7 replies to this topic

[james7ca]

This is a stack of 69 images taken using a Celestron EdgeHD with a 0.7X Reducer and a Nikon Z8 full-frame mirrorless camera. Image processing was with PixInsight, Siril, and Photoshop 2024. There are additional details in the image captions (click the below preview to see a larger image). There are around ten background stars, the faintest being about magnitude 12.

 

I have some ideas on how to make a better image, but that will have to wait for another day (of processing).

Attached Thumbnails

• 

Edited by james7ca, 13 April 2024 - 06:52 PM.

[flt158]

Utterly exquisite image from you, James!

I am wondering what that "bump" is coming from at 3 o'clock on the bright side. 

Could it be a mountain?

 

Not to worry if you cannot find out, dear friend. 

 

Clear skies from Aubrey.  

[gstrumol]

It's just this region:

 

 

from Virtual Moon Atlas

[james7ca]

Utterly exquisite image from you, James!

I am wondering what that "bump" is coming from at 3 o'clock on the bright side. 

Could it be a mountain?

 

Not to worry if you cannot find out, dear friend. 

 

Clear skies from Aubrey.  

It appears to be some highlands just to the north of Mare Humboldtianum (Gary had the right area). It shows up in the single exposures but was emphasized by the stacking of the images. It also shows up in NASA's Visualization Studio renderings of the moon for April 11th at 3AM UTC (which would be 8PM PDT, April 10 for me).

 

Here is a link to the NASA site (very useful for these kinds of questions): https://svs.gsfc.nas...lery/moonphase/

 

More interesting to me is the southern limb of the moon where you can see some large variations along the earthlit limb and some tiny sunlit peaks that are far from the remainder of the sunlit portion of the moon's limb (probably over-the-horizon peaks).

Edited by james7ca, 14 April 2024 - 12:40 AM.

[Borodog]

Fantastic!

[james7ca]

flt158, Gary, and Mike, thanks for the comments and also a good nod to everyone else for the likes.

 

I know of at least two more methods to stack these frames to see if I can produce a better result. The issue is that I know of no way to actually align to the features on the earthlit side of the moon. So, you need some kind of proxy for that alignment or registration. My original attempt used an offset from a registration on the background stars using PixInsight's comet registration process and that seemed to work pretty well for the single point of reference that was used for the "comet's" registration. However, I think it would also be possible to use SiriL to register to at least two points on the sunlit portion of the moon and of course there is always AutoStakkert!. I also tried PixInsight's FFTRegistation script but for some reason that produced some really bad results (but perhaps I need to revisit that method since I know it has worked in similar cases before).

 

One other complication is that I was using a standard T-adapter (M42) between the EdgeHD reducer and the full-frame Nikon Z8 and that caused quite a lot of vignetting in the corners of the frame. That didn't affect the image of the moon but it did add artifacts around the corners of the frame and what I really need is a so-called "wide" T-adapter that uses M48 threading. However, those are currently hard to find for Nikon's Z mount and I haven't yet decided on which adapter to try.

 

Anyway, I don't think I'm completely done with this data set so I may have something better in the coming week or two.

[Borodog]

I’m confused. Autostakkert will align on the earthshine side just fine, and will almost certainly produce the best result of any option, so what is the dilemma?

As for the vignetting, have you tried sky flats? They work a treat.

[james7ca]

I've used AutoStakkert! before for similar images (many times, and here is a post on CN from 2017), but I'm not sure it will work that well with such a dim and ill defined image give the base exposure that I used. It should work given a longer exposure, but then the moon will have a greater amount of blur from the seeing and the errors in tracking. In any case I still want to try a variety of stacking methods to see which one works best and AutoStakkert! is on the list for such work. I also need to see whether it will stack better in a RAW, linear format or if it would be better to start with a non-linear, stretched image. The latter might be better for Autostakkert! or some of the other techniques that I may try. Photoshop produces a really good looking result with its RAW processing and its RAW denoising option may be the best I've ever used (although it is very slow, it would probably take two hours for it to finish all of my subs given that each is 46 megapixels in size).

 

For this first round I also wanted to do all of the processing on my Macbook Air and I've never tried to run AutoStakkert! on a Mac using some kind of emulator for a Windows PC.

 

Interestingly enough, the sequences where I used a normal exposure for the sunlit side of the moon also show some traces of the earthlit side and while I've done images before that revealed feature on both sides they've never been to my complete satisfaction. But, this is the first time I've used the Nikon Z8 on the EdgeHD so this configuration may provide some advantages over what I've done before.

 

As is, this is still a pretty sharp image (IMO) and the question would be whether it can be made any better with a different alignment method.

 

Yes, flats should work to even out the background, but the vignetting was so severe that I couldn't even record stars in those areas (i.e. the corners had near zero exposure). Also, the camera has already been removed so it's probably too late to get any quality flats and one issue that I had not yet mentioned is that these same areas that begin to show vignetting also show flare and ghosting from the grossly overexposed moon. What all of this means is that I couldn't use the entire frame from the Z8, but little if any of that affected the moon but it would have been better to have had a clean, full-frame result.

Edited by james7ca, 15 April 2024 - 05:38 PM.