44 replies to this topic

[Tom Glenn]

Here are a few lunar images that I took on November 27, 2018.  Overall the seeing was poor, but I could tell there was an occasional decent frame so I was able to salvage the run by being much more selective than usual, stacking only 100-250 frames out of 5000.  This leads to much lower SNR than I typically like, but was doable because of low gain.  The first image shows Theophilus as the prominent crater at the bottom, and the Apollo 11 landing site region is visible at the top of the image, with the craters Armstrong, Collins, and Aldrin, as well as Cat's Paw all visible.  (See my Apollo 11 post from months ago if you want to see this region annotated).  Click to enlarge all images. 

 

Edited by Tom Glenn, 01 December 2018 - 10:18 AM.

[Tom Glenn]

This image shows the crater Piccolomini at the bottom right, filled with shadow, and the large escarpment extending to the northwest is Rupes Altai.  Both this image and the one above are presented at 63% capture scale and compressed. 

 

Edited by Tom Glenn, 01 December 2018 - 02:37 AM.

[Tom Glenn]

This image shows the region stretching from the crater Aristoteles in the upper left to Posidonius in the lower right, on the northern edge of the Sea of Serenity.  The prominent crater at the top center is Burg, which is situated in a lava flooded region named Lacus Mortis (Lake of Death).  This image is presented at 54% capture scale with compression. 

 

Edited by Tom Glenn, 01 December 2018 - 02:37 AM.

[Tom Glenn]

This image shows the southern region of the Moon and is presented at 29% capture scale and compressed.  Edit: If you click the image it should now link to Flickr where you can access the full size image.  

 

Edited by Tom Glenn, 01 December 2018 - 02:09 PM.

[Tom Glenn]

This image is a cropped region from the image above, and is presented at 72% capture scale.  The prominent crater near the center is Moretus, and Clavius is in the lower right.  Past Moretus, located on the limb, are several prominent mountains that rise 8000m above the terrain below them.  

 

 

 

And that's it from this run.  I did not collect panels for the entire Moon, as conditions were not to my liking but I was happy to salvage a few decent shots from it.  

Edited by Tom Glenn, 01 December 2018 - 02:47 AM.

[aneeg]

Great images, Tom! Feels like sitting in the Lunar Orbiter.

 

Arne

[agmoonsolns]

These are amazing, some of the best! I keep looking at them, so much detail.

[aeroman4907]

Hi Tom, love the images as always.  Can you better describe just how poor the seeing was and how you were able to collimate and critically focus in those conditions?  I find I need seeing to at least be 3.5/5 in order to collimate and properly focus in order to come close to the resolution of my sensor.  Even then it is a pretty big challenge that takes about 20 to 30 minutes.  Seeing that 'good' also appears to occur only once or twice every few months, so I'd love to 'see' if I can be more productive in somewhat less seeing (pun intended).

[Tom Glenn]

Thanks for the comments everyone!

 

Aeroman, the number rating system for seeing is flawed because it depends on a subjective assessment that is relative to what one is used to encountering.  I only describe seeing as good or excellent if the live view is sharp and has a high degree of stability.  Basically you look at the live view and say WOW, I need to capitalize on this moment before it disappears.  Also problematic when trying to describe seeing is that there are several types of bad seeing.  On the worst end of the spectrum is the type of seeing that you seem to get a lot of in the Rockies, which is extreme turbulence and garbage across the board with no quality frames whatsoever in the entire run.  I occasionally encounter this myself, and there is simply no processing that can save the data.  Then on the more positive side (of seeing that I would still describe as poor) is the type in which the live view occasionally snaps into sharp focus, but only rarely.  Maybe once every 10-15 seconds or so there is a brief 0.25s where things look good.  This is where collecting many frames and being selective can help.  But the problem with this approach is that a shallow stack leads to poor SNR and a disappointing outcome unless the quality of those few frames was high enough.  Also it is only possible if the exposure parameters used low gain and longer shutter speed to fill the histogram.  This is why you can get away with stacks of only 100 frames on the Moon (although not ideal), but this is not possible with planets.  You need far more frames with planets, because the images were captured at a much higher gain.  In this case, it worked out well for me, but there are no guarantees in these situations.  You have to make a judgement call on whether it is worth imaging.  My scope was already setup before I went to sleep, and when I woke up at 3:30am and looked at the live view I was disappointed, but I thought it would be worth it to take a few captures.  If seeing had been any worse and I was unable to discern any fine detail while focusing, I would have just abandoned the session entirely.  Also note the Moon was over 70 degrees altitude.  

 

Here are two quality graphs from data taken that fits the above description.  The first is from September 30, and I would describe the seeing as excellent, at least during this capture.  The second is from the November 27 data.  In the first graph, there are 3000 frames represented and the green line is at the 500 mark, whereas in the second graph there are 5000 frames represented and the green line is at the 200 frame mark.  

 

 

 

As you can see, the first graph from the September 30 data allows for a much deeper stack.  The graph from November 27 shows much less stability in the seeing.  If the best frames are good enough to drive the stack, you can still get a good result from such a run, but often this is simply not possible.  It all depends on the quality of those frames at the far left of the graph.  Much has been written about how these AS!3 quality graphs cannot tell you how good the seeing was, and that's true, they just tell you how consistent it was.  But the best seeing is usually both very good and very consistent.  

 

Regarding collimation, it was not possible to see the Airy pattern so I had to get as close as I could with a slightly defocused star and centering the Poisson point.  There was considerable guess work here because the point was moving about.  Same with focusing, you just do the best you can.  I find that focusing is a skill that gets better with experience.  Also, if seeing is poor, it helps to zoom out a bit, certainly no more than 100% image scale in the live window.  

[aeroman4907]

Thanks very much Tom, that was a very helpful answer.  I too have found similar experiences with the quality graphs.  Yes, they don't give a true representation of actual quality; however, I've never had a consistent quality graph in poor seeing.  Your first quality graph is amazing and one I've never been able to obtain during imaging.  My very best has the 75% quality mark around the 15% to 20% of total frames and the 50% quality mark reached about to the right side quarter mark (75% of total frames).

 

Seeing is definitely a waiting game here.  If Sirius is twinkling badly then you know not to waste your time setting up.  That happens quite a bit here.

[Tom Glenn]

No problem.  I also just edited post #4 so that if you click the image it should link to my Flickr page where you can access the full sized image.  And I didn't mention earlier in the post, but I used a green filter (bandpass 500-575nm) and a 1.3x barlow (Siebert Optics).  

[aeroman4907]

How do you like the Siebert Optics 1.3x barlow?  I unfortunately have a beautiful and expensive 2x Televue Powermate that won't see much action, not that the 1.3x barlow will either, but just curious.

[Tom Glenn]

How do you like the Siebert Optics 1.3x barlow?  I unfortunately have a beautiful and expensive 2x Televue Powermate that won't see much action, not that the 1.3x barlow will either, but just curious.

I like it.  It seems very well made.  Some measurements I did indicate that in the configuration I have it in, it is providing 1.4x magnification, which is actually perfect for my ASI183mm.  I also have a Televue 2x barlow that I use for my ASI224 but that provides too much image scale for the ASI183.  Unfortunately I haven't had perfect seeing yet to test the Siebert with.  Would have been nice to have it for my data on September 30, but it hadn't arrived yet.  I should add that I only use it with the green filter.....because at F/10 the ASI183mm is already maxed out in resolution when using my 610nm filter and there is no improvement to be had.  But with the green filter you can get a little more, and this also extends to drizzle integrations if you do shoot at F/10 in green.  I never noticed any improvement with drizzle when using my 610nm filter, but with the green filter I do notice a bit.  But when using the barlow there is no need to drizzle, which is nice.  

Edited by Tom Glenn, 01 December 2018 - 02:34 PM.

[TareqPhoto]

I like it.  It seems very well made.  Some measurements I did indicate that in the configuration I have it in, it is providing 1.4x magnification, which is actually perfect for my ASI183mm.  I also have a Televue 2x barlow that I use for my ASI224 but that provides too much image scale for the ASI183.  Unfortunately I haven't had perfect seeing yet to test the Siebert with.  Would have been nice to have it for my data on September 30, but it hadn't arrived yet.  I should add that I only use it with the green filter.....because at F/10 the ASI183mm is already maxed out in resolution when using my 610nm filter and there is no improvement to be had.  But with the green filter you can get a little more, and this also extends to drizzle integrations if you do shoot at F/10 in green.  I never noticed any improvement with drizzle when using my 610nm filter, but with the green filter I do notice a bit.  But when using the barlow there is no need to drizzle, which is nice. 

The moon is a great target really, i won my first image in astrophotography with my moon shot, and so far my best images are my moon shots, i just got sharper and nicer results using my new filter which is ProPlanet IR742nm, i want to buy another filter to have colors of the moon with my color camera, IR742 kill the colors because it goes beyond red or spectrum so no colors.

[kbev]

I can only hope for a graph like your first one Tom, mine normally resemble your second graph and I have to limit myself to the best 5% to maybe 15% of the total frames which even with a capture of 20-30,000 frames still puts you at a deficit.  In the six years I've been imaging I think I've seen two or three nights of seeing that would be deemed 'good' or 'excellent' - most tend to be 'poor' to 'meh'.

[MADRID SKY]

Amazing shots, Tom!! WOW!!! I will have to test the moon under the green filter when weather gives us a break here. Do you usually image moon through green? I guess resolution through green is a bit better than through red... but I am not sure it really pays off... what's your experience?

[aneeg]

One thing I particulary like about your images, Tom, is the subtle greyscale. You seem able to produce every 256 levels of grey.

 

Arne

[Tom Glenn]

Amazing shots, Tom!! WOW!!! I will have to test the moon under the green filter when weather gives us a break here. Do you usually image moon through green? I guess resolution through green is a bit better than through red... but I am not sure it really pays off... what's your experience?

Thanks for the kind words!  The choice of filters is not always easy.  From a theoretical perspective, a green filter will have more resolution potential than a red filter, followed by IR.  But in the real world, the results are often reversed because of the reciprocal relationship between wavelength and turbulence, with longer wavelengths doing a better job of fighting the turbulence.  But if the seeing is excellent, the green filter will have a slight edge.  Whether or not you will be able to detect the difference, however, will depend on many factors, including the image scale you are imaging at (sampling rate) as well as your processing choices, and of course issues like collimation and focus.  My default filter for the Moon is the 610nm long pass filter, as it is an excellent all around filter.  It does a good job in mild turbulence (heavy turbulence is not worth imaging in my opinion....too much effort for little success).  I have only had the green bandpass filter for 2 months, but that is the filter I am now using if the Moon is above 50 degrees altitude and seeing is good.  It does reduce the amount of light that reaches the sensor, so you have to change exposure parameters (increase exposure or gain), which is only worthwhile if seeing is very good.  

 

It's always difficult to compare results from different imaging sessions, because the seeing conditions alone can account for any differences observed in the outcome, but here are some examples of images that I took with the 610nm filter, and the green filter, on similar regions of the Moon.  This comparison is further complicated by the fact that different lunar phases have different illumination patterns, leading to different contrast, but my feeling is that although all of these images look equally sharp when viewed in their entirety (zoomed out), if you look at the full-sized image, the green filter has a slight edge in resolution.  When initially looking at an image however, it's easy to confuse contrast with resolution, so the difference is rather subtle.  Also noteworthy is that when using the 610nm filter, I have seen no benefit to imaging at anything above F/10 with my scope (C9.25) and camera with 2.4um pixels, however with the green filter there is some room for improvement when imaging at up to F/13 or doing a drizzle integration on data gathered at F/10, but only if the seeing is excellent. 

 

610nm filter examples:

 

https://flic.kr/p/GjtcFv

https://flic.kr/p/GVTMcF

 

Green filter examples:

 

https://flic.kr/p/296jBM7

https://flic.kr/p/R9TRJq

Edited by Tom Glenn, 03 December 2018 - 12:28 AM.

[Tom Glenn]

One thing I particulary like about your images, Tom, is the subtle greyscale. You seem able to produce every 256 levels of grey.

 

Arne

Thanks for that very nice comment, Arne.  I do spend a lot of time thinking about the tonal variation of my images.  It can be a tricky thing to settle on, because I don't know what type of monitor people will be using (different resolutions, different brightness settings, etc), and these factors can greatly affect how the image looks.  I can't say I always get it right, but I do try to achieve an aesthetic yet realistic look.  

 

And thanks to all others for the additional comments and "likes"!

[MADRID SKY]

Thanks for the kind words!  The choice of filters is not always easy.  From a theoretical perspective, a green filter will have more resolution potential than a red filter, followed by IR.  But in the real world, the results are often reversed because of the reciprocal relationship between wavelength and turbulence, with longer wavelengths doing a better job of fighting the turbulence.  But if the seeing is excellent, the green filter will have a slight edge.  Whether or not you will be able to detect the difference, however, will depend on many factors, including the image scale you are imaging at (sampling rate) as well as your processing choices, and of course issues like collimation and focus.  My default filter for the Moon is the 610nm long pass filter, as it is an excellent all around filter.  It does a good job in mild turbulence (heavy turbulence is not worth imaging in my opinion....too much effort for little success).  I have only had the green bandpass filter for 2 months, but that is the filter I am now using if the Moon is above 50 degrees altitude and seeing is good.  It does reduce the amount of light that reaches the sensor, so you have to change exposure parameters (increase exposure or gain), which is only worthwhile if seeing is very good.  

 

It's always difficult to compare results from different imaging sessions, because the seeing conditions alone can account for any differences observed in the outcome, but here are some examples of images that I took with the 610nm filter, and the green filter, on similar regions of the Moon.  This comparison is further complicated by the fact that different lunar phases have different illumination patterns, leading to different contrast, but my feeling is that although all of these images look equally sharp when viewed in their entirety (zoomed out), if you look at the full-sized image, the green filter has a slight edge in resolution.  When initially looking at an image however, it's easy to confuse contrast with resolution, so the difference is rather subtle.  Also noteworthy is that when using the 610nm filter, I have seen no benefit to imaging at anything above F/10 with my scope (C9.25) and camera with 2.4um pixels, however with the green filter there is some room for improvement when imaging at up to F/13 or doing a drizzle integration on data gathered at F/10, but only if the seeing is excellent. 

 

610nm filter examples:

 

https://flic.kr/p/GjtcFv

https://flic.kr/p/GVTMcF

 

Green filter examples:

 

https://flic.kr/p/296jBM7

https://flic.kr/p/R9TRJq

Wow, Tom. That last green capture is something... that camera and scope really show forth. It is difficult to get better at 9.25". Besides, your processing is good. Do you use any barlow or eyepiece projection, or is it direct focus? As for green resolution... I have here an old spanish magazine with a color spot diagram study of several telescope designs. Curiously enough, Schmidt-Cassegrain design shows green/blue MTF lines fit together and ABOVE the red lines, which show lower. So, at a theoretical level, green has better lines and reaches better resolution. I agree with you contrast is not the same as resolution. I will always switch to higher resolution over better contrast as long as seeing permits. Digital processing gives us the chance to push up contrast, whereas better resolution is harder to acheive if you do not get it right at the spot. As you say, it is weather and altitude that marks the way to go. I will definitely have to test the moon under the green filter. Last time I captured, seeing was so bad that I didn't have a chance.

 

I envy your achievements on the moon, man.... Your pictures really pop-up. As a side note, I would just give you a little advice: do not push up so much the 1st wavelet layer. As you do this, some "patches" show up here and there. Go a little easier on that layer.

 

I have another question... you guess visually drizzle will benefit you, or do you test drizzle and non drizzle outputs spearately and then take decision?

And another one... I have seen your camera and sensor combo (ASI183) only reaches 19fps when at full resolution... so I guess you are capturing at 19fps, is that right?

 

God bless,

Sam

Edited by MADRID SKY, 03 December 2018 - 10:58 AM.

[TareqPhoto]

Wow, Tom. That last green capture is something... that camera and scope really show forth. It is difficult to get better at 9.25". Besides, your processing is good. Do you use any barlow or eyepiece projection, or is it direct focus? As for green resolution... I have here an old spanish magazine with a color spot diagram study of several telescope designs. Curiously enough, Schmidt-Cassegrain design shows green/blue MTF lines fit together and ABOVE the red lines, which show lower. So, at a theoretical level, green has better lines and reaches better resolution. I agree with you contrast is not the same as resolution. I will always switch to higher resolution over better contrast as long as seeing permits. Digital processing gives us the chance to push up contrast, whereas better resolution is harder to acheive if you do not get it right at the spot. As you say, it is weather and altitude that marks the way to go. I will definitely have to test the moon under the green filter. Last time I captured, seeing was so bad that I didn't have a chance.

 

I envy your achievements on the moon, man.... Your pictures really pop-up. As a side note, I would just give you a little advice: do not push up so much the 1st wavelet layer. As you do this, some "patches" show up here and there. Go a little easier on that layer.

 

I have another question... you guess visually drizzle will benefit you, or do you test drizzle and non drizzle outputs spearately and then take decision?

And another one... I have seen your camera and sensor combo (ASI183) only reaches 19fps when at full resolution... so I guess you are capturing at 19fps, is that right?

 

God bless,

Sam

Right, right and right

 

Honestly speaking, i really don't care at what focal length or frame speed or filters when i shoot the moon, for experimenting and test i do anything, and whatever nice results i came up with i just use it.

 

I shoot at full resolution for very slow frames maybe even less than 15fps, and i drizzle later, sometimes i like it and sometimes not, i decide after i finish stacking with and without drizzling and then i go with the best, i prefer drizzled one if it is good enough because it will give me larger or more resolution on the image.

[Tom Glenn]

Wow, Tom. That last green capture is something... that camera and scope really show forth. It is difficult to get better at 9.25". Besides, your processing is good. Do you use any barlow or eyepiece projection, or is it direct focus? As for green resolution... I have here an old spanish magazine with a color spot diagram study of several telescope designs. Curiously enough, Schmidt-Cassegrain design shows green/blue MTF lines fit together and ABOVE the red lines, which show lower. So, at a theoretical level, green has better lines and reaches better resolution. I agree with you contrast is not the same as resolution. I will always switch to higher resolution over better contrast as long as seeing permits. Digital processing gives us the chance to push up contrast, whereas better resolution is harder to acheive if you do not get it right at the spot. As you say, it is weather and altitude that marks the way to go. I will definitely have to test the moon under the green filter. Last time I captured, seeing was so bad that I didn't have a chance.

 

I envy your achievements on the moon, man.... Your pictures really pop-up. As a side note, I would just give you a little advice: do not push up so much the 1st wavelet layer. As you do this, some "patches" show up here and there. Go a little easier on that layer.

 

I have another question... you guess visually drizzle will benefit you, or do you test drizzle and non drizzle outputs spearately and then take decision?

And another one... I have seen your camera and sensor combo (ASI183) only reaches 19fps when at full resolution... so I guess you are capturing at 19fps, is that right?

 

God bless,

Sam

 

Sam,

 

Drizzle can sometimes smooth things out by adding image scale, although I never found that it added any resolution when using the 610nm filter.  I feel that at that wavelength, the system is already maxed out at F/10.  With the green filter, I somewhat feel like a 1.5x drizzle processing is nice, but the difference is very subtle.  As a general matter, I don't drizzle integrate all the data, because the ASI183 camera produces files that are so big, that Autostakkert cannot drizzle the entire image without crashing.  So in order to drizzle, you have to either capture smaller ROIs, or use PIPP to split up the recording into several smaller videos and drizzle.  This turns into a lot of work!  The main reason I got the ASI183 was for the large field of view for making lunar mosaics (such as this recent post).  So I usually am using the camera at full size.  Each file takes about 4 hours to process in Autostakkert, and so with a typical mosaic requiring 4-6 panels, this is 16-24 hours of time just to stack the images.  Then they need to be sharpened, blended together, and finished in Photoshop.  Splitting the panels up into even more sub-panels greatly increases the effort and time involved, so I usually don't do it for all of the data, but only for favorite regions that I want to process independently.  This is why I recently bought a 1.3x barlow to use with this camera, as critical sampling occurs at ~f/12 wth the 2.4um pixels.  This way I can maximize resolution without having to drizzle, although I haven't had perfect seeing to test this with yet, although the images in this post were taken using this barlow.  I don't think the barlow would have any benefit when using the 610nm filter, but it would when using green in excellent conditions.  

 

The camera does only record at 19fps when using the full sensor.  This, and the large file sizes, lead to some other technical difficulties, and may actually be the cause of the "patches" you refer to above, although I'm not exactly sure about your use of that term.  Processing is always a balancing act between detail and artifact, and using a large camera adds some additional difficulties (while saving time in other respects if the goal is a large FOV!).  For one thing, I can never collect as many frames as I would like, due to speed and file size.  A stack of 5000 frames is already 100GB in size!  The large sensor size also leads to significant issues with dynamic range, because of the huge difference in tonal value of the brightest spots in frame.  The raw data indicates that when captured at a histogram of 70%, most of the image is very dark, and this requires significant adjustments in Photoshop to bring up those shadows, especially near the terminator.  This may lead to some noise and artifact, and contribute to the patches you describe, rather than the sharpening.  One solution to this would be to capture additional panels covering the terminator, taken at much higher exposure values, and then blend the result together.  However, here again I run into problems with file sizes and time.  Adding more panels is not always possible, and in fact, during my last mosaic, after my first two panels covering the terminator were recorded, the seeing deteriorated somewhat significantly.  I had originally intended on revisiting some regions to take additional images, but I was running out of hard drive space in addition to declining seeing.  So there are a number of challenges when trying to make high resolution images that cover large regions of the Moon, but it is a fun challenge, and I appreciate your comments!

[MADRID SKY]

Tareq: yes, I agree with you that testing is the way to go. I am still in "testing mode", but I want to learn to master my equipment and processing and come up with a stable and reliable workflow. I need a lot of experience as I am a naive in practice (I have lots of theory and I need to engage more practice)...

 

Tom: no,  by "patch" I don't mean "artifact". Your images are very artifact free. It is really not a great deal and a bit of an aesthetical opinion. I see a bit of oversharpening / overcontrast of the "smallest elements". This oversharpening has a bit of an aesthetical "dust effect"  (the smallest elements in your pic resembles "dust in optics" and not so natural elements). But you really have to be very nerd and demanding. It is more of an aesthetical thing really. I would sharpen a BIT less the "smallest elements". Either if you use sharpening radius or wavelets, I would be a bit more careful with the smallest details and be a bit less agressive. You can live with it the way it is, and it is more of a taste thing to me. My personal taste is not to oversharpen and get as "natural" as possible so that picture can "talk" to you instead of "processing talking to you". It is difficult to explain this.  Pics are gorgeous anyway. Some of them seem like LRO images, boy. Your technique is great. FOUR HOURS in AS! you say? OMG!! My camera produces HD images (around 1900x1200px) in a powerful desk computer, and they last about 1 minute to do all the job (analyzing and stacking a 500 frame video, for instance). 4 HOURS!! That's really crazy... how do you live with that?

[TareqPhoto]

Tareq: yes, I agree with you that testing is the way to go. I am still in "testing mode", but I want to learn to master my equipment and processing and come up with a stable and reliable workflow. I need a lot of experience as I am a naive in practice (I have lots of theory and I need to engage more practice)...

 

Tom: no,  by "patch" I don't mean "artifact". Your images are very artifact free. It is really not a great deal and a bit of an aesthetical opinion. I see a bit of oversharpening / overcontrast of the "smallest elements". This oversharpening has a bit of an aesthetical "dust effect"  (the smallest elements in your pic resembles "dust in optics" and not so natural elements). But you really have to be very nerd and demanding. It is more of an aesthetical thing really. I would sharpen a BIT less the "smallest elements". Either if you use sharpening radius or wavelets, I would be a bit more careful with the smallest details and be a bit less agressive. You can live with it the way it is, and it is more of a taste thing to me. My personal taste is not to oversharpen and get as "natural" as possible so that picture can "talk" to you instead of "processing talking to you". It is difficult to explain this.  Pics are gorgeous anyway. Some of them seem like LRO images, boy. Your technique is great. FOUR HOURS in AS! you say? OMG!! My camera produces HD images (around 1900x1200px) in a powerful desk computer, and they last about 1 minute to do all the job (analyzing and stacking a 500 frame video, for instance). 4 HOURS!! That's really crazy... how do you live with that?

It is really very simple, get a good scope like that SCT or even a good Mak, get a good camera, one of those ZWO planetary cameras, and a filter IR PASS such as 610 or like mine IR 742 and just wait the good seeing condition, then ...... BANG!!!!!

 

Believe it or not, past years i was trying on the moon with different setup not astro, mainly DSLR and large lenses, even last year when i started, only DSO astro mono camera and i damaged my ASI120MC even before i give it a try, this year i bought 7" Mak, last year on December i bought another ASI camera, then i just gave them a try and WOW, i recently won with my moon shot, so it wasn't really that difficult at all, good for me that the moon is clear in our sky, and the quality i saw using that IR742 made me thinking that i was like in darkness shooting the moon with wrong items or not the good ones, believe me, get those items and thank me later.

[Tom Glenn]

Tom: no,  by "patch" I don't mean "artifact". Your images are very artifact free. It is really not a great deal and a bit of an aesthetical opinion. I see a bit of oversharpening / overcontrast of the "smallest elements". This oversharpening has a bit of an aesthetical "dust effect"  (the smallest elements in your pic resembles "dust in optics" and not so natural elements). But you really have to be very nerd and demanding. It is more of an aesthetical thing really. I would sharpen a BIT less the "smallest elements". Either if you use sharpening radius or wavelets, I would be a bit more careful with the smallest details and be a bit less agressive. You can live with it the way it is, and it is more of a taste thing to me. My personal taste is not to oversharpen and get as "natural" as possible so that picture can "talk" to you instead of "processing talking to you". It is difficult to explain this.  Pics are gorgeous anyway. Some of them seem like LRO images, boy. Your technique is great. FOUR HOURS in AS! you say? OMG!! My camera produces HD images (around 1900x1200px) in a powerful desk computer, and they last about 1 minute to do all the job (analyzing and stacking a 500 frame video, for instance). 4 HOURS!! That's really crazy... how do you live with that?

Sam, I agree completely with your thoughts about Moon processing.  The Moon has a certain aesthetic look to it that can easily be lost in the processing, and there are far too many over-processed Moon images.  I will admit that I sometimes straddle the limits of processing, and it can be tempting to get carried away when the quality of the data is high!  Sometimes, only after weeks or months go by and I revisit my old images do I come to a final conclusion about whether I like my version (sometimes I do, but often I think I could edit and do better).  Fortunately, it is very easy to play around with changing small details, and this also gives me something to do when the weather is bad!

 

Regarding the crazy amount of time that AS!3 takes with the files, that is something that just comes with the territory when using a larger sensor camera.  The ASI183 sensor is 20 megapixels, and this equates to several thousand alignment points in AS!3.  I know that number sounds crazy, but the APs really add up because the number required is proportional to the area of the image, which increases as the square of the linear dimensions,  My captures are typically between 3000-5000 frames, so each video is 60-100 GB.  On top of that, the initial alignment is very slow.  I go to the "Advanced" menu bar and click on "Experimental features" and select the 1.5 pixel vertical and horizontal Gaussian blur.  This adds a blur to the data which helps prevent AS!3 from aligning on noise (the blur is only added for alignment, and is removed before stacking so the data is not altered).  Before I did this, I was frequently getting pattern noise (like a grid) in the maria, and this pattern noise is the worst type of noise to get.  I got the idea to use this method in stacking from a CN post from Bart Declercq.  I hadn't previously been aware of this feature.  But it does seem to slow down the alignment, at least in my experience.  Although the stacking is very slow with these file sizes no matter what I do.  I then will typically do stacks of 250, 500, 1000, and 2000 for comparison, although in the case of the images in this post, I did stacks of 100 and 250 as well and ended up using those, because the seeing was unsteady.  Although the files take a long time to process, it's better than the alternative of using a small sensor camera and requiring 50 panels for a mosaic.  In that case the tedious part is not simply the processing, but also the data acquisition itself.  

Edited by Tom Glenn, 05 December 2018 - 04:18 AM.