The nearly-full moon last week

StewartR

StewartR

Messages
11,513
Name
Stewart
Edit My Images
Yes
While I was out hunting the International Space Station one evening last week, I noticed that the moon was looking lovely.

I took 116 images and put them through Registax to try to bring out more details. It seems to have worked.

26231964_1994165563942012_5800426346665454118_o.jpg
 
Last edited:
chriscross said:
What were your settings, I can never get it right - cheers.
Click to expand...
It's not that hard an object to photograph, because it's in full sun. So there's a wide range of settings that are workable.

The biggest error that some beginners make is allowing the camera's meter to be confused by the large amount of black sky in the frame, which can cause the moon itself to be hugely over exposed. That's easily fixed by spot metering on the moon itself, or by shooting manually with 'Sunny-16' type settings and tweaking from there. If you're not making that beginner error, but you still aren't getting good results, then perhaps you should create a thread here, post an example of what you're getting, and ask for help.

Anyway these images were shot with:
  • Canon 7D Mark II
  • Canon EF 600mm f/4 L IS II USM
  • Canon EF Extender 2x III
Actual focal length was 1200mm, and full-frame equivalent focal length was 1920mm. The diameter of the moon in the original images was about 2550 pixels.

The EXIF data says ISO 800, 1/500th at f/8. That surprises me a little as it seems a bit high - it's 2½ stops brighter than the Sunny-16 rule would suggest, whereas I'd have expected to be about 1½ stops brighter (based on the moon's low albedo) - but it seems to have worked.
 
Last edited:
Moon Tonight by Mark Johnson, on Flickr

I took this a night or two ago. Still very noisy.

Used my Em1 mk 2, ISO 64, 1/80 sec at F7.1. Panasonic 100-400mm at full throttle (400mm)

On a tripod, but still used IBIS as the tripod seemed a bit shakey. Tripod is a Manfrotto thing. used a remote control release, and took maybe 20 shots in swift succession. Stacked in Registax.

have had better results handheld to be honest.

Just so noisy.........
 
Stewart, how many shots do you think something like Registax actually NEEDS to pull that sort of detail? Do the images need to be accurately overlaid or does the software do the registration of images automatically? Is the software easy and intuitive to use? Sorry for the questions but this is something I'd like to try but I'm not too good at new things if they're complicated!

That image looks oversharpened at a first glance but peering a bit closer, I can see that it's just more detailed than I'm used to seeing!
 
Nod said:
Stewart, how many shots do you think something like Registax actually NEEDS to pull that sort of detail?
Click to expand...
I have no idea whatsoever. I expect it depends strongly on how turbulent the atmosphere is whilst the images are being captured. I used 116 because that's how many I shot at 10 frames per second before I got bored. There's a feature in Registax where you can choose how many images it will stack - either the best "N" images, or however many are within "X%" of the quality of the best one. (Don't ask me what "best" means here though!) In principle I could reprocess the image using the best 100, the best 80, and so on, and see what I get - but I think that will just tell me how many were needed in my circumstances on that particular night, in the conditions as they were then, and it won't provide any general insights.

Nod said:
Do the images need to be accurately overlaid or does the software do the registration of images automatically?
Click to expand...
Registax seems to handle alignment pretty well. There's a parameter that allows you to specify how far it's allowed to shift each image in order to align it, so it probably makes sense to check that's large enough. But I didn't have any problems.

Nod said:
Is the software easy and intuitive to use?
Click to expand...
Not even remotely. It's complicated and confusing. There are tons of parameters which are not explained and I've never found any good manuals or tutorials. It seems to be incredibly powerful if you can use it properly, but the learning curve is long and steep.

Having said that, the first few phases of the process - image selection and alignment - aren't too bad and I've managed to muddle through them. But the final phase is enhancement, and that's really impenetrable.

Nod said:
That image looks oversharpened at a first glance but peering a bit closer, I can see that it's just more detailed than I'm used to seeing!
Click to expand...
Indeed! When I get some time I'll create a before-and-after image to compare the originals with the Registax output. The difference is quite striking.
 
Looks like I'll have to download it and have a play! I might even remember to have a notepad handy so I can jot down some settings and see what makes enhancements that are worth having. Long, steep learning curves really mess with my brane these days. I'm fine with older skills but pretty hopeless with new ones.
 
Well as someone has already said wow. The detail is amazing.
 
Nod said:
Stewart, how many shots do you think something like Registax actually NEEDS to pull that sort of detail?
Click to expand...
StewartR said:
I have no idea whatsoever. I expect it depends strongly on how turbulent the atmosphere is whilst the images are being captured. I used 116 because that's how many I shot at 10 frames per second before I got bored. There's a feature in Registax where you can choose how many images it will stack - either the best "N" images, or however many are within "X%" of the quality of the best one. (Don't ask me what "best" means here though!) In principle I could reprocess the image using the best 100, the best 80, and so on, and see what I get - but I think that will just tell me how many were needed in my circumstances on that particular night, in the conditions as they were then, and it won't provide any general insights.
Click to expand...
Having said that it's probably not useful, I just performed some experiments. I performed the same processing with various sizes of image stack:
  • best 90% of frames (105 frames)
  • best 75% (87 frames)
  • best 50% (59 frames)
  • best 25% (30 frames)
  • best 10% (12 frames)
I'm really struggling to see any differences between them at all. Even the 12-frame one (which looks virtually identical to the image I posted to start this thread) is much much better than anything I can do in Lightroom with just one image. But I think the 'wavelet' enhancement processing in Registax might be playing a big part in that; I don't think it;'s necessarily the stacking per se that's making the difference. At least, not with this image of the moon. YMMV.
 
  • Like
Reactions: Nod
Thanks for taking the time to go the extra mile, Stewart. Still haven't got around to D/Ling Registax let alone having a play myself!
 
Its a good job its and wife is in bed and not looking or she will be nagging me, can we do that, how is it done etc etc etc . Amazing picture, thank you for the info on set up etc.
 
You must log in or register to reply here.
Back
Top