Very impressive Burt. A true roadmap for the area.
How do you do your mosaic intergration - with Registar? Does it cope ok with the curvature for such huge FOVs?
Very impressive Burt. A true roadmap for the area.
How do you do your mosaic intergration - with Registar? Does it cope ok with the curvature for such huge FOVs?
That is a very good question Rob.
The problem of course is we are trying to map a spherical surface onto a flat one. Distortions will occurr.
I use Registar and as a start I register the left hand side three images starting with the centre image as reference image. I then calibrate the two top and bottom images to the centre image. Same procedure for the RHS.
I then use PS to rotate and crop so we have straight edges for the LHS and RHS where they overlap.
Registar is then used to align and calibrate these two images to produce the final mosaic.
We are not done yet as you can see from the map in the third post. There is noticeable curvature when the image is registered with a map from Star Atlas Pro.
I recently bought Starry Night Pro and have found that the projections match quite well with Star atlas Pro. See first image below.
I used an upsized image from SNP as a reference to bring the mosaic image back into a more rectilinear projection.
The 300mm F2.8L Canon lens I use is rectilinear to better than 1% ie very tiny barrel or pincushion distortion. It is still projecting a spherical surface onto a flat one.
The second picture shows SNP as the starting image with SAP and the final mosaic.
Here is the much better corrected version of the mosaic 14MB
That's a high quality image Bert.. star colours also looking natural which is not always easy with widefield imaging.. often end up with halos that discolour the image. very nice work indeed. would look good in a huge frame :-)
The original is a 550Mb Tiff. I need a 64 bit version of RegiStar as I have to downsize the original frames to make the mosaic. The dreaded 32 bit memory limit.
Although I think I am finally getting on top of what I am trying to do.
Wow, thanks for outlining your workflow Bert. I thought there would be challenges, but there's certainly some original and delicate work to get to the finished product obviously.
I just took the time to download the 35meg version and load it up in Picasa's zoomable viewer. What a stunning view. Not surprising I guess, but the 200k image here really doesn't even start to do just justice. Thanks again for showing Bert.
The original is a 550Mb Tiff. I need a 64 bit version of RegiStar as I have to downsize the original frames to make the mosaic. The dreaded 32 bit memory limit.
Although I think I am finally getting on top of what I am trying to do.
Bert
The individual stars are discreet and they are colourful. There is no background noise. Your processing routine here is just perfect. I wouldn't change anything. Often images of dense stars don't show the stars as discreet items and there is often blurry colour noise or dark coloured noise that makes it worse. Even with dedicated astro CCDs.
This is one of the most amazing, breathtaking astro pics I've seen and has touched me quite profoundly as this is my favourite region of the sky.
I beg you, I plead with you, please, sell me this photo.
If you do so, you will make this astro girl extremely happy.
Suzy you are welcome to use the picture any way you like for personal use, say make a print or a poster or whatever. If you cannot down load the big image send me your address by PM and I will post a CD with an even larger image at full resolution and quality. about 550MB 16bit tiff or about 75MB 8bit jpg at max quality.
For commercial use I generally only need the usual acknowledgment.
I would not put up high resolution images if I wanted to hide my efforts in a locked drawer. As I am retired I do this for fun and to keep me as 'sane' as last week.
Steve, Ron and Rob being of the old school in science I am a firm believer in free exchange of all information especially if it can help others with their own efforts. Most of my knowledge came to me from very many smart people that were way ahead of me in time and expertise, so if I figure out something new it is my duty to pass it on to try and balance the books even a small bit.
My lecturers and colleagues always told me very good science was about very good questions, not answers!
That's a high quality image Bert.. star colours also looking natural which is not always easy with widefield imaging.. often end up with halos that discolour the image. very nice work indeed. would look good in a huge frame :-)
Phil
Sorry Phil I must have missed your post.
There are a few things working in my favour.
Optics the best available a this focal length. I have baffled the lens by exterior aperture at f/3.6 and extended lens hood so there is no scatter inside the lens of light that does not contribute to the image. The lens only 'sees' the bit of sky it is imaging.
I now use an 82mm Hutech filter accurately mounted in front of the lens. This further contributes to contrast enhancement as no light due to light pollution that is not forming the image enters the lens.
As for haloes. I dither at about the 25 pixel level between exposures. When the upsized images X1.6 are stacked it minimises 'bleeding' of star colour into darker pixels next to bright stars. this also has the effect of improving resolution. They are the same thing.
The real control I have is my HDR method. When adjusting contrast, levels, stretching and saturation I am drawing on six 190MB images with exposures from 7s to 240s at the same time! The trick was to work out what the starting images should have done to them.
I may write a full 'how to' but at the moment it is still in development.
:jawd rop:
