Thanks Ross Marc and Ken. I was trying to get a decent mosaic of the Coal Sack or Emu Head but the bright stars of Crux produced faint reflections. There are still a few there but we take what we can get.
You are correct Ken it wasn't until I saw the whole mosaic that my mental version of this region changed completely. The wispy dust lanes are clearly separate from the stars. There seems to be a better sense of the 3D nature of the Milky Way.
Here is panel 4 at the real resolution of this data. It shows the lovely 'reflection' nebula due to a bright star of Crux.. 14MB
Awesome work Bert, and I thought I had a nice image of the Crux region, holy cow, don't I have some work to do Superb detail all the way around. So much to see also... Please if you can, can you bore us with details, I'd love to know how you do it...
OK each panel is 20x(3s, 7s, 15s, 30s, 60s and 120s) all at 1600 ISO at f/3.6. with the 300mm F2.8L Canon lens and Canon 5DH.
I dither between each image. This minimises sensor error.
Images Plus is used for flat and dark correction all in fits form. This way correction is done before any interpolation. Images Plus is then used to convert to TIFF and upsize by a factor of 1.6 using bicubic interpolation.
These sets of twenty images are then stacked using Deep Sky Stacker. I use med kappa sigma and balance colour with RGB background calibration.
Each of these stacks are now stretched with Images Plus Break point 5000 and background weight of 0.5550.
Photo Shop is now used to adjust levels to get the histogram to the left and balanced for colour.
The set of stacks are now put in a new directory.
Registar is now used to get a common area of all the stacks. I use the sixty second exposure as a reference image. PS is then used to trim this image so there are no black bits.
Run RegiStar again with the trimmed image as the reference. Now combine each exposure with the reference image by setting the weight of the reference image to .0001 and intersection combine.
This will give you a set of identically sized images that only differ in exposure.
Thanks for the explanation of how it all comes together Bert, really appreciate it. I am sure this method is quite time consuming but well worth it in the end. again.. Thanks.
This post will explain how I stitch it all together.
This mosaic is made up of twelve panels. Below is the spatial arrangement.
Top row 1 2 7 11 12
Centre row 4 8 13
Bottom row 9 10 14
The residual gradients due to light pollution are in the direction of the rows and get worse from left to right.
Bitter experience has taught me to stitch panels together in the direction of the gradients.
So we will start with the top row.
Put panels 1, 2 and 7 into a directory.
Using panel 2 as reference register panels 1 and 7 and save registered panels.
PS can then be used to crop panel 2 and the two registered panels 1 and 7 to get rid of reflections in overlap regions.
These crops are then put into a new directory and again using panel 2 as reference register and then calibrate both 1 and 7 to panel two.
Average combine panel 2 with the registered and calibrated panels 1 and 7.
We now have 1+2+7 and the overlaps are now invisible or matched for colour and brightness.
This combination can now be trimmed to get rid of black areas. We can call this topC1.
Put topC1 and panel 11 into a new directory.
Using topC1 as reference then register and calibrate panel 11 and combine topC1 with the registered and calibrated panel 11.
We can call this topC2.
Again trim to get rid of black areas and put this trimmed topC2 into a new directory with panel 12.
repeat as for topc1 and 11.
We now have the top row all nicely matched and registered.
Below is panels 1, 2, 7, 11 and 12 with the final top row all registered and calibrated.
This procedure is then used to stitch the centre and bottom rows much the same way.
Below are the three rows top mid and bottom.
Each one of these is now registered with the template I made from Starry night pro and saved.
These three saved registered rows can now be cropped to get rid of the black bits.
These three cropped rows can now be stitched together using the centre row as reference and registering each and calibrating.
Again combine the reference centre row with the calibrated top and bottom row.
You now have a mosaic that is rectilinear and perfectly matched for colour and brightness. Even the nasty gradient is matched!
Gradient exterminator used at the coarse and low settings get rid of the gradient quite easily.
Brilliant explanation Bert, never thought of doing separate mosaics then combining the separate ones later, actually makes quite a bit of sense, as you have shown. Thankyou once again.
Superb detail all the way around. So much to see also... Please if you can, can you bore us with details, I'd love to know how you do it...
