Re: HD 145913

[Michael Koppelman <lolife@bitstream.net>]

Thanks guys. My goal, as always, is to understand. I understand 
Michael's reply regarding the different shape of the PSF. I hadn't 
thought this was important since people stress keeping the apertures 
the same size. (I realize this is not contradicting that.) This made me 
think that regardless of the aperture size, all will be well. I've seen 
people use extremely small apertures in some cases, like 1x FWHM.

I also understand that with a bright star like this, a little more sky 
will not significantly contribute to the noise.

I also understand that experimentation is good. The problem is, I don't 
necessarily know how to recognize what is better and what is worse. I 
might succeed in driving a certain standard deviation down, but what if 
it's true value has more variation?

Finally, I co-added a bunch of images and there are other stars near 
the program star and near the comp stars. Taken together I doubt that I 
am seeing any true variation (yet).

Cheers,
Michael


On Sunday, May 11, 2003, at 06:34 PM, Arne Henden wrote:

> As Michael R. mentioned, the important qualifying word in my
> reply was "bright".  MK was talking about the measurement
> of a 7th magnitude object, bright for just about any telescope.
>   The goal is to increase signal/noise and to remove
> systematics when you are trying to obtain 0.01mag
> or better photometry.  The other key word is "experiment."
> Try different approaches and see what works best for you
> for this specific project.
> Arne
>
>
> Stupendous Man wrote:
>>   Arne Henden wrote:
>>> I usualy also recommend apertures around 5x fwhm diameter
>>> when trying to do accurate photometry of bright objects, so you
>>> might experiment with aperture size and see what happens.
>>   Michael Koppelman replied:
>>> There is nothing going on (that I can see) after 4 or 5 pixels. You 
>>> are suggesting a radius (in this case) of 7 pixels or so. Seems like 
>>> I would get a lot more noise and a lot less signal in that case.
>>   There are two reasons Arne (and others) suggest using such large
>> apertures.
>>       a) if you are trying to calibrate your photometry of stars in   
>>        the field of a variable against, say, Landolt standards
>>          in a different set of images, then the PSF might change
>>          significantly from the target star images to the standard
>>          star images (especially at large airmass, as Michael K is
>>          observing).  In this case, a different fraction of light
>>          might fall within a "small" aperture, and so one would
>>          not make a fair comparison between target and standard
>>          frames.
>>       b) even if you are doing purely differential photometry of stars
>>          in a single field, there can be small differences in the PSF
>>          across the field.  Again, using large apertures protects you
>>          from this effect.    Note that Arne qualified his statement 
>> "photometry of bright objects"
>> with the word "bright".  For bright stars, the majority of the noise
>> in the measurement is due to variations in the photon count from the
>> star; noise from the sky is negligible.  In this case, using a large
>> aperture doesn't hurt.
>>   If you are working near the plate limit, trying to measure the 
>> faint stars, then, yes, using a small aperture may be necessary to 
>> avoid the extra noise from the sky.
>>                                 Michael Richmond
>
>
>