|
Thanks to everyone for the feedback.
As I have said, I've learned a lot about processing over the last 12 months. Having said that, I recently found out how much I don't know! I met a guy from ASNSW at our last imaging meeting who blew me away with what he knew about photoshop and the processing of asto images. I think his name is Mick McCollough. I picked a few bits but just scratched the surface.
Anyhow, a couple of people requested an explaination of the differences in Processing these images. It was back in February that I did the first one, so I'll do my best:
The February image was converted from RAW to linear 16 bit Tiff using ImagesPlus. The subexposures were calibrated using a masterdark made of about 5 dark frames and a master flat made with about 8 or 9 flats. The calibration worked reasonably well, but left a gradient that I had no way of dealing with. I combined the images using average combine and then used IP's Digital Development to stretch them. All I did was to use the Auto settings and it produced a pleasing looking image. It was some time later that I learned that using the auto setting actually resulted in severe clipping of the faint data. I then finished off the image in Photoshop, doing my best to bringout what I could using levels and curves.
The main difference in basic processing this time around was that I used No White Balance CFA (Colour Filter Array) conversion on the lights and the calibration frames. This Pre-Bayer processing allows for much more accurate calibration. Also I have constructed a new set of Flats using 22 subframes and I used a masterdark made from 20 dark frames. These calibration frames were applied using ImagesPlus 2.75 bets 4, which has a really neat Auto dark frame matching routine. This allows for much more flexibility in the temperature variation of the darks. Also my techinque of using Digital Development has improved as well. I now only use Auto as a starting point, and keep a close eye on the histogram while performing the stretch. This way, I concentrate on the histogram shape rather than what the image looks like, thereby avoiding the clipping. Once I took the Image into Photoshop, I discovered that the image was much easier to adjust, and interestingly the gradient was gone. I can only put that down to a better flat frame. The lower background noise is probably due to the more accurate calibration and the use of auto darkframe matching. The master dark was much smoother too.
Another difference was the application of ImagesPlus Star Size and Halo reduction tool (also in beta). About the only other thing that was different was the Highpass filter mask that I used in Photoshop to enhance the detail in the dustlanes. I have briefly describe that in another recent thread, so I won't bore you with that again.
If you have any questions, ask away.
Cheers
|