The dust and other matter that causes the need for the taking of flats lies within the optical train, mostly on the glass surfaces of the focal reducers, filter, and ccd chip.
I usually take flats for each filter at the end of an observing session and if the focus position does not change much I will use them again on another night. If there is a big change I will retake them
But I have been thinking is there a need to retake take them if there is a reltaively large change in focus position.
In other words are the flat fields appearance and form changed by a change in focus. Would one set of flat fields for each filter last several sessions irrespective of the focus postion.
I've never found focus-specific flats to be needed. But it's easy enough to be sure for your own system, by comparing flats taken at different focus positions. [The easy way: use one flat as image, and calibrate it using another flat as a flat.]
I take flats and use them for weeks even months.
I also never focus except there was a change in the optical trian. I have a carbon tube and even though temperature changes all night it keeps focus very well.
my second telescope about to "come online" is a carbon fibre truss for that very reason, but my C14 Edge unfortunatly is not and there is focus change with temperture, particularly nigh to night.
My scope is a C14 Edge with focal reducer; you'll be happy to hear that it's not extraordinarily temperature-sensitive. My entire annual focus range is only about 5.5 mm from mid-summer dusks to mid-winter dawns (New Mexico). Even with my extremely shallow Atlas focuser, I never need to adjust the primary mirror position, all year.
More relevant: focus just doesn't matter for defects on the CCD window or filters, as their distance to the CCD chip doesn't change with focus. If your f/ratio is independent of focus (and my C14 is), all those shadows will be constant too. And while defects that are skyward of the focuser will theoretically change with focus, they are already so diffuse that they are very faint, and any focus effects would probably be very hard to see experimentally.
But again, you can verify this experimentally by taking flats at the focus limits.
Hi again Eric,
My C14 Edge is housed within a dome here in the UK and over a year my temperature variation is about 20C. During a cold winter's night it can drop from ~ 10C to 0. The focuser is a moonlite and is fixed-in-line during focusing so there is no primary mirror movement.
I will certainly verify experimentally as you suggest.
My nightly variation of temperature is 20 deg C and yearly it varies much more.
I am remotely observing from the Atacama desert at 2500 m elevation. Still I do not focus and can only recommend a scope with a carbon truss or tube.
which all my future telescopes will be and I have had 16" carbon truss ready for this year once the observatory is completed which incidentally is a trial set up for a planned 0,.65m telescope. But for the time being I can only play the cards I'm dealt :)
Good and bad focus cause some subtle effects that require ray-tracing to fully understand with most optical systems. I usually prefer dawn twilight flats because they occur after you've been carefully focussing through the night. That said, there are many times when you must take dusk flats (or lightbox flats, or EL panel flats, or whatever), before you've had a chance to look at a star and determine whether your system is in focus. As Eric suggests, you can experimentally test focus changes by taking flats on one night at large excursions from proper focus, along with focused flats, and then dividing the two flats to see any residual pattern. Dust is usually the thing that you see most often with change, depending on what surface it resides and where in the optical path - before or after a filter or corrector for example. My general rule of thumb is to take flats close to the correct focus, so that you don't have to make assumptions.
Likewise, how often to take flats depends on your particular circumstances, and how much of a gambler you may be. For some closed systems - say, an SCT with a focal reducer, filter wheel and camera that are never opened - new dust is rare and you can go long periods without new flats. For others - say, an R/C system where the filter wheel sees raw sky, sited in a desert climate - new dust (or fly wings, or cat hair...) is common. The amazing thing is that flats are free. You take them during periods that you would not otherwise be doing imaging - dusk and dawn, cloudy nights with a light box, whatever. Again, a good rule of thumb is to take flats as often as possible. Then you don't have to make assumptions, and possibly avoid multiple nights of compromised data.
Guidelines are just that, and you can take any approach that you feel comfortable with. However, calibration images are important, so monitor them regularly.
I am considering an upgrade to my system and have looked at a scope with a rotating Nasmyth focuser . With the Nasmyth the CCD and optical tube do not stay aligned. Before I waste a considerable amount of money, can you use this system for differential photometry?
There are at least two potential problems with a Nasmyth focus:
- if you are not accurately collimated, the optic axis will not be in the exact center of your CCD, and as the field rotates, any vignetting pattern will also rotate, and not symmetrically, so a single flat field won't work.
- as the field rotates, the spyders holding the secondary and tertiary mirrors will rotate in the field of view. They are usually ~1-2% issues, but will affect your photometry at that level if you are using a single flat.
Some people get around this by taking a library of flats at each rotation angle. Some people just ignore the effects - they may be smaller than your photometric errors.
Alt-az and Nasmyth both have these issues. Your best bet is to make sure the system is carefully collimated, and that any field rotator that you have is also rotating about the optic axis and not miscentered. Every telescope configuration has its issues, so I wouldn't get to worried about Nasmyth.
GEMs have their own problems. The axis flip through the meridian means fields get rotated 180 degrees, and the gravity load on the mirror is different, often causing enough flexure that photometry across the meridian can jump by a measurable amount. My preferred mount is the fork, if the arms are long enough that the camera can pass through!
What Arne says about GEMs is generally the case, but not always. I have 3 Astro-Physics GEMs (good grief!), of varying sizes. All can run a full 360 degrees, with no need to flip at the meridian or anywhere else: provided your camera clears the pier. I designed my (welded steel) piers to ensure there is clearance. With an Cass-type scope (weight is to the rear, so OTA tends to be mounted "forward"), I can observe through 360 degrees. I often observe through the nadir (for high dec objects), and can start a run with the OTA in either the "normal" or "flipped" orientation, at any hour angle.
(By "observe", I mean do photometry with a camera: it would be very inconvenient to observe through an eyepiece at such crazy positions.)
This is definitely not possible with all GEMs. Perhaps not with most GEMs. E.g. I had a Paramount 1100 (original model) with a hard stop right at the meridian.
Of course, when rotating through large angles, managing cables is always something to be careful of, and one must be careful that there are no accumulated full rotations that are winding up the through-the-mount cables.
I guess this is a bit OT for "flat fielding", but justed wanted to clarify.