Affiliation
British Astronomical Association, Variable Star Section (BAA-VSS)
Sat, 02/17/2024 - 20:52
As a beginner in CMOS photometry I am beginning to get a feel for the limits of my system.
What I'm not sure of yet is how large FWHM can get before errors become too large.
I have experienced a range from 2 on a good steady night to spells of 8 when the sky is poor.
Oddly provided the aperture and sky rings are adjusted accordingly it does not seem to have a large affect upon VPhot calulated errors.
I read that some observers defocus slightly anyway. Maybe there is a good old rule of thumb. Simplistic?
Kevin
Kevin,
You ran an experiment along with all members of your vphot course that assessed the precision and accuracy of your photometry at different apertures (n x FWHM).
Do you not remember what the common rule of thumb was as a result of this experiment?
Your question is slightly different BUT as long as you stay away from a defocus that leads to a donut for star profile, you will be ok. Your defocus spreads out the signal over more pixels. As long as the pixel counts do not get too small in the individual pixels you should not have problems with sky noise or read noise, etc., that become too large relative to your star signal.
IMHO, in most cases you would keep your images focused properly. Saturated stars obviously need shorter exposures or more defocus to avoid non-linear photometry. The key is that you do not want to have an image scale that is undersampled (<2 pixels / FWHM). That is where you do want to defocus to yield 2-3+ pixels / FWHM! CMOS cameras today have very small pixels and often present problems related to too much oversampling but this can be resolved by binning the pixels to better match your typical seeing.
As you noted, the focus is not critical. You can be more oversampled. I do not find that I worry much about my focus until I again can see star donuts on my image. The need to refocus depends on whether your scope tube is metal or carbon fiber. I rarely need to refocus my carbon tube scope over months. My metal tube scope is a different story. IF, the temperature changes significantly during a night, I may refocus every 90 minutes.
Some imaging software will run an autofocus either based on time delay or measured FWHM (seeing). And ,the software may allow an automatic defocus step on top of perfect focus.
You probably remember that I am not a believer in using some ones 'rule of thumb' when you can run a simple experiment with your own system and calculate your own results and reach your own conclusion. Sounds like you have actually already done this? Trust what you get!
Ken
Defossing has been used to optimize the precision of exoplanet photometry (Mon. Not. R. Astron. Soc. 396, 1023-1031, 2009). Flat fielding problems are substantially reduced, and exposure times can be longer. The benefit of the latter is that the total photon count for the star is greater, and hence signal/noise is increased.
Roy
I like to stay as focused as possible. I cannot guess what possible uses that I will have for the images at some later date. If the image is focused, I may be able to see a star that was dim then, but exploding now. If the image were defocused, I may not be able to see that dim star before it went nova. Also, I like using the default 5 pixel FWHM of VPHOT without having to muddle around to find the optimum setting for a series of defocused images. Saves a lot of time when doing a lot of photometry. I can see a 10 mmag change on a 9th mag star just fine while focused.
Ray
Ray,
I'm not implying that defocussed images should be used as a routine by everyone, just stating what was found in one circumstance by astronomers who needed optimal precision.
Knowing the principles involved may be useful to someone.
Roy
Just…
All good Roy.
Just explaining how I do it here and reasons why I don't defocus.
If you have reason to defocus, then you should.
Ray
Ray,
I should explain that I try to get as close to focus as possible, and only defocus deliberately for bright stars when necessary.
I use a 12 bit camera, started out a few years ago defocussing routinely, and promoted that practice on this forum. My most precise photometry was for a defocussed 8th mag variable with peak ADUs not far below saturation. Ìt was Ken Menzies who suggested to me that I should try in-focus images instead, despite the 12 bit ADC.
Roy
Caculad wrote: "I have experienced a range from 2 on a good steady night to spells of 8 when the sky is poor."
You don't give units here, but I'm assuming pixels. If so, your seeing must get pretty bad sometimes!
I have been observing the 6th mag K-supergiant variable BM Sco in Messier 6 for many years using telescopes large enough (0.7- amd 1.1-m) that it is necessary to defocus the telescope considerably in order to avoid overexposure. As part of this I also carry along CPD-32 4716 = NSV 9424, which is a mag 10 background M-giant in the cluster field. It turns out to be not variable. The reason for doing so is that the available comp stars in the cluster are all blue mid/late B-type stars, whereas BM Sco and NSV 9424 are fairly red. The systematics in the observing set-ups remain unexplored for the moment. I end up using a very large measuring aperture (about 25" diameter, and wide sky annulus, compared to the 1".1/pixel image-scale) on the defocused, grossly annular images. Given that the field is at >2.5 airmasses even on the meridian, the differential photometry is satisfactory (night-to-night rms ~0.015 mag from batches of five exposures).
As part of this, over the last couple of seasons I have obtained in-focus images on the same nights with our 1.1-m telescope to try getting calibration on the standard system. NSV 9424 and the comp stars are faint enough that I measured the NSV star on several nights on both the out-of-focus and in-focus images, taken back-to-back, the latter with typically a 14" diameter measuring aperture (3"-4" fwhm on the stars). Happily, the differential photometry showed no offset between the two series within a few millimags. So, at least this circumstance, it looks as though somewhat defocused images are OK to work from, although 'funnies' having to do with scattered light, flat-fielding errors, and so on may come into effect with whatever system you are using.
\Brian