HFR should actually be HFD

In SGP the size of a star image is frequently referred to as HFR (half flux radius); e.g. it is used in the HFR module or for the autofocus module. All very convenient!

I decided to see how HFR would compare to FWHM (full width half max) calculated by APP (atropixel processor software).

I made 100 exposures (10sec exposures of M44 with 7 different filters over a temperature range of 10 degrees); per exposure I noted down the HFR values of SGP and the FWHM calculated by APP.

With considerable accuracy I found that HFR is 83% of FWHM. Which is odd, since it should theoretically be only 47% (see youtube or other forums for explanation).

The only conclusion I can draw is that HFR in SGP is wrongly named that way and it should be HFD (Diameter instead of Radius -which is half of diameter-).

Everybody agrees?

Not sure if I agree that it should be half. But I do agree that the HFR value does not line up with other metrics. I used PixInsight to measure FWHM on a number of subs and compared with SGP’s values. I found the reverse ratio that you found. The FWHM measured in PI is about 0.8x the HFR reported by SGP . In other words, my FWHM is LESS than the HFR.

I assumed that the issue was units of measure. In PI, the FWHM is shown in arc seconds in their Subframe Selector process, if you enter in the correct image scale. SGP does not seem to specify the units so I don’t know if it’s pixels, arc seconds, or furlongs. :slight_smile:

In any case, I just know that if I want to get an estimate of FWHM, which is a good proxy for seeing (as long as you’re properly sampled or over sampled), then I just multiply SGP’s HFR by 0.8x.