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

Fixed tripod and lucky imaging

  • Please log in to reply
13 replies to this topic

#1 Mrcloc

Mrcloc

    Vendor - SvBony Eyepieces

  • -----
  • Vendors
  • topic starter
  • Posts: 188
  • Joined: 19 Jun 2020
  • Loc: Pretoria, South Africa

Posted 15 June 2021 - 02:04 AM

I'm really a visual astronomy guy, but I do love a good experiment. I can't afford a mount which can track, let alone one which can do resolution I'm looking for for this experiment.

 

I was wondering, would it work if I used the lucky imaging technique with a fixed tripod? So I would take 5s video, move the scope, etc. I think perfect polar alignment would go a long way in making this possible... Then I could track the frame as if the scope was moving. I believe it would be possible to calculate how many pixels to move by for each frame using some sort of algorithm. Either that, or split the video into individual frames and align them. Again, perfect polar alignment would make this task so much easier.

 

I would like to try this, but I suspect I might need a very high ISO. Unfortunately I don't have a good sensor for high ISO (although it's not that bad), but as far as I understand, the signal increases proportionally to the number of frames while the noise increases with the square root of the number of frames, or something like that, so a few hours worth of good frames will be what I need. Be that as it may, I'm not sure a moving object would provide enough signal as it passes over the sensor pixels for the short time it does. Do you need the sensor to have the image static so that it can pick up the image? Or will high ISO work?



#2 Hesiod

Hesiod

    Aurora

  • *****
  • Posts: 4,844
  • Joined: 13 Jan 2013

Posted 15 June 2021 - 05:41 AM

A photographic tripid or alt/az mount can not (and need not to be) aligned to the pole.
Untracked imaging of deep sky is feasible, but usually is done at very short focals (most of Milky way shots are taken without tracking devices).
Anyway there is a very skilled guy (the one who wrote autostakkert) who did nice lucky imaging shots of galaxies.
He uses much shorter exposures (1" or less), "planetary" or other high framerate cameras and his software to stack and mix the frames.
He uses rudimentary tracking as well, but also quite large telescopes (e.g. 16" Dob) so could try w/o drive if have a smallish telescope
  • Mrcloc likes this

#3 Mrcloc

Mrcloc

    Vendor - SvBony Eyepieces

  • -----
  • Vendors
  • topic starter
  • Posts: 188
  • Joined: 19 Jun 2020
  • Loc: Pretoria, South Africa

Posted 15 June 2021 - 06:17 AM

Thanks for the useful info. The telescope is on an EQ mount, and having it polar aligned makes stacking easier (at least for me with my limited experience) because the orientation of the object doesn't change.

Interesting that he uses 1s or less. It means I probably can get away with tracking the frame rather than live tracking. I'm going to give this a try sometime, and play around with Autostakkert. I think my weakest point at the moment is my ability to use Autostakkert. I haven't had much luck figuring it out yet. I think if I just put in the time, I'll master it.



#4 james7ca

james7ca

    Cosmos

  • *****
  • Posts: 9,832
  • Joined: 21 May 2011
  • Loc: San Diego, CA

Posted 15 June 2021 - 06:49 AM

A photographic tripid or alt/az mount can not (and need not to be) aligned to the pole...

That type of mount won't work very well for long sequences since you'll get field rotation. Post processing registration can handle the rotation, but you'll still end up losing the corners of your frame since they will have unequal exposure coverage.

 

That said, I've done long exposure sequences on a simple, non-motorized alt/az mount and I've posted images from those sessions here on CN. Below are links to two of those images (somewhat similar to what the OP is proposing, using short sequences where the object drifts, you then reposition the mount, then another short sequence of frames, etc.):

 

  https://www.cloudyni...d/#entry8881987

 

And, one on the Orion Nebula using the same technique:

 

  https://www.cloudyni...d/#entry8888970



#5 klaussius

klaussius

    Apollo

  • *****
  • Posts: 1,272
  • Joined: 02 Apr 2019
  • Loc: Buenos Aires

Posted 15 June 2021 - 07:05 AM

I have tried this, and it's possible, but challenging.
 
Here my first (and only) attempt:

get.jpg?insecure
 

In terms of resolution, it blows long exposures out of the water. But getting the SNR is quite the challenge. And since you'll be wanting to apply deconvolution to extract the maximum resolution allowed by the diffraction limit of your system, a high SNR is paramount.

 

My limiting factor ATM with my astroberry-based setup, is storage and storage speed. Taking that many subs that quickly can easily use up all the available storage and bandwidth in a low-power computer like a Raspberry Pi.

 

A full-blown laptop may fare better, but then you'll have to stack all those subs, which is no small feat. At a minimum frame rate of 1 subs per second (any less and you don't overcome seeing often enough to be worth it), read noise is high, and you need hours and hours of total integration time, easily racking up 10k images or more. At 24MB per image, that's 24GB of data to stack.

 

So, the above prevented me from improving on that experiment, but as you can see, the resolution gain is quite considerable.


Edited by klaussius, 15 June 2021 - 07:07 AM.


#6 jdono

jdono

    Sputnik

  • -----
  • Posts: 47
  • Joined: 12 Mar 2017

Posted 15 June 2021 - 08:31 AM

From what I understand shot noise decreases based on the square root of the total integration time and read noise decreases based on the square root of the number of frames taken.

So ideally you would have a bright image and a camera that has low read noise to begin with for this. And with many cameras the read noise decreases with the ISO so that could help out a bit. I assume blowing out the image won't be too likely with very short exposure times but you would have to experiment with this

And like someone mentioned previously file sizes can be an issue. I suppose you would want the camera with the smallest frame that captures your image if possible too (but with low read noise, not sure if they make them like this as I would think that cameras optimized for lucky shot viewing would have small file sizes but tend not to care about read noise too much as they are in general planetary cameras)

 

One additional issue might be if there is a delay between the frames then that would really cut down on total integration time. For some reason I think there is (so if you take .5 sec exposures and there is a 1 sec delay between exposures then thats a 2/3 decrease in exposure time). maybe SER cuts down on this?

 

I've done imaging with alt-az and you can indeed do it and it works OK but obviously not even close to some of the great images you see on here.  One thing to watch out for is that if there are any aberrations they will get rotated around and be smeared all around the image making a semicircle or circle, vs being located in one spot. so good calibration frames (esp flats) are necessary (which often causes problems for me). the corners won't necessarily be lost but they will have less total exposure time as stacking can preserve them but just not use the images that don't have the corners.

 

for me its sort of fun experimenting like this, it seems worth a try at least with something like orion or a bright globular cluster. It wouldn't be great for really faint details though



#7 Alex McConahay

Alex McConahay

    Voyager 1

  • *****
  • Posts: 10,415
  • Joined: 11 Aug 2008
  • Loc: Moreno Valley, CA

Posted 15 June 2021 - 12:40 PM

Mrcloc,

 

I see in post #3 that "The telescope is on an EQ mount." 

 

If so, why are you not using that instead of trying to use a fixed tripod? 

 

Even an undriven (non motorized) equatorial mount is better than a fixed tripod. 

 

So, are you asking about a camera attached to something on an undriven eq mount, a driven eq mount, or a fixed tripod?

 

 

Alex



#8 Trenor

Trenor

    Vostok 1

  • *****
  • Posts: 118
  • Joined: 01 Mar 2021

Posted 15 June 2021 - 01:49 PM

I started out doing a tripod, 200mm Canon lens, and a Canon 5Dmark2. I took this image of Orion from 1015 X 1.3 second subs moving the camera manually to reframe every 10-15 mins. This was my "first" stacked image. I did have to crop pretty heavy once I got done stacking since there was a lot of overlap. With the wide field though it wasn't a much of a problem.

 

get.jpg?insecure

 

If you are using a polar aligned EQ mount you should be able to take much longer exposures which will improve your image.



#9 james7ca

james7ca

    Cosmos

  • *****
  • Posts: 9,832
  • Joined: 21 May 2011
  • Loc: San Diego, CA

Posted 15 June 2021 - 02:22 PM

Mrcloc,

 

I see in post #3 that "The telescope is on an EQ mount." 

 

If so, why are you not using that instead of trying to use a fixed tripod? 

 

Even an undriven (non motorized) equatorial mount is better than a fixed tripod. 

 

So, are you asking about a camera attached to something on an undriven eq mount, a driven eq mount, or a fixed tripod?

 

 

Alex

I think the OP's use of the term "fixed tripod" may have confused some of the responders. By "fixed" I think the OP's meaning was a mount without a motor drive. So, an equatorial mount and if polar aligned then that would be a much better case than trying to use an alt/az mount.

 

When I'm doing some quick imaging without guiding on my Celestron AVX I will sometimes let it track for a few minutes capturing a short sequence using one or two second exposures and then I will recenter a reference star manually to prevent field drift. This way the object stays well centered with very little walking noise in the final stack.



#10 Trenor

Trenor

    Vostok 1

  • *****
  • Posts: 118
  • Joined: 01 Mar 2021

Posted 15 June 2021 - 03:06 PM

I think the OP's use of the term "fixed tripod" may have confused some of the responders. By "fixed" I think the OP's meaning was a mount without a motor drive. So, an equatorial mount and if polar aligned then that would be a much better case than trying to use an alt/az mount.

 

When I'm doing some quick imaging without guiding on my Celestron AVX I will sometimes let it track for a few minutes capturing a short sequence using one or two second exposures and then I will recenter a reference star manually to prevent field drift. This way the object stays well centered with very little walking noise in the final stack.

That makes more sense now. With that setup they should be able to get decent exposures and have better alignment requiring less cropping once stacked.



#11 rj144

rj144

    Viking 1

  • -----
  • Posts: 749
  • Joined: 31 Oct 2020

Posted 15 June 2021 - 10:28 PM

I have tried this, and it's possible, but challenging.
 
Here my first (and only) attempt:

get.jpg?insecure
 

In terms of resolution, it blows long exposures out of the water. But getting the SNR is quite the challenge. And since you'll be wanting to apply deconvolution to extract the maximum resolution allowed by the diffraction limit of your system, a high SNR is paramount.

 

My limiting factor ATM with my astroberry-based setup, is storage and storage speed. Taking that many subs that quickly can easily use up all the available storage and bandwidth in a low-power computer like a Raspberry Pi.

 

A full-blown laptop may fare better, but then you'll have to stack all those subs, which is no small feat. At a minimum frame rate of 1 subs per second (any less and you don't overcome seeing often enough to be worth it), read noise is high, and you need hours and hours of total integration time, easily racking up 10k images or more. At 24MB per image, that's 24GB of data to stack.

 

So, the above prevented me from improving on that experiment, but as you can see, the resolution gain is quite considerable.

Isn't this usually done with frames from video like planetary imaging though?  Are you actually taking full resolution pics one at a time?

 

10k of photos at 24 MB each is 240 GB, by the way.


Edited by rj144, 15 June 2021 - 10:30 PM.


#12 klaussius

klaussius

    Apollo

  • *****
  • Posts: 1,272
  • Joined: 02 Apr 2019
  • Loc: Buenos Aires

Posted 16 June 2021 - 09:09 AM

Isn't this usually done with frames from video like planetary imaging though?  Are you actually taking full resolution pics one at a time?

 

10k of photos at 24 MB each is 240 GB, by the way.

 

Oops... forgot a 0  wink.gif

 

I took full resolution video, at ~1fps. Can you call that video?

 

The Raspberry Pi can't handle full resolution full speed video from that camera (QHY 163m), but the camera itself can't do a lot more at full resolution anyway (specs say 22FPS but I could never achieve that). It doesn't matter because for DSO you can't lower the exposure time a lot more. Below 800ms I couldn't see much, the signal was too low.


Edited by klaussius, 16 June 2021 - 09:14 AM.


#13 rj144

rj144

    Viking 1

  • -----
  • Posts: 749
  • Joined: 31 Oct 2020

Posted 16 June 2021 - 09:37 AM

Oops... forgot a 0  wink.gif

 

I took full resolution video, at ~1fps. Can you call that video?

 

The Raspberry Pi can't handle full resolution full speed video from that camera (QHY 163m), but the camera itself can't do a lot more at full resolution anyway (specs say 22FPS but I could never achieve that). It doesn't matter because for DSO you can't lower the exposure time a lot more. Below 800ms I couldn't see much, the signal was too low.

Cool, but your stills from the video were 24 MB?



#14 klaussius

klaussius

    Apollo

  • *****
  • Posts: 1,272
  • Joined: 02 Apr 2019
  • Loc: Buenos Aires

Posted 16 June 2021 - 12:43 PM

Cool, but your stills from the video were 24 MB?

Images are 4656 x 3522 x 16bit = 32MB

 

24MB compressed on average (gzipped).


Edited by klaussius, 16 June 2021 - 12:44 PM.

  • rj144 likes this


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