Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

Copernicus, Stadius, Eratosthenes

  • Please log in to reply
19 replies to this topic

#1 lakeorion

lakeorion

    Apollo

  • -----
  • topic starter
  • Posts: 1227
  • Joined: 03 Aug 2010
  • Loc: Lake Orion MI

Posted 17 December 2018 - 12:18 PM

After a three month drought (of clear skies at the appropriate time of the month - plenty of precipitation in various forms) I managed to get the clubs new planet buster out and attempt an image.  In between banks of light clouds I pulled this out.

 

2018-12-17 Copernicus-Stadius-Erastosthenes copy.jpg

 

Meade 14" fork mounted, QHY 183M, Baader 685 nm filter.  Best 50% of 2000, 40 ms, gain 4.  Acquired with FireCapture, stacked with AS!3, post in PI.

IMO my best lunar picture to date, hopefully starting to get closer to the capability of the equipment.  Thanks to all who've given me pointers on this forum.

 

Unfortunately I missed getting a decent picture of the nice moonbow that happened right before this.  Ran in the house and grabbed the DSLR but handheld focusing and rushed HDR just wasn't in the cards last night.


Edited by lakeorion, 17 December 2018 - 04:46 PM.

  • Carol L, roelb and jlund like this

#2 lakeorion

lakeorion

    Apollo

  • -----
  • topic starter
  • Posts: 1227
  • Joined: 03 Aug 2010
  • Loc: Lake Orion MI

Posted 18 December 2018 - 04:30 PM

What a difference a day makes in the lighting.  Same equipment, similar workflow.

 

2018-12-14 Copernicus-Stadius-Erastosthenes copy.jpg


  • Carol L, neaptide and PatrickVt like this

#3 Tom Glenn

Tom Glenn

    Apollo

  • -----
  • Posts: 1414
  • Joined: 07 Feb 2018
  • Loc: San Diego, CA

Posted 18 December 2018 - 07:01 PM

Yes, the terminator moves quite a bit in the course of a day.  Since you mention trying to get the maximum performance out of your equipment, how would you rate the seeing, collimation, and thermal stability on these days?  Was it easy to focus?  You have done a nice job producing a low-noise result, although the fine detail that your equipment is capable of is absent.  I would guess that based upon your location and the time of year that good seeing is not common, although collimation, focus, and thermal issues can cause identical symptoms.  Did you try stacking less than 50% of the frames?  If seeing was poor, a stack this deep might not be optimal, although you may not see any improvement with smaller stacks.  A 40ms exposure is also somewhat long for high resolution imaging of the Moon.  I routinely do 15ms exposures on the Moon if conditions are excellent.  Although a super short exposure is not advisable (and cannot in fact freeze heavy turbulence) you probably don't want to be much slower than 15ms.  



#4 lakeorion

lakeorion

    Apollo

  • -----
  • topic starter
  • Posts: 1227
  • Joined: 03 Aug 2010
  • Loc: Lake Orion MI

Posted 18 December 2018 - 10:56 PM

Seeing the night of the 16th was fair, the 17th poor to horrible.  I was out tonight and I'd rate seeing good actually - gigabytes are still transferring to the main machine.  Also not sure if it makes a difference with the Moon but I'm in a suburban front driveway with Christmas lights all over and cars driving by.  The Moon is pretty bright and I am shooting in IR so hopefully those don't matter much.

 

Tonight I gave the scope an hour before attaching camera to give it more time to cool.  It's stored in a garage so the temperature difference is minimal.  Watching Capella got boring, but it looks like a small amount of collimation is in order for this machine, it's hard to tell with the mirror flop.

 

And in the bin of parts I did manage to find a t-thread Borg helical focuser.  I still have to touch the scope so it's not as ideal as a motorized Moonlite (like I have for my refractor) but it helps tame the Meade mirror.  It's a club scope so an SCT Moonlite won't happen until I have my own SCT - and mount it on a proper equatorial - I'm learning to dislike forks.

 

This is a 100% single frame showing what I've been able to get for focus from try #2 last night - still WIP.

Untitled-1.jpg

 

I was going for low gain / big histogram.  At gain 1 there's room to fiddle if I want to target 15ms or less.  Thanks for the pointer.

 

The pic from the 17th was 25 ms / gain 1, 25% of 2000.  Tonight I ran 20 or 15 ms / gain 1, 1000 frames because I figured with it so 'relatively' still I wouldn't need 2000.  (And 2000 frames / 76 gb takes so horribly long run through AS!3 shocked.gif )

 

I've been using PixInsight Multiscale Linear Transform for sharpening, the ability to make previews and see setting effects on the fly I think is better than anything else I have.

 

I'm just happy to have been out three nights in a row (without having to go to work) and burn some hard drive space.  I'm going to have to clean my RAID soon.  2TB just isn't as big as I thought it was any more.



#5 Tom Glenn

Tom Glenn

    Apollo

  • -----
  • Posts: 1414
  • Joined: 07 Feb 2018
  • Loc: San Diego, CA

Posted 18 December 2018 - 11:35 PM

Yep, it's nice to be outside having fun...I'm glad you are getting some clear nights!  From the frame you presented above, if that is the best possible focus, and collimation and thermals were good, then I agree that seeing was horrible, and there's nothing you can do to improve your image.  As for histograms, you can definitely use low gain on the Moon, but you don't need it to be zero.  Once you stack several hundred to 1000 frames, the noise is reduced considerably.  I typically set the exposure to 15ms and see what kind of gain I need to get to a histogram of 70%.  I'm not familiar with the gain settings on your camera, but on mine I'm running anywhere from 10% to 40% of the available gain, depending on the region of the Moon that's in frame and the filter.  When I use a green bandpass filter, the amount of light getting to the sensor is much lower than with an IR pass filter, so the gain needs to be higher, or the shutter longer, but I try to stay below 20ms at all times, even if seeing is excellent.  In another post from just a few weeks ago, I presented a series of images as a single frame, and with stacks of 5, 10, 100, and 1000 frames.  In that post, I didn't specify the gain or shutter, but I just went back and looked, and the gain was 14% and the shutter was 15ms.  Seeing was excellent.  This just demonstrates that you don't need to stack too many frames to reduce noise, as there isn't much difference between the 100 and 1000 stacks.  You can also add a Gaussian blur to reduce noise.  The comparison starts here:

 

https://www.cloudyni...more/?p=8992897

 

But in any event, these details apply only if you get some better seeing than from what you had last night.  Good luck with tonights imaging!  Also, light pollution, street lights, car lights, etc. won't matter at all on the Moon, as long as the lights aren't shining directly into your scope (a dew shield is sufficient to block these).  The Moon is bright enough to overpower anything other than someone shining a flashlight directly into your aperture.  



#6 lakeorion

lakeorion

    Apollo

  • -----
  • topic starter
  • Posts: 1227
  • Joined: 03 Aug 2010
  • Loc: Lake Orion MI

Posted 19 December 2018 - 12:47 AM

Green filter huh?  I have a Baader Solar Continuum loaded into the filter wheel. It didn't work at all for deep sky - too reflective on the front and I got huge halos around the brighter stars.  Briefly looked at it tonight but it was noticeably jumpier than the Baader 685.  I could give it a try.  The reflectivity probably wouldn't affect an SCT because the secondary mirror is so far away.

 

And yes I was using the dew shield on the 14".  Big floppy takes three hands to do the Velcro thing.

 

The other thing I need to figure out (but it can be a daylight thing) is why I'm not mustering better than 4-5 fps lately.  The very first time I hooked it up to Fire Capture I hit 9.8 fps and figured those problems were over.

 

First try from tonight is going through the glacially slow Surface Image Stabilization process.  CPU is loading one core, RAID is nearly going to sleep...


Edited by lakeorion, 19 December 2018 - 08:36 AM.


#7 Tom Glenn

Tom Glenn

    Apollo

  • -----
  • Posts: 1414
  • Joined: 07 Feb 2018
  • Loc: San Diego, CA

Posted 19 December 2018 - 12:53 AM

If seeing is poor I would stick with the 685nm filter, especially with 14 inches aperture.  Even in good seeing the 685nm filter is a good choice with that scope for the Moon.  Firecapture has some issues, and slow frame rates with the IMX183 sensor is quite common, because of the large pixel count.  Unfortunately, I don't really have any good solutions, but you could try SharpCap or another software.  



#8 lakeorion

lakeorion

    Apollo

  • -----
  • topic starter
  • Posts: 1227
  • Joined: 03 Aug 2010
  • Loc: Lake Orion MI

Posted 19 December 2018 - 02:27 PM

Another night another Copernicus, Stadius is fading out due to the solar angle.  Clouds predicted for foreseeable future, so unlikely to continue this trend.

2018-12-18 Copernicus-Stadius-Erastosthenes copy.jpg

 

This one was 20ms gain 1 50% of 1000 frames (still a 38 Gb file shocked.gif).  Went with a less contrast-y look this time, goes against my deep sky processing impulses to make contrast where this is none.


  • Luc CATHALA likes this

#9 lakeorion

lakeorion

    Apollo

  • -----
  • topic starter
  • Posts: 1227
  • Joined: 03 Aug 2010
  • Loc: Lake Orion MI

Posted 20 December 2018 - 09:29 AM

And southern region from the 18th.

 

15 ms / gain 1 75% of 1000.

2018-12-18 Tycho copy.jpg

 

Computer rebooted last night (thank you Mr. Gates) interrupting processing of the next one.


Edited by lakeorion, 20 December 2018 - 05:41 PM.


#10 lakeorion

lakeorion

    Apollo

  • -----
  • topic starter
  • Posts: 1227
  • Joined: 03 Aug 2010
  • Loc: Lake Orion MI

Posted 21 December 2018 - 10:02 AM

And the north on the 18th.  I wanted a big chip for deep sky pictures and small pixels for planetary pictures.  I settled with the IMX 183 sensor thinking it was a little of both but, dang it's a lot of pixels to run through AS!3.

2018-12-18 Northern copy.jpg


Edited by lakeorion, 21 December 2018 - 04:42 PM.

  • Carol L likes this

#11 aeroman4907

aeroman4907

    Viking 1

  • -----
  • Posts: 760
  • Joined: 23 Nov 2017
  • Loc: Castle Rock, Colorado

Posted 21 December 2018 - 11:29 AM

And the north on the 18th.  I wanted a big chip for deep sky pictures and small pixels for planetary pictures.  I settled with the IMX 183 sensor thinking it was a little of both but, dang it's a lot of to run through AS!3.

The IMX 183 sensor is amazing, but it will take a huge amount of heavy lifting on your computer to process it through AS!3.  It takes my laptop a couple of hours per panel to process because I use some experimental settings as well.

 

Keep up the good work on the progress.  You've done a nice job not blowing out the highlights.  I'd surmise that for a 14" scope, you'll need to get better collimation and/or focus, even will less than average seeing.  I have an 8" scope, and under quite poor seeing conditions, I would estimate you should be able to get at least double the resolution I see in your posted images.  Here is the thread Tom Glenn started where I posted one of my images and quality graphs.  That being said, poor conditions will make proper collimation and focusing much, much more difficult.

 

https://www.cloudyni...cember-16-2018/

 

I don't know how good the seeing is in Michigan, but I suspect you'll have very few nights where you'll max out the capability of your 14" scope.  I would never dream of purchasing a 14" scope here in Colorado as there are only a few nights a year where I can get somewhat close to the limit of an 8" scope.  In the 11 months since I purchased my scope, we've never had a single time of excellent seeing.  Good seeing may occur on 4 to 5 nights a year here.



#12 lakeorion

lakeorion

    Apollo

  • -----
  • topic starter
  • Posts: 1227
  • Joined: 03 Aug 2010
  • Loc: Lake Orion MI

Posted 21 December 2018 - 05:11 PM

Meant to say it's a lot of pixels to run through AS!3.  My computer is an AMD 6 core, the data is on a 4x RAID 1 (striped) array.  Wasn't too expensive and was close to state of the art - 7 years ago.  And yes I have a tendency to click all the experimental features - the one most required is rotation because the scope is on a fork.  Without that one properly done the stack looks like I had six too many...  I also have to wonder if the rotation is killing a little of my resolution.  Maybe with a better frame rate I could get shorter 'grabs' and that would be less of an issue.  I tried an ROI to shrink the file size and up the frame rate but the frame rate didn't change.  Not sure if that's a camera / laptop USB / acquisition software issue.  I haven't had the time to check that out.

 

I read the other post already - good info there.  When the scope gets put into it's permanent location I'll work on better collimation.  It was donated to my club, just happens to be resting in my garage for a few months wink.gif.  The upside is that it was free, the downside is that 'accessories' are minimal.  I did scrape together a rear focuser but it still requires touching the mount and, yeah, focus is a matter of patience.

 

If and when I get to purchase one for me it will probably be a C8 with a motorized focuser on the back - my mount probably can't lug a C9.25.  I have a Moonlite motor for my little refractor and it's definitely the way to fly - 'look Ma no hands!'

 

In the meantime, I can say, "These are the best I've done so far."



#13 lakeorion

lakeorion

    Apollo

  • -----
  • topic starter
  • Posts: 1227
  • Joined: 03 Aug 2010
  • Loc: Lake Orion MI

Posted 21 December 2018 - 07:20 PM

OK now comes the dense question.  My FPS seems to be directly related to bit depth:  8 bit -> 10-11 fps / 16 bit -> 4-5 fps in both SharpCap and FireCapture.  Not quite half but there's probably some additional checksum overhead.

 

My camera has a 12 bit ADC.  So for the Moon, which is better?

  1. Capture at 16 bit -> sacrificing 25% of data bandwidth (and storage size) that my camera will never properly populate, or
  2. Capture at 8 bit -> sacrificing 25% of available intermediate intensity data?

For DSO photography the answer is easy - 16 bit because bandwidth is not an issue with multiple minute exposures.

For planetary (non lunar) I've already read the consensus is 8 bit because you're not really going to capture that much fine tonality anyways.

 

But for Lunar, tonal variation is part of the end goal.  I can see both sides.  What is the consensus of experts?



#14 lakeorion

lakeorion

    Apollo

  • -----
  • topic starter
  • Posts: 1227
  • Joined: 03 Aug 2010
  • Loc: Lake Orion MI

Posted 22 December 2018 - 12:22 AM

Well color me silly.  I'd been using the wrong sharpening in PI.  I wanted MMT but was using MLT.  One letter makes a large difference.  Not sure if that difference will show up after the shrinking and JPEG compression but here are the last three with a sharpening algorithm that doesn't destroy small detail.

2018-12-18 Copernicus-Stadius-Erastosthenes copy.jpg


  • Carol L likes this

#15 lakeorion

lakeorion

    Apollo

  • -----
  • topic starter
  • Posts: 1227
  • Joined: 03 Aug 2010
  • Loc: Lake Orion MI

Posted 22 December 2018 - 12:23 AM

2018-12-18 Tycho copy.jpg


  • Carol L likes this

#16 lakeorion

lakeorion

    Apollo

  • -----
  • topic starter
  • Posts: 1227
  • Joined: 03 Aug 2010
  • Loc: Lake Orion MI

Posted 22 December 2018 - 12:24 AM

2018-12-18 Northern copy.jpg


  • Carol L likes this

#17 Tom Glenn

Tom Glenn

    Apollo

  • -----
  • Posts: 1414
  • Joined: 07 Feb 2018
  • Loc: San Diego, CA

Posted 22 December 2018 - 01:12 AM

But for Lunar, tonal variation is part of the end goal.  I can see both sides.  What is the consensus of experts?

There's not a consensus, but what I can say about it is that for most people, you will never have the seeing conditions for 16 bit capture to be beneficial, even on the Moon.  It should be noted that no camera is sufficient to capture the dynamic range of the Moon in a single exposure.  The human eye has about 20 stops of dynamic range, which is why you can easily see both the earthshine side and illuminated side of the Moon at the same time.  It should also be noted that although the 16 bit mode records a file in 16 bits, the amount of real data is typically only 10-14 bits depending on the camera.  Most CMOS cameras only record in 10-12 bits if you select the 16 bit mode, but the file has to be saved in 16 bit format.  

 

The reason that extra bits don't typically matter is that atmospheric turbulence causes slight fluctuations in the tonal value that registers on each pixel.  This contributes to noise, and it doesn't take much noise to render the extra bit depth provided by 16 bit files to be meaningless.  Emil Kraaikamp (the creator of Autostakkert) had an analysis on this several years ago.  The summary is that unless you have almost zero noise in your images, which is only possible if you use low gain during truly extraordinary seeing (the likes of which most never see) then 16 bits is not going to do anything.  If there is some noise in the system, then 8 bits is just as good if you stack about 1000 frames.  This is also why going above 8 bits is never beneficial for planets, because the gain always has to be high in order to keep exposure low (which is not true for the Moon).

 

https://www.cloudyni...small-analysis/

 

 

 

If you do some reading on the web and in books, you will see that many people have traditionally recorded the Moon in 16 bits, but it was common in the past to only stack 100 or so frames out of maybe 1000.  Now with faster cameras and computers, it is more typical to record between 5000-10,000 frames and stack at least 1000, so this changes things (bit depth actually increases with stack depth....an 8 bit recording becomes more like 12 bits if there is some noise....which there usually is).

 

As always though, everyone has to experiment to come to their own conclusions. If you were going to work from a single raw image, or if for some reason you were limited to only a few dozen frames, you would want the 16 bit files for sure.  But when you stack 1000 frames or so, the bit depth of the 8-bit files increases to basically match the same bit depth of a 16 bit recording.  

 

There are significant disadvantages to recording in 16 bits with the IMX183 sensor.  One is the frame rate, which will almost certainly drop to single digits.  The other is the file size.  Instead of 20GB per 1000 frames, the files are 40GB per 1000 frames.  Even in truly excellent seeing (rare, even here in San Diego), I like to capture at least 2000-3000 frames per file.  For a full frame shot, this is 80-120 GB file size in 16 bits, but only 40-60GB in 8 bits. When capturing several files in a session this size does matter from a practical standpoint.  However, in more typical seeing conditions, I like to capture 5000 frames.  16 bit recordings turn this from 100GB into 200GB, and this fills up hard drives fast.  When I first started using this camera, I recorded in 16 bits, but now I do 8 bits because of the file sizes.  I haven't noticed any meaningful difference between my 8 bit and 16 bit processed recordings.  

 

But these are just practical matters.  If you had a relatively unlimited number of external SSDs (required to avoid snail pace file transfers) as well as an internal SSD hard drive of at least 2 TB, I would probably choose to record in 16 bits if I knew the seeing was extraordinary, as Emil's analysis does show that if the noise is very low, then 8 bit recordings can fail to capture all of the tonal data.  So it becomes a bit of circular logic......8 bits will be just as good if you can stack enough frames and you have some noise........BUT you might not be able to stack enough frames if seeing isn't excellent........BUT if seeing isn't excellent you won't be getting a good result anyway and any bit depth over 8 is meaningless........AND if you take 16 bit files this limits your frame rate and increases your files sizes (which limits frames!!!!). Not so simple.  Bottom line, 16 bit captures are probably better on the Moon if you have extraordinary seeing and have the hard drive space to spare and you computer can handle the frame rates.  


Edited by Tom Glenn, 22 December 2018 - 01:21 AM.


#18 Tom Glenn

Tom Glenn

    Apollo

  • -----
  • Posts: 1414
  • Joined: 07 Feb 2018
  • Loc: San Diego, CA

Posted 22 December 2018 - 02:20 AM

I should also add, because I didn't state it above, that increasing the bit depth will only (potentially) allow you to stretch the histogram without incurring as much quantization error in the shadowed regions.  It won't affect your ability to resolve fine detail, unless that detail happened to be buried in shadow.  But in most circumstances, you can improve the fine detail of your images more by capturing additional frames than by going to 16 bit, and there are other ways to address quantization error.  If I have to choose between capturing 2500 frames in 16 bit and 5000 in 8 bit, I would choose 8 bit unless the seeing is of the jaw dropping variety.  As in perfect Airy pattern during collimation, and unwavering image in the live feed (maybe a lunar equivalent of Ferret's Saturn video here).  



#19 aeroman4907

aeroman4907

    Viking 1

  • -----
  • Posts: 760
  • Joined: 23 Nov 2017
  • Loc: Castle Rock, Colorado

Posted 22 December 2018 - 09:44 AM

Good to know Tom.  I've been recording 16 bit at 6.8 fps.  It's real hard to get 3000 frames before movement of the moon in that 7.5 minutes starts to constrict my frame sizing from overlap of images to stack in AS!3.  I already know I don't experience truly exceptional seeing here and likely never will.  I take it you capture 8bit AVI's?  Would there be any difference as I capture in color?  Thanks for your input as always!



#20 Tom Glenn

Tom Glenn

    Apollo

  • -----
  • Posts: 1414
  • Joined: 07 Feb 2018
  • Loc: San Diego, CA

Posted 22 December 2018 - 10:50 AM

Good to know Tom.  I've been recording 16 bit at 6.8 fps.  It's real hard to get 3000 frames before movement of the moon in that 7.5 minutes starts to constrict my frame sizing from overlap of images to stack in AS!3.  I already know I don't experience truly exceptional seeing here and likely never will.  I take it you capture 8bit AVI's?  Would there be any difference as I capture in color?  Thanks for your input as always!

I take Ser files, because AVIs have a size limit in Firecapture and cause it to crash.  Ser doesn't have this problem for me.

 

Based on what you have told me about your frame rate, and your seeing conditions (and the impact and limitations this has already), I wouldn't even consider taking 16 bits files unless you were only going to work from between one and a dozen frames.  There's no benefit.  I may have said something different in the past, but this is my current feeling.....at least until someone proves me wrong.  The IMX183 is capturing in 12 bits if you use the 16-bit mode, but as Emil explained in his post, unless you have noise-free raw frames, the first several bits are completely consumed by noise, rendering the extra bits useless.  So you get no benefit, but all the downsides to the larger files.  And as I said, since I started taking mostly 8-bit captures, I haven't noticed any impact on the outcome.  In the past year, I can count on one hand the number of CN posts I've seen with lunar images taken in conditions in which people should even entertain the idea of 16 bit captures.  


Edited by Tom Glenn, 22 December 2018 - 02:04 PM.



CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics