[Aavso-photometry] Re: Dark Sky Annulus

Wed Jan 4 20:16:27 EST 2006

Arne,

Yes, it appears I did misquote you. Sorry about that. If the wings are large
then you need to use a larger aperture. As you know, the relative importance
of the wings is strongly affected by the seeing---the worse the seeing, the
more extended are the wings. On the other hand, too large a radius and you
may not be measuring the same background as that underneath the star. A
person really has to understand their instrumentation and know the extent of
the wings on the stellar profile. This brings us into the black art part of
photomety that you referred to. Said another way, doing really good
photometry takes some time. Dave Crawford and I used to get a real kick out
of comparing pairs of published papers measuring the same stars in a cluster
and seeing that the two sets of supposedly precise measurements were
sometimes 10 times (or more!) different than the quoted errors.

To illustrate the 3 stars for which the FWHM was measured, here is an image
that shows all 3 of them using apertures set to 8, 32.5, and 42.5 pixels
based on your quoted sky annulus criterion. Personally, I would put the sky
apertures closer in this case, but I think you'll agree that somewhere
between your rule of thumb and the one I mentioned is the "sweet spot" for
this image.
    http://www.mirametrics.com/pub/SkyAnn7.png

Michael Newberry

----- Original Message ----- 
From: "arne" <arne at aavso.org>
To: "Aavso-Photometry" <aavso-photometry at mira.aavso.org>
Sent: Wednesday, January 04, 2006 5:50 PM
Subject: Re: [Aavso-photometry] Re: Dark Sky Annulus

> Actually, Michael has misquoted me.  I said 5xfwhm radius, not diameter,
> for the inner sky annulus radius.  I'm more conservative than Michael is.
> Typically, fwhm is 3 pixels for my telescope, and I set the inner radius
> to 15 pixels, outer radius to 24 pixels (10pix width).  What you use is
> a matter of preference; the further you go from the star, the less 
> influence
> the star wings have on the sky background but the less correlation the sky
> background has with the sky underneath the star.  The important points
> are to be concentric with your target and to have sufficient pixels 
> included
> in the annulus to get good signal/noise determination of the mean sky
> background.
> Arne
>
> Michael Newberry wrote:
>> Wolfgang,
>>
>> You are right about FWHM being independent of the brightness (exactly the 
>> same if everything is ideal). I showed a graph of the radial profile and 
>> FWHM for a star that peaked at 668 ADU above background. You asked about 
>> a bright star. The brightest star in that part of the same image peaks at 
>> 13,200 ADU above background. So I measured that one, plus one super faint 
>> star just to illustrate this independence of FWHM on brightness. These 
>> were all measured in a similar area of the image so that focus 
>> differences, tilt, and field curvature do not change the actual PSF 
>> appreciably.
>>
>> Here are 3 stars ranging from bright to faint, measured using Mira Pro 7:
>>
>> Star peaks at 13,200 ADU, FWHM = 6.48:
>>    http:/www.mirametrics.com/pub/SkyAnn4.png
>>
>> Star peaks at 668 ADU, FWHM = 6.50:
>>    http:/www.mirametrics.com/pub/SkyAnn1.png
>>
>> Star peaks only 33 ADU above background, FWHM = 6.6:
>>    http:/www.mirametrics.com/pub/SkyAnn5.png
>>
>> That is a range of 400 times in star brightness (6.5 magnitudes). Here is 
>> an image showing the very faint star, marked by the aperture set using 
>> Arne's criteria with aperture radii of 8, 16, and 26 pixels. The "very 
>> bright" star to the right is the 668 ADU star that I originally measured:
>>    http:/www.mirametrics.com/pub/SkyAnn6.png
>>
>> This last image shows the contrast boosted way up to show just how faint 
>> that last star really is! Looking just at relatively faint stars, it is 
>> obvious that a person would be inclined to "eyeball" the FWHM as smaller 
>> than it really is. Even measuring it, without a robust FWHM tool, what 
>> would you get? Answer: It would be systematically smaller for fainter 
>> stars. And then you would use an inner sky aperture that actually rides 
>> up on the profile of star, rather than being at the edge of the sky. This 
>> would bias the sky to higher value. As a side point, one thing this 
>> exercise shows is how robust the FWHM calculation is in Mira. I am not 
>> trying to brag about Mira here, just make a point that your software 
>> should give you the same answer for FWHM independent of brightness, 
>> within variations caused by higher noise at the faint end. It is a common 
>> problem in the software world for different FWHM values to be reported 
>> for stars of different brightness.
>>
>> The point of all this is to illustrate that Arne's R = 2.5 x FWHM 
>> criterion is also independent of the brightness of the stars.
>>
>> Michael Newberry
>>
>>
>> ----- Original Message ----- From: "Wolfgang Renz" <w_renz at onlinehome.de>
>> To: "Michael Newberry" <mnewberry at mirametrics.com>
>> Sent: Wednesday, January 04, 2006 4:25 PM
>> Subject: Re: [Aavso-photometry] Re: Dark Sky Annulus
>>
>>
>> Hello Michael
>>
>> The FWHM example is one of a pretty faint star.
>> Do you also have one for a bright star with much
>> larger count values (say >= 30000 ADUs) ?
>>
>> The FWHM should be about the same (if its from
>> the same image), but I'm sure it will extend much
>> farther outside significantly above the sky back-
>> ground.
>>
>> Clear skies
>>  Wolfgang
>>
>
> _______________________________________________
>
> Aavso-photometry mailing list
> Aavso-photometry at mira.aavso.org
> http://www.aavso.org/mailman/listinfo/aavso-photometry
>
>
> 