by Chris Peterson » Thu Nov 01, 2012 2:48 pm
flash wrote:In this image (and others too) some stars appear not as points of light but show an extent, as if they are large enough to subtend an angle. I believe that Rigel's diameter is around .7 AU or roughly 11 light minutes. Given that the Witch Head Nebula in the image is estimated to be 50 light years across,and the Nebula spans (say) 10k pixels in the image, that means the image of Rigel ought to be no larger than (say) 2 pixels in diameter. Is the apparent enlargement of Rigel due to image processing artifacts or some physical process during the image capture? I believe Chris has mentioned that bright objects cause saturation of CCD elements. Is that what's happened here? Is there any way for a (naive) viewer of such images to know that the apparent extent of stars is not real?
No star will ever have its diameter resolved in a wide field image. The largest, closest stars subtend angles that are much, much smaller than a single pixel.
Because the light of a star passes through an aperture, it is diffracted. This turns a point source into an extended spot. In the ideal case, this is an Airy diffraction pattern- a central spot surrounded by successively fainter rings. In practice, this pattern is smeared out by atmospheric motion into a big blob of illumination, and the brighter the star, the farther away from the image center you can get and still have enough light to rise above the background noise level. This is why bright stars look bigger in images. The dimmest discernible stars will be smaller than a pixel (but of course, we will see them as exactly one pixel in extent). As stars get brighter, their diffracted light will spill over onto more pixels.
This pattern of bloating is unrelated to saturation, although it's true in many cases that the brighter stars in an image will saturate their central pixels, resulting in several effects which are apparent in today's image- the loss of color information in the saturated area, and an apparent white central disc.
There's a completely different effect that also leads to bloated appearing stars, and that is light scatter in the optics. This usually shows up as halos around brighter stars, and is apparent in this image. Scattering in the optics of the telescope used here should be very small, although front surface scatter is likely if the front lens element isn't kept scrupulously clean. Scatter and internal reflections in the filters, sensor cover glass, and top layer of the CCD itself are very difficult to avoid, and probably all contribute to the halos we see (which are also larger with brighter stars- hundreds of pixels in the case of Rigel).
[quote="flash"]In this image (and others too) some stars appear not as points of light but show an extent, as if they are large enough to subtend an angle. I believe that Rigel's diameter is around .7 AU or roughly 11 light minutes. Given that the Witch Head Nebula in the image is estimated to be 50 light years across,and the Nebula spans (say) 10k pixels in the image, that means the image of Rigel ought to be no larger than (say) 2 pixels in diameter. Is the apparent enlargement of Rigel due to image processing artifacts or some physical process during the image capture? I believe Chris has mentioned that bright objects cause saturation of CCD elements. Is that what's happened here? Is there any way for a (naive) viewer of such images to know that the apparent extent of stars is not real?[/quote]
No star will ever have its diameter resolved in a wide field image. The largest, closest stars subtend angles that are much, much smaller than a single pixel.
[float=left][img]http://upload.wikimedia.org/wikipedia/en/thumb/e/e6/Airy-3d.svg/200px-Airy-3d.svg.png[/img][/float]Because the light of a star passes through an aperture, it is diffracted. This turns a point source into an extended spot. In the ideal case, this is an Airy diffraction pattern- a central spot surrounded by successively fainter rings. In practice, this pattern is smeared out by atmospheric motion into a big blob of illumination, and the brighter the star, the farther away from the image center you can get and still have enough light to rise above the background noise level. This is why bright stars look bigger in images. The dimmest discernible stars will be smaller than a pixel (but of course, we will see them as exactly one pixel in extent). As stars get brighter, their diffracted light will spill over onto more pixels.
This pattern of bloating is unrelated to saturation, although it's true in many cases that the brighter stars in an image will saturate their central pixels, resulting in several effects which are apparent in today's image- the loss of color information in the saturated area, and an apparent white central disc.
There's a completely different effect that also leads to bloated appearing stars, and that is light scatter in the optics. This usually shows up as halos around brighter stars, and is apparent in this image. Scattering in the optics of the telescope used here should be very small, although front surface scatter is likely if the front lens element isn't kept scrupulously clean. Scatter and internal reflections in the filters, sensor cover glass, and top layer of the CCD itself are very difficult to avoid, and probably all contribute to the halos we see (which are also larger with brighter stars- hundreds of pixels in the case of Rigel).