Aperture radius settings in the Seeing Profile dialog

[Richard Lee]

Hello

Question about aperture radius settings in the Seeing Profile dialog, Fig 16 in the extended AIJ paper 2017.

Based on the aperture radius-to-FWHM factors in the figure script & A.5 Radial Profile, I get the following aperture radius values for FWHM = 10.5 pixels, compared with figure annotations (nn).
     Source radius =              1.7 x 10.5 = 17.9   (20)
     Bkgnd inner radius =       1.9 x 10.5 = 20.0   (35)
     Bkgnd outer radius =      2.55 x 10.5 = 26.7   (53)

The inner & outer radii are roughly 2x the computed values, so maybe the figure shows diameter instead of radius?

Running version 3.4.0.01 I get a similar discrepancy, the source aperture matches but the bkgnd apertures are about twice the expected radius.

Regards

Richard Lee

[karenacollins]

Hi Richard,

Wow, good catch. It's been calculated in an unexpected way for many years due to a bug found looking into your report. The bug is fixed now, but it was decided to change the original 1.7-1.9-2.55 multipliers as part of the bug fix, because in our experience with bright stars attempting to achieve mmag time-series photometry, those values are too small. For now, I have changed to 2.0-4.0-6.0 for the three FWHM multipliers.

These values will be too large for a fainter star, so really what is needed is an interface to allow the multipliers to be set by the user. Unfortunately, I don't have time to implement that right now. Of course the paper will still be wrong until I have a chance to update it also.

If you are interested in the new version of AIJ, you can update to the daily build (3.4.0.04 or later).

Karen

On 3/15/2020 12:35 PM, Richard Lee [via AstroImageJ] wrote:

Hello

Question about aperture radius settings in the Seeing Profile dialog, Fig 16 in the extended AIJ paper 2017.

Based on the aperture radius-to-FWHM factors in the figure script & A.5 Radial Profile, I get the following aperture radius values for FWHM = 10.5 pixels, compared with figure annotations (nn).
     Source radius =              1.7 x 10.5 = 17.9   (20)
     Bkgnd inner radius =       1.9 x 10.5 = 20.0   (35)
     Bkgnd outer radius =      2.55 x 10.5 = 26.7   (53)

The inner & outer radii are roughly 2x the computed values, so maybe the figure shows diameter instead of radius?

Running version 3.4.0.01 I get a similar discrepancy, the source aperture matches but the bkgnd apertures are about twice the expected radius.

Regards

Richard Lee

---

If you reply to this email, your message will be added to the discussion below:

http://astroimagej.1065399.n5.nabble.com/Aperture-radius-settings-in-the-Seeing-Profile-dialog-tp1304.html

To start a new topic under AstroImageJ, email [hidden email]
To unsubscribe from AstroImageJ, click here.
NAML

[Richard Lee]

Hello Karen

Thanks for confirming, seems like a simple error which was hiding in plain view for several years ..  

In trying to get to grips with Multi-Plot, I tripped over the issue while trawling though the extended AIJ paper. Downloaded 3.4.0.04 and the apertures are consistent with 2-4-6x FWHM on one of my test images.

FYI 1 .. BAA member Eliot Hall has developed a working version AIJ-java plugin to convert AIJ photometry results into a format compatible with the British Astronomical Association database. I'll ask Eliot to get in touch when it's ready for release.

FYI 2 .. I've developed a collection of Excel-based vba macros to automate session planning, download and creation of radec files etc to support AIJ photometry. Somewhat inspired by Eliot's work .. started to consider  converting some of the routines to java to access as ImageJ plugin(s). Will email the details separately.

Richard

[karenacollins]

Hi Richard,

Please do post back with updates on your and Eliot's work. If the features are useful to many and become a fairly stable release, we can discuss integrating your code into the AIJ menu system, if that would be useful.

Thanks for your support of AIJ!

Karen

On 3/17/2020 4:56 PM, Richard Lee [via AstroImageJ] wrote:

Hello Karen

Thanks for confirming, seems like a simple error which was hiding in plain view for several years ..  

In trying to get to grips with Multi-Plot, I tripped over the issue while trawling though the extended AIJ paper. Downloaded 3.4.0.04 and the apertures are consistent with 2-4-6x FWHM on one of my test images.

FYI 1 .. BAA member Eliot Hall has developed a working version AIJ-java plugin to convert AIJ photometry results into a format compatible with the British Astronomical Association database. I'll ask Eliot to get in touch when it's ready for release.

FYI 2 .. I've developed a collection of Excel-based vba macros to automate session planning, download and creation of radec files etc to support AIJ photometry. Somewhat inspired by Eliot's work .. started to consider  converting some of the routines to java to access as ImageJ plugin(s). Will email the details separately.

Richard

---

If you reply to this email, your message will be added to the discussion below:

http://astroimagej.1065399.n5.nabble.com/Aperture-radius-settings-in-the-Seeing-Profile-dialog-tp1304p1314.html

To start a new topic under AstroImageJ, email [hidden email]
To unsubscribe from AstroImageJ, click here.
NAML

[karenacollins]

As it turns out, the method used before the recent change to calculate suggested photometric radii in the seeing profile code was *not* a bug afterall. It was a deliberate change to attempt to accommodate both focused and defocused images. I have now reverted to the original code starting with the 3.4.0.05 and later.

The method I have reverted back to looks for the radius where the average radial flux level drops by ~1% of the peak ADU in the aperture (by default). This radius r is the suggested aperture radius. It has nothing to do with FWHM.

If you prefer a different value of flux cutoff threshold, the default value can be changed in the "Aperture Photometry Settings"  "Radial profile model normalized flux cutoff" setting. The default value is 0.01 (i.e. 1%). To make the suggested aperture radius smaller, change to a larger number such as 0.02 (2%) and to make the suggested apertures larger, change the value to a smaller number such as 0.005 (0.5%).

Then the suggested inner background radius is defined as r1 = r x 1.75, and the suggested outer background radius is defined as  r2 = r1 x 1.5.

Finally, please note that these values are only suggested starting points. If you need the absolute best photometric results, you should try a series aperture radii smaller and larger than the suggested value to see which produces the lowest model residual (RMS).

Note: The current AIJ paper states that r = 1.7 x FWHM, r1 =1.9 x FWHM, r2 = 2.55 x FWHM. This is not correct and has not been correct for many years. However, it is the paper that needs updating and not the code.

Karen