Posted by
karenacollins on
Apr 26, 2018; 5:19am
URL: http://astroimagej.170.s1.nabble.com/Variable-apertures-tp989p991.html
Hi Craig,
Answers embedded:
On 4/25/2018 5:46 AM, CraigNZ [via AstroImageJ] wrote:
Some questions about the variable aperture feature.
1. Is the variable aperture size constant, for all stars on a given image, of a series? In other words, is the aperture size = FWHM_mean x scaling factor.
Yes, all apertures within an image have the same size aperture. The variation is from image-to-image.
2. In the measurements report, column Width_xx, is this the diameter of the FWHM? As I understand, FWHM is a radius and I guess width is then the diameter (FWHD)?
Width_xx is an estimate of FWHM derived from the center-of-light centroiding algorithm. It may not be an accurate representation of FWHM, depending on the level of defocus, but it scales well from image to image and works well for detrending photometric inaccuracies
due to changes in seeing from image to image.
3. Using data from a recent measurement report which used 1 target and 7 comparisons across 15 images, image 1 I have:
Width_T1: 8.06 (I assume this is the FWHM diameter in pixels)
It is an estimate of FWHM, but as can be seen here, is apparently off by a factor of about 2 for this series of images.
FWHM_T1: 4.05 (in pixels)
This should be an accurate measurement of FWHM.
FWHM_Mean: 4.18 (average across all comparison stars)
Yes
Source_radius: 20.70 (?)
This is the actual radius of the photometric apertures in this images (in pixels)
Source_radius(base): 25 (a radius in pixels, so FWHD = 50 pixels)
This is the radius you have set for the photometric aperture. It is important that this value be set large enough to properly measure the centroid and to estimate Width_T1. In general, set this larger than the largest scaled aperture that you expect. Since
you are getting Source_radius = 20.70, I expect Source_radius(base) = 25 is a good setting.
FWHM_mult: 1.4
I think I read Source_radius = FWHM_xx x FWHM_mult. But using the data above I don't think this is right. Assuming these are all in pixels (except for the multiplier), shouldn't Source_radius = 4.18 x 1.4 = 5.85?
I'm hoping you can follow my pseudo-code here: in the current version of AIJ the variable aperture radius = FWHM_mult * MaxOf( Mean(X-Width_xx), Mean(Y-Width_xx) ). In a future version, AIJ will offer the option to use variable aperture radius = FWHM_mult *
FWHM_mean, but in some testing I ran a while back, for some reason the latter version does not seem to work as well as the current version (thus my slowness in introducing the option).
4. Is the Sky_Rad(min) and (max) scaled using FWHM_mult?
No, these stay constant, so they should be set large enough to be outside the largest expected scaled aperture radius. Of course this could require a repeated photometry run if you guess wrong the first time.
Craig