Been a long time since I posted anything. Won't go into reasons apart from the bloody weather!
ANYWAY!
This is 10 minute subs, not a lot of data (about 1.5 hours of each channel)
BUT, for those of you with keen eyes, you wi;ll notice slight (or great) misalignment of stars in the image, giving a distinct blurred appearance.
I currently use DSS, FITS Liberator and PSCS6 for processing.
Looking at adding Pixinsight to the mix, but really need to justify the cost to the better half.
WILL IT FIX THIS MISALIGNMENT???!!! I know and appreciate that there are no guarantees in life, but generally speaking...
What I notice on the image is that the alignment of stars (colour channels) is pretty good in the center, but progressively gets worse towards the edges. It's almost as if each channel has a different degree of curvature. But looking further I now notice that the red channel is always at the 4 o:clock position, just more so on the right side than the left side. Might your red filter be tilted somehow? And the upper left corner looks quite suspect. My guess is that something in your system is not aligned optically given the eneven degree of distortion. I hazzard a guess that if the distortions were symetrical your alignment optically would be good and the distortions could be a spacing issue of a flattener or reducer, etc. What is your gear and are you confident about these issues. How does this photo compare to previous attempts? Did something change?
Peter
PS I don't think this is a software issue at the root. Not saying software won't possibly help a bit however I highly doubt any software can fix what is on display. I think you must sort out the optical issue as a first priority.
Hi Peter. Thanks for taking the time to look and respond.
Here is my current setup and what I consider to be the limitations with my rig, to which I can make no further modifications or improvements without causing a divorce...
SW ED80, with Moonlite focuser. Manual filter wheel, ED80 0.85 reducer/flattener. Thin OAG with Atik 314L guiding and Atik 11000 imaging.
The size of the 11000 sensor is a problem. There is going to be some distortion in the corners no matter what I do with flatteners etc. I have got the field as flat as I am going to get it.
The red channel was not taken in the exact same position as the green and blue. It was framed more to the top left rather than centred. Don't know why. Just did it that way. That would explain the difference in the positioning of the red stars... I think.
So, my question was would PI, or some other program, have the power to accurately align the images? From what I understand, there are programs out there that can do this from different focal lengths, cameras etc. Is PI the best program to do this with?
I have broached the issue of purchasing this with the chief financial advisor, and I must say the response was less than pleasant, but not a NO... Have to save. Is the money worth it?
As Peter said, you're not going to get a perfect result from any software unless you can improve the quality of the raw data. Your sensor is much larger than your imaging circle, so cropping or divorce may be your only short term options if you're chasing perfection
Thanks Rick. I never knew you did those comparisons. Thanks. The difference is pretty amazing.
I am almost converted.
I think with proper care and attention to framing, and as you suggested, cropping it might be a short term solution, BUT I think I am going to have to bite the bullet.
Rick, you mentioned that you had tried Registar. I tried playing with a trial version last night, but is frustrating to figure out. What are your thoughts on it?
Thanks Rick. I never knew you did those comparisons. Thanks. The difference is pretty amazing.
I am almost converted.
No problem...
You can always sign up for a free trial and see if you like PI. Some people really don't
Quote:
Originally Posted by DJScotty
Rick, you mentioned that you had tried Registar. I tried playing with a trial version last night, but is frustrating to figure out. What are your thoughts on it?
Registar is really good at registration but I find it difficult to use and it doesn't handle floating point FITS data (I hate to throw away the hard earned precision of my calibrated subs.) I own a copy but I haven't used it since I switched to PI. I guess I might dust it off one day if I have to deal with some particularly tricky data.
I agree that the large chip is going to give you distortion at the edges. What I don't understand is why the distortion isn't symmetrical. If everything is centered properly it ought to be (symmetrical). So, I'd say that is worth investigating and should not cause a divorce to fix!! Maybe the focuser is sagging. You could test this theory with CCDInspector (free trial) by rotating the focuser and seeing how the field distortions change (or don't).
When you say you "framed" the red filter subs differently that ought not to be causing you any difficulty. I use CCDStack and I think there is also a free trial. You must align all your color channel subs at the same time and then use either bicubic b spline (or nearest neighbor if you have enough subs). After that combine each channel, and finally create the RGB combine. Alternately you can align each channel separately (again use NN or bicubic b spline) but you will need to repeat the alignment step before combining into RGB. For that reason alone I would align all the subs first.
Yes, for sure that looks a lot better. I just don't understand why you need to do that. I don't know anything about DSS. Does it try to align/register all your rgb subs? Is it failing to do so?
Yes, for sure that looks a lot better. I just don't understand why you need to do that. I don't know anything about DSS. Does it try to align/register all your rgb subs? Is it failing to do so?
Since the size of the sensor significantly exceeds the image circle of the scope/reducer there are going to be significant aberrations affecting the corners of the subs, and these won't affect the colours equally. No doubt DSS is doing its best but its a difficult problem.
It may be possible to improve things by adjusting spacing between the reducer and sensor and making sure there is no tilt but it's never going to be wonderful
You are having problems because your scope and reducer cannot handle the size of the 11000 sensor which is full frame in size.
Full frame sensors require a 3.5 inch focuser as a minimum and similarly sized reducer.
The ED 80 does not come with that sort of focuser or accessories. So you have a mismatch. But you can crop the bad bits out.
Also it seems you have some tilt happening which is making it harder.
Software won't correct it although you could try using Lightroom with lens correction and perhaps the distortion slider. I might improve but will not fully handle it.
I went though all this years ago with a Tak FS152 2.7 inch focuser which was not adequate with an SBIG STL11. When I upgraded to a 4 inch focuser and 4 inch flattener all was well.
It may be possible to improve things by adjusting spacing between the reducer and sensor and making sure there is no tilt but it's never going to be wonderful
Thanks Rick. I wish I could adjust the spacing but it needs to ge a bit closer to the sensor and unfortunately, I have no room let, what with the Filter wheel, and the thin OAG, to move it closer (unless I cut the filter wheel in half!)
You are having problems because your scope and reducer cannot handle the size of the 11000 sensor which is full frame in size.
Full frame sensors require a 3.5 inch focuser as a minimum and similarly sized reducer.
The ED 80 does not come with that sort of focuser or accessories. So you have a mismatch. But you can crop the bad bits out.
Also it seems you have some tilt happening which is making it harder.
Software won't correct it although you could try using Lightroom with lens correction and perhaps the distortion slider. I might improve but will not fully handle it.
I went though all this years ago with a Tak FS152 2.7 inch focuser which was not adequate with an SBIG STL11. When I upgraded to a 4 inch focuser and 4 inch flattener all was well.
Greg.
That's excellent information Mr Bradley. I really appreciate you taking the time to comment.
I do wish that when I was discussing the purchase of this camera with the retailer that they made this focuser size restriction clear to me. Basically, about a third of the sensor is "unusable".
OH WELL! We live and learn.
Cropping the images is a good solution me thinks
Thanks Rick. I wish I could adjust the spacing but it needs to ge a bit closer to the sensor and unfortunately, I have no room let, what with the Filter wheel, and the thin OAG, to move it closer (unless I cut the filter wheel in half!)
Nothing worse than running out of back focus, Scotty!
