I'm not really happy with this, but thought what the heck, an image is an image. Shot in Ha last week while the moon was tormenting us (2hrs worth of 5min subs, QHY9/MPCC through the 8" Newt). I don't have a good Western horizon at home (umm, any horizon ) so I was surprised to find myself struggling to find a decent Nebula/Ha target on the night.
Brighter stars have shadows and contrast hasn't come out so well.
Anyway, comments welcome
Seems like this is a popular area of the sky at present! I did some test images with my GSO RC10 and focal reducer that turned out OK (at least, I thought so...)
My first attempt at Ha was 1.5 hours. Adding another couple of hours certainly helped with noise. I also found a lot of OIII in NGC 346 itself, but not so much in the rest of the area. Three hours of OIII added some nice detail. It's a challenging area of the sky!
That's still a pretty respectable image with plenty of detail. And, hey, we gotta keep ourselves busy while that pesky moon is up anyway, right?
There are tricks to get rid of shadows like that by the way;
Create a mask with the offending stars in it.
Save the mask from step 1
Grow the mask by as many pixels as are required to make the mask cover the shadows.
Subtract the saved mask of step 2 from the mask of step 3 (you should now have masked rings of pixels around the offending stars).
Feather it just a tiny bit.
Perform a median filter of only the masked pixels. You'll want to pick the smallest kernel size that still gets rid of the shadows.
Layer the original image on top of the 'fixed' image and set the layer to 'lighten only'.
Beer.
Sample attached.
Cheers,
Thanks Ivo - I better Startools out and try harder!
Looks good at first glance, and I expect if I go back to the FITS/TIFF even better.
Final step sounds like an essential addition to my processing too
Quote:
Originally Posted by RickS
G'day Rob,
Seems like this is a popular area of the sky at present! I did some test images with my GSO RC10 and focal reducer that turned out OK (at least, I thought so...)
My first attempt at Ha was 1.5 hours. Adding another couple of hours certainly helped with noise. I also found a lot of OIII in NGC 346 itself, but not so much in the rest of the area. Three hours of OIII added some nice detail. It's a challenging area of the sky!
Cheers,
Rick.
Thanks Rick - I enjoyed the detail you pulled out of 346 with the GSO. I guess I've got used to brighter Ha areas (you get spoiled when the Milky Way is overhead) but its nonetheless a very interesting area.
Gee I don't know, the original looks pretty good to me. The repro by Ivo has the stars looking better but the background looks a bit mottled now (may not be from his processing).
I am not sure what you are seeing in that Ha that is so bad. That is a standard looking Ha to my eye.
Despite narrowband being pushed as image through the moon, they definitely work better when there is no moon! I feel it really isn't totally possible (perhaps 3nm Ha may work) to image with any filter with a full or large moon close by the target.
You can try 2x2 that helps build the signal to noise ratio faster and helps but then with smaller aperture you will be giving away precious resolution.
Overall I think its a good result and perhaps you are expecting too much.
There are gradient removal techniques that remove even the grossest gradients very tidily that become more important to use in moon weeks or light pollution areas.
Greg.
Last edited by gregbradley; 18-09-2011 at 10:24 PM.
Thanks Greg. It's pleasing to hear you say that, as I think I was converting over to the religion that you can "image anywhere through anything" with NB. There was some decent glare in the finderscope frames from this night from memory, so the moon was definitely out and doing its "thang".
Thanks guys. I think the last time I played with Ha I had a good couple of hours or so, so have probably forgotten how important it is to have good long subs with plenty of total exposure. Didn't want to risk 10 or greater min subs on the night. Wasn't feeling lucky...
Hi guys, just wanted to point out that I really only performed the steps as outlined, modifying only a handful of pixels around the stars in that way. ST normalized the image automatically afterwards which probably accounts for a slight contrast enhancement.
It's just the JPEG compression that changed anything else. Bear in mind the repro is a heavily compressed JPEG that was modified and subsequently heavily compressed again. All credit for this image goes to Rob!
Good point Ivo - you did pull off a remarkable repro considering just the tiny jpg was used. Has prompted me to pull the lid off Star Tools again and dig deeper.
) so I was surprised to find myself struggling to find a decent Nebula/Ha target on the night.
I like them both. Although Ivo did get rid of some noise and cold pixels.
