there seems to be a lot of comment about taking good flats . could we start a thread with every ones comments on the best way of achieving good flats.
philip
For many years I put flat frames in the too hard basket. I never got around to building a light box, etc.
Then I started using CCD Commander. One of it's features is the ability to automatically take flat frames at the time of twilightwhen the sky is not so bright it saturates your camera but is bright enough to not have stars showing through. By this way you get a perfect flat field. Easy!
You don't need to use CCD Commander, it just makes it easy. You point the telescope relatively high in the sky but on the opposite side to the sun (east at sunset, west at sunrsie).
As an example it has calculated that at 6:02 tonight it will start taking frames, and will continue until the brightness count is at my desired level (20,000 ADU).
thanks roger ive been taking flats just at sunset when the sky in the west is white. I got it wrong? by the way ive also wondered about the virtue of converting them to black and white would that be useful? Also how many flats do you take and do you average them?
regards philip
Providing you end up with an image that is not fully saturated in any part and had an even amount of light (although possibly an uneven flat of course), you can do it at whatever time of sunset/twilight suits your equipment.
You don't want stars or other objects showing through, and you don't want it over saturated so you don't get a correct gradient.
I use a combination of 10 flat frames, averaged. I picked that number because I just generally try to have 10 flat, 10 bias and 10 dark, I figure it's a good number to average out the errors.
My ST7 camera that I do the flats for is only greyscale so I can't comment on the comparison between colour & greyscale. I suspect that there would be little advantage in colour, becuase unless there's something very interesting hapenning optically all the colours would be flat or un-flat in the same way, so a greyscale image would do the job. If there's any advantage in a greyscale image I'm not sure.
I cant comment if it's a good way but what I do is point the scope to Zenith either early in the morning or late afternoon when the sun is going down.
I put a white sheet over the scope.....take some short exposures and pick the timed exposure that give me between a 3rd and half way on the histogram...take about 10 then average stack them to make a master Flat.
I also make sure the whole imaging train is identical to what I was using whilst imaging.
I am working on a top secret method where you take flat frames with a ground glass screen in front of the optical train only illuminated by a dark sky and/or moon for the same exposure conditions as the light frames. That is same long exposure say six minutes. This means any fluctuation due to noise is also taken care of in the flat calibration stage. I always use in camera noise reduction and in conjunction with this would be even better.
Below is a 100% crop of a six minute flat with a tissue (not as good as ground glass screen) over a Canon 85mm F 1.8 at F2.8 with in camera noise reduction on. This shows the 'holes' that noise reduction produces.
You can work out the rest!
i can see that there are many ways of achieving a flat . i suppose the proof is in the final images. im not sure which way to go . i might do some experimenting myself having seen the lateral thinking thats being used .
philip
