Overriding pre processing script behaviour with flats

[stevek]

Gee, second post tonight!
Not sure if this is possible but appreciate your thoughts please?

I want to use the same flats for all my colour channels (RGB) - so I want to use the Blue flats.  All binning is 2x2 so all correct there.

I first calibrate my Blue lights.  Fine - all goes through.

I now pick green lights and add BIAS.  But now I add the Blue Flats (the blue flats filenames have BLUE in the filename itself).  Preprocessing script complains that no flats have been selected for the Green channel.  This wound imply that PI sanity checks the file names, and will not permit me to do want it to do.

Anyone else noticed this please?
Thanks, Steve

[marekc]

Hi Steve,

I would be curious to know your reason for wanting to apply the blue flats to all of the color channels. Are you shooting through separate R, G, and B (+/- narrowband) filters?

The reason I ask is because I've often heard that one should shoot a separate set of flats for each filter. A flat-field frame compensates for uneven illumination, including dust on the filter. Therefore (it is often argued), one should shoot a set of flats for each filter, since each filter could have a different set of dust particles on it, producing different shadows on the sensor.

I shoot a separate set of flats for each color filter in my filter wheel. I've been pleased with the results. I must admit, though, that the pattern of dust spots looks pretty much the same from filter to filter, so I suspect I'm really mostly seeing dust on the CCD sensor window. But, just in case there are some dust spots from the filters that I'm not noticing when visually inspecting my raw lights, I shoot flats for each filter.

- Marek

[Geoff]

A workaround would be to make two copies of the blue flat master, then change the filter type in the fits header file to 'red' or 'green', although as Marek says, you should do flats through each filter.
Geoff

[stevek]

Thanks guys, there are various people who think that, as long as your filters are spotless (and they normally are when kept inside a sealed filter wheel) that doing flats for each channel is a waste of time and that luminance flats or flats of the same binning level work just as well.  Most dust bunnies are on the primary or reducer glass and not on the filters.  Hence me wanting to check this out.

Do you guys use the preprocessing script or do calibration manually?

Rgds, Steve

[jkmorse]

Steve,

Your fliters must be clean room clean for there not to be any differences.  I see subtle and sometimes not so subtle differences in all my flats and  since they are only short iterations of less than 10 seconds and you only need a couple of dozen, its really not that big a waste of time.  IMHO, I would rather spend those extra few minutes than find out later I have a bunch of wasted lights because my blue flats leave artifacts in the Red and Green light frames.

May be just me, but i have had too many nights' work ruined by such little things that those kinds of shortcuts strike me as really dangerous.  For example, ruined two nights worth of work recently when focus slipped in night two when I decided that one time to shoot all Rs, then all Bs then all G's.  As a result, every start had a fat green halo.  Fixed it in the end but it ruined the color is every star and I will never make that mistake again.  The gods of astroimaging are a very unforgiving lot, especially with new, time saving techniques.

Which ever way you go, best of luck.

Jim

[pfile]

the 'speed' of the imaging system has a huge impact on dust bunnies. at f/2.8 i *never* see any dust in my images and in fact i have use Ha flats to calibrate OIII images. not perfect, but close enough in a pinch - there are gradients in the resultant image that must be removed after calibration with mismatched flats, but manageable with DBE.

as the f/ ratio increases the light cone gets more and more shallow and the effect of dust is more and more pronounced. with my f/8 system it would be a disaster to mismatch flats as dust on the filters is very obvious in the image.

if you want to see if there is dust on your DSLR sensor, the thing to do is stop your lens all the way down to f/22 (or whatever the max is) and it will be painfully obvious

rob

[jkmorse]

Rob,

I am in the same boat and that is where I am coming from.  I am at f8 with my CDK12.5 and not using separate flats would be disasterous.  Thanks for the explanation about the fast systems though.  I had not realized that was one more benefit to a fast system.

Jim

[Geoff]

Thanks guys, there are various people who think that, as long as your filters are spotless (and they normally are when kept inside a sealed filter wheel) that doing flats for each channel is a waste of time and that luminance flats or flats of the same binning level work just as well.  Most dust bunnies are on the primary or reducer glass and not on the filters.  Hence me wanting to check this out.

Rgds, Steve

There are at least three reasons for taking flats:
1. To correct intrinsic deficiencies such as vignetting in your optical train. The flats will be wavelength dependent because typically ccd sensors haved different QE at different wavelengths.
2. To remove the effects of dirt on your optical surfaces (filter dependent)
3. To correct the (wavelength dependant) pixel to pixel variation in QE--typically 2%--which limits your SNR, no matter how long your subs are.

You need to take flats through each filter.  Why spend hours gathering data, then blow it by begrudging a few extra minutes taking proper flats.
Geoff