avandonk
25-04-2009, 12:02 PM
I reprocessed the M7 widefield. Here it is 10MB
http://d1355990.i49.quadrahosting.com.au/2009_04/m7large.jpg
When correcting for flats and darks in Images Plus I always do this with the data in fits form. You get a far better result as the corrections are done at the original pixel data level rather than the interpolated colour data that a tiff has.
Now I have found if digital development or stretching is done in the fits form and THEN convert to 16 bit tiffs for stacking in RegiStar. The image has far fewer artefacts due to the interpolation step occurring after stretching.
The other trick is to make a reference image with Registar by median stacking about four images from the sequence you want to stack. These should be chosen thus. If you have sixteen images say use 1,4,8,12, and 16. This works even better if you have bothered to move the imaging rig slightly after about five images say or you have some slight drift due to non perfect polar alignment.
RegiStar has a nasty habit of stacking any residual hot pixel noise as real star data so it helps to use one of the two noise supression settings.
Bert
http://d1355990.i49.quadrahosting.com.au/2009_04/m7large.jpg
When correcting for flats and darks in Images Plus I always do this with the data in fits form. You get a far better result as the corrections are done at the original pixel data level rather than the interpolated colour data that a tiff has.
Now I have found if digital development or stretching is done in the fits form and THEN convert to 16 bit tiffs for stacking in RegiStar. The image has far fewer artefacts due to the interpolation step occurring after stretching.
The other trick is to make a reference image with Registar by median stacking about four images from the sequence you want to stack. These should be chosen thus. If you have sixteen images say use 1,4,8,12, and 16. This works even better if you have bothered to move the imaging rig slightly after about five images say or you have some slight drift due to non perfect polar alignment.
RegiStar has a nasty habit of stacking any residual hot pixel noise as real star data so it helps to use one of the two noise supression settings.
Bert