FWHM, seeing, and image scale...
- Hi, all,
I'm imaging at 1.36 asp. I just measured the FWHM in my image as 2.4 arcsecs.
I know the definition of FWHM, but I'm wondering what the above numbers really tell me.
If I understand correctly, FWHM can be affected by seeing, star brightness, tracking, and exposure length. Since an imager can influence all but the seeing, I was wondering if anyone has ever done anything akin to a "reference frame" and maybe some magic software like Ron's CCDCalc where, for a given starfield, pixel size, focal length, (and assuming good focus and tracking), there is a resultant benchmark FWHM, so that anything over that number is a seeing measurement.
Is it possible to do something like this?
- --- "Robert" <robert.hertel@...> wrote:
> I'm imaging at 1.36 asp.FWHM tells you 2 important things: when measured in pixels, FWHM indicates how well the camera is sampling the information delivered by the scope; when measures in arcsec, FWHM indicates the quality of actual resolution.
> I just measured the FWHM in my image as 2.4 arcsecs.
> I'm wondering what the above numbers really tell me.
FWHM measured in pixels tells you how well the camera is capturing the information being delivered by the scope/mount/seeing/etc. The above numbers mean that your FWHM = 1.8 pixels, which is significantly undersampled and thus the camera is failing to capture all of the resolution available to it; i.e. information is blurred by the camera because the pixels are too large to differentiate fine details. See:
So if you are interested in resolution then you should consider using a Barlow in order to adjust your image scale so that the pixels are not so large (angular size).
> If I understand correctly, FWHM can be affected by seeing,Actually PSF (FWHM) is independent of brightness, which is why it is such a useful measurement. And exp time only affects PSF indirectly, in that longer exps have more opportunity to go wrong. As a measurement of resolution, FWHM is primarily affected by the scope quality (aperture and optical integrity), the accuracy of tracking/guiding, and the quality of the seeing.
> star brightness, tracking, and exposure length.
> ... there is a resultant benchmark FWHM, so that anythingImage FWHM is basically the quadratic sum of the blur components. So if the scope/mount/etc. delivers resolution of 2" and the seeing is also 2" then the image FWHM will be near sqrt(2^2+2^2) = 2.8". You can use this formula to back-out the seeing but only if you are confident in the quality and stability of your scope/mount/etc.
> over that number is a seeing measurement.
> Is it possible to do something like this?
If you have a scope/mount/camera that is proven to be capable of genuine high-res and you are well experienced with it then you can develop a knack for assessing the seeing based on FWHM. But this method is vulnerable to various mistakes; e.g. if the scope did not fully cool-down or is slightly out-of-focus then it may produce bad resolution even in good seeing.
I have an exotic system (intensified video) that has been proven to resolve D.S. to better than 1 arcsec FWHM; so if I get 2.5" at the scope then I know the seeing is significantly greater that 2". But very few amateurs have such a system. Most long FL scopes top out near 2 arcsec (even in very good seeing). Most short FL scopes top out near 3-4 arcsec, which pretty much makes them immune to all but the worst seeing.
Because of the complicating factors, it is often not possible to accurately assess the seeing from normal images. To distill seeing effects, remove or pin as many factors as possible. For normal imaging, one of the worst and most variable sources of degradation is the mount and tracking, so take several very short exps (0.1 sec) of bright star clusters and measure the FWHM to get a notion of what the scope is capable of in that particular seeing.
For an accurate assessment of seeing it is advisable to use a seeing monitor.