by Nitpicker » Tue Oct 21, 2014 9:01 pm
Chris Peterson wrote:Nitpicker wrote:Having just noticed that the stacking software I typically use (Deep Sky Stacker) calculates the FWHM for me in pixels, I did a quick comparison between my reduced image and the APOD (which I assume has also been reduced) and was surprised that they both had similar FWHM values just under 7 arcsec. And if I check my full resolution images at 0.6 arcsec/pixel, raw and stacked/stretched, things improve (as I expected, though not linearly) to between 4 to 5 arcsec. But I'm struggling to relate these values with the FWHM values typically quoted for astronomical seeing, where 3 arcsec is rather average and anything above that is worse. I considered the seeing to be better than average on the 19th when I took my images. Might the discrepancy in the numbers be mainly due to tracking errors, or something else I haven't considered? Edit: perhaps the low elevation angles too.
And you're using a DSLR, right? So you're going to see a loss of resolution from the color sensor. And depending on the model, there may be internal processing (Nikon cameras, for instance, aren't great for astroimaging). And the raw images from a DSLR typically don't have the same dynamic range as from a good astronomical camera, which can mess up FWHM measurements somewhat. And different FWHM algorithms give different results, especially if the data isn't linear, which is often the case with DSLRs. I'd say there's a good chance that a measured value of 4 arcsec in a DSLR raw equates to what I'd call 3 arcsec seeing.
A good test of the quality of the image is to measure the FWHM of multiple stars. They should all be the same to within a few tenths. If you're seeing more than that, it means there are nonlinearities in the underlying data.
Good advice. I'm using a Nikon D5100. I've read about older Nikon DSLRs which had something of a reputation as "star crunchers" (relating to how the Nikon converted the raw image to its own NEF raw format, I think). That reputation seems difficult to shake, but the newer models have been improved with a better conversion to a 14 bit NEF format, which is not a problem with my not-so-deep astrophotos, anyway. Certainly, the software and controls available for astro imaging with Canon cameras leaves Nikon for dead, but other than that I'm very happy with the D5100. (We had the Nikon before I ever developed an interest in astronomy. I've also bought an inexpensive planetary imaging video camera since, which is a pretty good match for my modest 6" scope.)
[quote="Chris Peterson"][quote="Nitpicker"]Having just noticed that the stacking software I typically use (Deep Sky Stacker) calculates the FWHM for me in pixels, I did a quick comparison between my reduced image and the APOD (which I assume has also been reduced) and was surprised that they both had similar FWHM values just under 7 arcsec. And if I check my full resolution images at 0.6 arcsec/pixel, raw and stacked/stretched, things improve (as I expected, though not linearly) to between 4 to 5 arcsec. But I'm struggling to relate these values with the FWHM values typically quoted for astronomical seeing, where 3 arcsec is rather average and anything above that is worse. I considered the seeing to be better than average on the 19th when I took my images. Might the discrepancy in the numbers be mainly due to tracking errors, or something else I haven't considered? [i]Edit: perhaps the low elevation angles too.[/i][/quote]
And you're using a DSLR, right? So you're going to see a loss of resolution from the color sensor. And depending on the model, there may be internal processing (Nikon cameras, for instance, aren't great for astroimaging). And the raw images from a DSLR typically don't have the same dynamic range as from a good astronomical camera, which can mess up FWHM measurements somewhat. And different FWHM algorithms give different results, especially if the data isn't linear, which is often the case with DSLRs. I'd say there's a good chance that a measured value of 4 arcsec in a DSLR raw equates to what I'd call 3 arcsec seeing.
A good test of the quality of the image is to measure the FWHM of multiple stars. They should all be the same to within a few tenths. If you're seeing more than that, it means there are nonlinearities in the underlying data.[/quote]
Good advice. I'm using a Nikon D5100. I've read about older Nikon DSLRs which had something of a reputation as "star crunchers" (relating to how the Nikon converted the raw image to its own NEF raw format, I think). That reputation seems difficult to shake, but the newer models have been improved with a better conversion to a 14 bit NEF format, which is not a problem with my not-so-deep astrophotos, anyway. Certainly, the software and controls available for astro imaging with Canon cameras leaves Nikon for dead, but other than that I'm very happy with the D5100. (We had the Nikon before I ever developed an interest in astronomy. I've also bought an inexpensive planetary imaging video camera since, which is a pretty good match for my modest 6" scope.)