Return to Home Page
(fwd) Re: [vsnet-chat 6004] Re: CCD-V Vs visual observations From: "Richard Berry" <rberry@wvi.com> Subject: Re: [vsnet-chat 6004] Re: CCD-V Vs visual observations Date: Sun, 16 Feb 2003 12:20:43 -0800 Hello Taichi-- > Long focal-length instrument is not adequate for this purpose, as you > can see. In practice with short-focus lens with a CCD, the accuracy > is known to strongly depend on the combination and the condition. > This is probably because the stellar PSF can become comparable or smaller > than the CCD pixel size, which makes precision photometry very difficult > or sometimes impossible. I know similar experiments which could only > yielded a 0.1 mag precision. I wonder if there is a well-established > setup for wide-field precision CCD photometry. Yes, about 15 years ago I found very large scatter when I used a TC211 camera with short focus optics. The scatter was about 0.1 mags. Because of this, I was not interested in CCD photometry for the next five years. I recently made some long images sequences to examine making images with short-focus lenses and CCDs. This morning I ran some tests on the photometric properties of the star images in these sequences. The images were made with an Starlight Xpress MX716 camera. This is a CCD with microlenses and interline readout, so it is probably an example of the worst case for photometry. The lens is 25 mm focal length f/1.4 video lens. It is sharp in the center, but gives bad images in the corners. I measured the star image sizes as the sigma of an assumed Gaussian profile. The star image FWHM = 2.34 * sigma(star). The integration time was 60 seconds. The tracking mount allows the star images to drift about 1 pixel in 15 minutes at this image scale. The first image sequence was 100 images of central Orion. I selected three stars with good images (sigma(star)) =0.66 pixel. for these images: sigma(V-C) = 0.039 mag sigma(K-c) = 0.070 mag Then I selected three stars with moderately aberrated images, with sigma(star) = 0.91. The photometry: sigma(V-C) = 0.026 mag sigma(K-c) = 0.051 mag Next I selected three stars with very aberrated images, with sigma(star) = 1.02. The photometry: sigma(V-C) = 0.037 mag sigma(K-c) = 0.074 mag (This particular pair of stars showed a large periodic fluctuation as the star image drfited across the pixel structure of the CCD, presumably because they were "in phase" with the structure.) The second image sequence was 60 images centered on M31. I selected the nucleus of M31 as C, M32 as V, and a nearby star as K. sigma(M32) = 1.62, so the image of this small galaxy was very soft. sigma(V-C) = 0.026 mag sigma(K-c) = 0.021 mag There is a reasonably clear trend toward lower photometric scatter when the image is large and soft. The 25 mm lens is not optimal -- there are too many stars crowded too close togehter in the field. A lens with a focal length of ~100 mm stopped to 10 mm and set out of focus so the star images would be 4 to 5 pixels diameter might prove to be very effective for photometry of bright stars. This is an easy experiment to try the next clear night. --Richard
Return to the Daisaku Nogami
vsnet-adm@kusastro.kyoto-u.ac.jp
