noise accumulates. With a well exposed image, most noise comes from photon statistics. Technological improvements have allowed camera designers to reduce some instrumental noise, but there's still no assurance that a newer camera will perform better than the one being considered here. FWIW, even the worst digital camera is many times more sensitive than the best film emulsions.
A point-and-shoot digital camera attached to binoculars will probably take good pictures of the Moon. With a white light filter it could be used on the Sun. It will capture Venus, Jupiter, and Saturn, but with less magnification than most would desire. That's it, however. Anything more will require some sort of tracking. It is the need to collect more photons that drives larger apertures and fine mounts. The camera is secondary to that.
[quote="rstevenson"]I recall reading years ago that you need roughly 4.5 megapixels in a digital camera to duplicate the number of grains of silver in the emulsion of a reasonably good film.[/quote]
You need to distinguish between resolution and information content. When you make an astronomical image, there is a maximum resolution which is determined by your optics and the effects of the atmosphere. As long as your pixels are close enough together that their spacing isn't what limits resolution, that's all you need. This condition is generally easy to satisfy. The number of pixels is then equivalent to the size of film you might have once used, and matters in terms of how big an object you can image. With a few exceptions, astronomical targets are small. There are many cases where a few thousand pixels is all you need to get the highest possible resolution image of a target. In general, a 3.1 MP camera will be more that satisfactory for astroimaging.
[quote]As a previous poster mentioned, you're going to get false pixels in the dark areas. This is because the camera simply makes up something when it can't get enough light on the CCD to record any image data.[/quote]
All sensors, including film, add noise. Some of it comes from simply reading out the data, and is found in all cameras. Some comes from thermal effects, and is also found in all cameras. Cameras designed for astroimaging are cooled to minimize this; other cameras are restricted to short exposures to limit how much this noise accumulates. With a well exposed image, most noise comes from photon statistics. Technological improvements have allowed camera designers to reduce some instrumental noise, but there's still no assurance that a newer camera will perform better than the one being considered here. FWIW, even the worst digital camera is many times more sensitive than the best film emulsions.
A point-and-shoot digital camera attached to binoculars will probably take good pictures of the Moon. With a white light filter it could be used on the Sun. It will capture Venus, Jupiter, and Saturn, but with less magnification than most would desire. That's it, however. Anything more will require some sort of tracking. It is the need to collect more photons that drives larger apertures and fine mounts. The camera is secondary to that.