UKC

/ Photography of the moon, part 2

This topic has been archived, and won't accept reply postings.
The Lemming - on 01 May 2017
Seems like my initial OP on this subject has been archived. However I have been experimenting to see if I can improve on my image taking and stacking skills.

A couple of months ago I followed the advice of the kind people on UKC and Stargazers Lounge and had another go at capturing the moon. I took some shots and video during the day along with more shots and video at night. It then took me a month or so to find the time and then work out which software I needed along with using them without the computer freezing under the strain.

The moon at night was a stack of 1,000 images taken from a movie file. The daytime image is from 500 images taken from a movie file.

I'd appreciate any advice on how to improve.

https://www.flickr.com/photos/the1lemming/34239144161/in/dateposted-public/

https://www.flickr.com/photos/the1lemming/34239146971/in/dateposted-public/
richard_hopkins - on 01 May 2017
In reply to The Lemming:

Those pictures look good. What do you want to improve?

If you want more detail then a longer focal length or smaller pixel size will be needed. Eg a simple webcam has a much higher pixel density than a DSLR so you get a larger image, and the increased noise from the small pixels can be reduced with averaging due to stacking.
How does a single frame look compared with the total stack? Is there a significant difference or is the stack just overall a bit sharper?

What focal length & shutter speed did you use and did you track the moon's movement? For something bright like the moon you can use a fast shutter and high iso to minimise blurring due to movement (both tracking of the moon and the atmosphere) then the stacking process will then reduce the noise. However if you are taking videos, you might not have this flexibility on the camera.
Blue Straggler - on 01 May 2017
In reply to richard_hopkins:

> Those pictures look good. What do you want to improve?


He wants to remove a sharp white edge

richard_hopkins - on 01 May 2017
In reply to Blue Straggler:

It could be added by over enthusiastic sharpening in stacking process or even in the camera itself.
Are they present in the original movie?
The Lemming - on 01 May 2017
In reply to richard_hopkins:
> Those pictures look good. What do you want to improve?

I don't know. I've never done this before I don't know what pitfalls I could have made or how I could improve if I had another go at the project.


> How does a single frame look compared with the total stack? Is there a significant difference or is the stack just overall a bit sharper?

I took both RAW images and a single movie file with my mirrorless camera. I have to admit that the individual images of the movie ripped with software, and the RAW images that I took with the camera of the moon are, in my eyes, inferior to the finished composition by using software to stack 1,000 images. I had 10,000 to play with but only used 1,000 and then used 200 of the best images.


> What focal length & shutter speed did you use and did you track the moon's movement? For something bright like the moon you can use a fast shutter and high iso to minimise blurring due to movement (both tracking of the moon and the atmosphere) then the stacking process will then reduce the noise. However if you are taking videos, you might not have this flexibility on the camera.

As for focal length the movie and RAW images were shot at 400mm. With a full frame dSLR camera this equates to 800mm and even then the moon did not fill the frame. I have a small dobson telescope and maybe one day I will try and glue my camera to that. However I think this would be a bad idea seeing as the telescope mount is exceptionally basic and probably would not take the weight of the camera attached to it. But never say no.

When filming, I had a shutter speed of 50 frames per second. I make sure that everything is filmed at 50 frames per second so that I can get stuff to play nicely when video editing with the UK PAL setup. I'll be honest and say I don't have a scooby if this is the best practice especially as I publish everything for the internet rather than for viewing on the telly.

After my first attempt I asked for advice on here and my astro forums. Both sites mentioned that unless I had a mechanical way of tracking a subject then manually moving the camera would change perspective and add difficulties the editing phases. With all this in mind I chose to set my camera onto a tripod and keep it stationary. This allowed me 6 minutes for the moon to zoom across the viewfinder.

With the night shots in RAW the meta data was as follows:

ISO 200 (Lowest ISO the camera had)
F9 (Close to the sweet spot of f8)
100th second

I have to admit that the astro editing software did not like RAW files and had a hissy fit, crashing the computer. Probably as each RAW file was over 16Mb and there was a shed load to use. In the end I went for BMP files as the software played nicely with this format. TIFF was as unstable as RAW and I did not want to stack JPEGS as I was beginning from a lossy foundation and I'd rather start from as best a image format as possible. RAW was out so BPM was next best.

One thing that blew my little mind was how much the moon actually rotated during my little project. Yes I know the moon rotates but until I gave this project a go I did not actually witness it. During the day I shot the moon against a blue sky and a few hours later I shot it against a night sky. Even with a few hours difference I could visibly see a difference in the moon's rotation. Yay for science.

You also asked about over enthusiastic sharpening in the process. I will admit that when I used registax for a previous attempt there was an obvious white line which ruined my experience of viewing the image. I did not include that link as I was aware of this issue and saw no reason to add a link.

A kind individual on my astro forum explained how to reduce/remove the white line during processing/sharpening, and as far as I can tell I can not detect any sharp white lines to spoil the image that I gave a link to.

However I am open to comments and advice on improving my image further and would welcome comments on where sharp white lines can be viewed so that I can attempt to remove them.

I'll admit that the Blue Moon has some sharpening issues however I now know how to work on them.


I genuinely want to improve and get the best shot I can.

Cheers
Post edited at 18:09
Fredt on 01 May 2017
In reply to The Lemming:

> As for focal length the movie and RAW images were shot at 400mm. With a full frame dSLR camera this equates to 800mm...

That's different to my understanding. A 400mm lens on a full frame dslr equates to 400mm, unless I'm missing something.

(For comparison, I got this, not as good as yours in quality, with a single shot raw at 400mm, ISO 400, 1/400, f5.6 on a full frame 6D

https://flic.kr/p/McMVFC

The Lemming - on 01 May 2017
In reply to Fredt:
> That's different to my understanding. A 400mm lens on a full frame dslr equates to 400mm, unless I'm missing something.

You are correct, a 400mm lens on a full-frame dslr gives a focal length of 400mm. However I do not have a full-frame dslr but rather a micro four thirds mirror-less camera. This means that my lens which is 400mm has a similar crop to that used with a full-frame dslr and an 800mm focal length.


I also like your image of the moon.
Post edited at 22:23
malk - on 02 May 2017
In reply to The Lemming:

what software are you using? could you post example of single image used for stacking?
have you tried shooting through your dob eyepiece?
The Lemming - on 02 May 2017
In reply to malk:

> what software are you using? could you post example of single image used for stacking?have you tried shooting through your dob eyepiece?

To get the images ready for stacking I used PiPP or Planetary image pre-processor. This converted a movie file and turned it into various file formats of images. I found that I got stable results when I converted to BMP files. TIFF files crashed my computer if I used more than 5 images.

For stacking I used Registax, which gave best results when I coud adjust with wavelets. I tried a couple of other apps called Avistack and Autostakkert but they were not, in my opinion as visually good with their final results.

As for using my telescope, I don't think that it is suitable to film anything long enough with a heavy camera attached to it. It is a small telescope with a very basic mount and simple eyepiece attachment. However I don't mind experimenting if you think I could get good results.

Here is a single shot of the moon which I have tried to tart up in Lightroom. This is a sample of the 10,000 images I used.

https://www.flickr.com/photos/the1lemming/34245553162/in/dateposted-public/

And my telescope is a Skywatcher Heritage 130p. Its a great portable telescope but I don't think it is stable enough to mount a heavy camera and film objects without camera shake. However I am always willing to experiment if you think this is worthwhile.

https://www.firstlightoptics.com/dobsonians/skywatcher-heritage-130p-flextube.html
1
malk - on 02 May 2017
In reply to The Lemming:

thanks for info. maybe a processed raw image would be a better comparison
don't need much stability for the moon?
i was considering that scope a while back for portability. is it any good?
guess you've seen this thread?
https://stargazerslounge.com/topic/101464-skywatcher-heritage-130p-flextube-dslr-question/
The Lemming - on 02 May 2017
In reply to malk:
> thanks for info. maybe a processed raw image would be a better comparison don't need much stability for the moon?

I did take RAW images however the software could not handle them without crashing. I added an image that I used for the stacking process as an example. Even the RAW image wasn't as good as the stacked result.

I have to admit that I was very impressed with stacking and never thought that it would improve on a RAW image.


As for the telescope itself, I can confirm that it is exceptionally portable. It can be carried single handedly and set up in seconds. I also made a shroud from an off-cut camping mat to wrap around the extended telescope. My thinking was that it would reduce light leaking in while in an urban environment with street lighting.

I'm just a punter and found the added eye pieces very good however I did treat myself to an extra eye piece at 15mm. This gave me a 43 times magnification and with the extra barlow gives me 87 times magnification. I initially went for an eye piece that took magnification to the theoretical max of my telescope but I returned it a day later as it was virtually useless. You needed near perfect conditions and a bombproof mount otherwise the telescope shake and air movements simply distorted your viewing pleasure.

I went for this one as it got great reviews in magazines.

https://www.firstlightoptics.com/bst-starguider-eyepieces/bst-starguider-60-15mm-ed-eyepiece.html

I only get my telescope out a couple of times a year and with its small footprint, it does not take up a lot of space and get on Miss Lemming's proverbial's by taking up space.
Post edited at 14:17
malk - on 02 May 2017
In reply to The Lemming:

i looked up that crater on the the lower left where you can see the central peak in sun

https://en.wikipedia.org/wiki/Gassendi_(crater)

it has more than one peak- a test for resolution (at the right time of day)
alex_140 on 02 May 2017
In reply to The Lemming:

I'm surprised to see discussion around stacking images. Usually this technique is applies when photographing Andromeda or nebulae - the extremely faint light reaching us needs long periods of collecting the light to built up detail. A single exposure would need tracking and the scintillation due to the uneven air density would blur the image (this is lessened by stacking).

The light from the moon is intense enough to shoot a single exposure with low ISO and fast exposure - fast enough even, to shoot handheld. Here is a link to my last moon photo on flickr, which has probably been over sharpened in post-processing:

https://www.flickr.com/photos/116281704@N04/25045011024/

The detail on the moon is best highlighted (and captured) near the boundary of the illumination and shadow, seen best from earth during any lunar phase other than full or new moon. Capturing high detail require good atmospheric conditions, clear with low humidity, and being at higher altitude will also help.

Equipment-wise, a sharp lens (or telescope), with good light transmission, and a combination of high 35 mm equivalent focal length and high pixel density. My image was with a Canon 100d and a Tamron 150-600mm at 600mm and f 6.3, but f 8 would have been better as most lenses are sharpest at f 8. It was taken out of my bedrrom window in a city, so conditions could have been better, but again the luminosity of the moon will burn through mild light pollution or the atmospheric light during the day or twilight when stars and other features can't.

The image rotation was also commented on. The moon's rotation is not causing the detectable effect, but the earth's rotation, and therefore the camera, as I assume it is in a fixed orientation relative to the horizon across all shots.
richard_hopkins - on 02 May 2017
In reply to alex_140:

I think the Lemming's images are good and I doubt there is much improvement to be had, certainly it'll be a diminishing return. The processed image might benefit from some tweaking of the final Registax parameters (wavelets) to reduce the over sharp edges.

Stacking helps with lunar images because
a) You want to use a really short exposure to reduce atmospheric shimmer (see the DSLR image above with a 1/400s shutter), that looks sharp. Faster is better. It would be for faint objects too but they are just too faint to register.
b) The signal to noise ratio might not be great because of high camera gain (ISO) needed to get an image so stacking lots of images improves the signal to noise ratio exactly as it does for a faint object. You can see the noise in the image Lemming posted a couple of posts up.

However, with lunar images the real benefit in stacking comes from the fact that the stacking software (Registax or whatever) will chop the image into lots of tiny little sub-facets like a jigsaw and then pick the sharpest parts from many many images and chuck away the wonky ones which were distorted due to the shimmer.

At long focal lengths (especially onto sensors with small pixels like webcams) the atmospheric shimmer is so significant that the image looks like it is being viewed though a pond. Here is a random film clip from youtube showing what it's like https://www.youtube.com/watch?v=o90RyfjNYYM A long exposure with that level of shimmer will just result in a blurred mess, whereas the software identifies the sharp parts of each frame and makes a nice composite.

As a bonus, stacking many short frames means you can get away with poor camera tracking, perhaps even none at all, as the software tracks the movement from frame to frame, which is why it would be interesting to see what could be obtained with the Dobsonian telescope.

I think a series of RAW images might yield a better result than a movie because the movie mode in the camera almost certainly tampers with the image, reduces it in size and compresses it a bit whereas RAW will not (much). On a mirrorless camera I would imagine you could set the camera to shoot continuously as there is no mirror popping up and down, so you could get a series of RAW images at a few frames a second. There are plenty of scripts around that can turn each of those RAWs into a series of BMPs for feeding into Registax but as was mentioned on the older thread, that program will clunk with 20M+pixel frames as it's an old program.
The Lemming - on 02 May 2017
In reply to richard_hopkins:

Any ideas how I could bolt my micro four thirds camera onto my telescope?

I'm willing to give anything a try especially, as I can get exceptionally good close-ups of the moon. As for my mirror-less camera, it can take an awful lot of shots per second. The only problem is that the software has a hissy fit at the physical size of those images if shot in RAW.
The Ivanator - on 02 May 2017
In reply to The Lemming:

Have you seen this book? https://www.amazon.co.uk/d/cka/Full-Moon-Michael-Light/0224063049
Incredible collection of lunar photos from the NASA archives gathered during the Apollo missions. There was an exhibition of these images at the Hayward gallery about 15 years ago, extraordinary shots.
So if you want to improve your own lunar photography all you need is a rocket!
richard_hopkins - on 02 May 2017
In reply to The Lemming:

Use a t-ring adaptor to fix your camera onto the telescope. The exact combination of parts needed will depend on what you have but it should be on this list
https://www.firstlightoptics.com/adapters/t-rings.html

Other suppliers exist, perhaps even a local telescope shop if there is one near you, but these folk are helpful if you have any questions
Toerag - on 03 May 2017
In reply to richard_hopkins:

> If you want more detail then a longer focal length or smaller pixel size will be needed. Eg a simple webcam has a much higher pixel density than a DSLR so you get a larger image, and the increased noise from the small pixels can be reduced with averaging due to stacking.

correction - smaller sensor size needed, not smaller pixels. The problem is that using smaller sensors lack the number of pixels, so what you gain in effective focal length you lose in resolution. so you'd have to do a stitch to regain the resolution. Smaller sensors of decent resolution will have a smaller pixelpitch and be more prone to noise than Lemmings M43 sensor.

Lemming - why are you using F9? surely you're starting to get diffraction softening at this point? I suspect you'd be better off going for something like F5.6 / 8 and a shorter shutter speed, that's normally the sweetspot for m43. What does the MFT chart for the lens you're using look like? Is there any difference between F5.6/8/9?
As you probably know, going up in exposure time is a massive cause of noise for m43, so you're probably better off using a shorter shutter and wider aperture if the lens is still sharp at the wider aperture. Your current speed of 100/sec gives you lots of room to play with, you may find F5.6, ISO400 and 400/sec works better. Experiment!
In terms of the stacking software not handling the RAWs, do you really need to stack 1000 images? Can it handle stacking 20? If it can handle a lower volume, perhaps the best approach is to 'stack stacks' - stack a number of batches of 20 and then stack those images. I suspect you're best to avoid sharpening anything but the final image - sharpening beforehand will enhance noise which is what you're trying to avoid.

richard_hopkins - on 03 May 2017
In reply to Toerag:
Perhaps we are differing in definition of resolution. I am using resolution here in the physics definition way, ie the ability to separate closely spaced points. That is a bit different from the computer imaging definition of number of pixels in an image.
Small pixels improve the spatial resolution, big sensors with many pixels improve the image size.
The Lemming - on 03 May 2017
In reply to Toerag:

> Lemming - why are you using F9? surely you're starting to get diffraction softening at this point? I suspect you'd be better off going for something like F5.6 / 8 and a shorter shutter speed, that's normally the sweetspot for m43. What does the MFT chart for the lens you're using look like? Is there any difference between F5.6/8/9?As you probably know, going up in exposure time is a massive cause of noise for m43, so you're probably better off using a shorter shutter and wider aperture if the lens is still sharp at the wider aperture.

I have no defence for my choice of aperture apart from trying to remember so much, from reading it a month ago, and juggle it all in a short space of time while in the dark and cold. I'm guessing that f9 isn't that much of a deal breaker though.

> Your current speed of 100/sec gives you lots of room to play with, you may find F5.6, ISO400 and 400/sec works better. Experiment!In terms of the stacking software not handling the RAWs, do you really need to stack 1000 images?

I chose 200 of the best 10,000 images.

> Can it handle stacking 20? If it can handle a lower volume,

I could not get the software to play with any combination of RAW images, not even two.

> perhaps the best approach is to 'stack stacks' -

Have you had good results from trying this approach?

Thanks everybody as this information will help me to make a better image next time I try this. My next project is to try and capture the milky-way.

Toerag - on 03 May 2017
In reply to The Lemming:
>I have no defence for my choice of aperture apart from trying to remember so much, from reading it a month ago, and juggle it all in a short space of time while in the dark and cold. I'm guessing that f9 isn't that much of a deal breaker though.
Most m43 lenses reach optimal resolving power (sharpness) between f5.6 and 8, with some of the newer 'pro' lenses being designed to be sharp wide open. Instructions to use high F numbers are a hangover from the days of 35mm or larger film where you needed a small aperture to get enough depth of field or reduce exposure for longer shutter speed (fast shutters didn't exist).

>Have you had good results from trying this approach?
never tried it (or any stacking), it was just a theory I came up with. I don't know how it would compare to pure stacking, it should work.

My next project is to try and capture the milky-way.
Looking forward to it! Here's a starter for ten https://petapixel.com/2016/02/20/stack-photos-epic-milky-way-landscapes/
The Lemming - on 03 May 2017
In reply to Toerag:


Challenge accepted.




The Lemming - on 08 May 2017
In reply to The Lemming:
OK, I had one more stab at this project last night while it was hovering over Jupiter. I don't think that I can improve on this final go at photographing the moon, without buying more kit.

I'm quite pleased with how good this final image looks, especially compared to my other attempts that now look inferior to my final stab at this project.

https://www.flickr.com/photos/the1lemming/34373552252/in/dateposted-public/

I is quite chuffed
Post edited at 16:46

This topic has been archived, and won't accept reply postings.