Posted by Grant Privett at 23:13 on 2013 Dec 05
Think Martin nailed most of it. just a few comments as my scope is the same size as yours.A year or two back I had a bash at going deep from a fairly dark UK site (blue on the Phillips dark skies map). I took a couple of hundred 45 second images using a 10" f4.3 Newt and a Starlight MX7. By median stacking (though sigma clipped would have worked as well) the images in bunches of 10 (to get rid of artefacts and comic rays or satellites) and then stack-adding the resultant 20 or so frames I managed to get detection of stars at the mag 21 level with a SNR of about 5. So not spectacular, but real. Personally I would not use a UV/IR filter as many modern CCDs are surprisingly sensitive in the 650-950nm range and you are just throwing signal away for no gain – if it was a refractor it would be another matter all together. I reckon with my current set up, a good clear night, no moon, clean mirrors and exposures of 600s plus and a newer CCD like the 694, I might just hit 22.5. Similar to the photographic limit of the UK Schmidt survey, so respectable.One last thing, my old Polaris mount was pretty poor so I had to recentre the target often (I lost about 70% of images to trailing). In effect this dithered the hot/cold/dodgy pixels around the image and helped to supress background streaking which ruins so many attempts to go deep. Dithering long exposures is deeply tedious so, when I have mended my mount I shall be using the dither function of Nebulosity in cahoots with PHD autoguiding to help flatten the background. Worth a try.BTW Astroart can stack with images aligned with interframe offsets derived from the PA and speed of a candidate…Also remember a flat field really will help when looking for very small variations in the sky background….