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

AutoStakkert! 3

astrophotography
  • Please log in to reply
86 replies to this topic

#51 MvZ

MvZ

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1872
  • Joined: 03 Apr 2007
  • Loc: The Netherlands

Posted 30 March 2017 - 10:30 AM

Easy enough to just try it out I guess. Customers? The software is free, so you have nothing to lose by just trying it out!

 

I do suspect it can be a bit tricky though to do binary star stacking. Speckle patterns is not not the typical input of AS!...

 

@zAmbonii

You should indeed force a bayer pattern. It is the only way to be sure of what you get. Guessing the bayer pattern from pixel statistics only is difficult and sometimes fails. If you record in SER format, the bayer pattern should be encoded in the header file, which means it should default to using what the header says the bayer pattern is.There may be a bug there if this is not working...



#52 RedLionNJ

RedLionNJ

    Mercury-Atlas

  • *****
  • Posts: 2964
  • Joined: 29 Dec 2009
  • Loc: Red Lion, NJ, USA

Posted 30 March 2017 - 01:47 PM

Emil I have a question.  Would Autostakkert be suitable for stacking images of Binary Stars?  I do lucky imaging of binary stars with the ultimate goal of measurement of the pairs.  The camera is an ASI 224 MC.....

 

Have any of your customers, that you know of, used Autostakkert in this way?

 

Or, has anyone reading this thread stacked double star videos with Autostakkert?

 

Thanks,

 

Dave

I've done this from time to time - it works awesomely on stars, including very close doubles.

 

Last night, I had clear-ish skies but seeing was only so-so, even at high elevation. I played around with my 12-inch LX200ACF at f/7, using an ASI174MM for capture. Gain set to max (400), exposures of varying lengths. In particular, I spent a fair bit of time on two fields - around Mizar and around M97 (for different experimental purposes). For the Mizar field, I used mostly very short exposures, else Mizar A & Mizar B merged into one blob. I then took some longer exposures of the entire field for a later HDR merge attempt.

I realize Mizar is bright and relatively wide, so it's not quite in the same ballpark as you're likely referring to, Dave. But stacking just a very few frames does give nice, round stars with measurable centroids. I was able to measure Mizar A & B (with little difficulty) from my stack to be 14.3 arcsec separation. I could probably have gotten it more precise, but that wasn't really my purpose at the time.

 

The bright 8th mag star between Mizar and Alcor was coming across so brightly at first I thought it WAS Alcor, but the rest of the field stars betrayed its true identification. All the 14th and 15th mag stars catalogued in the field were readily apparent with really short exposures - no guiding - great work, Emil.

 

Grant


  • Lacaille likes this

#53 Cotts

Cotts

    Just Wondering

  • *****
  • Moderators
  • Posts: 8920
  • Joined: 10 Oct 2005
  • Loc: Madoc, Ontario

Posted 30 March 2017 - 02:57 PM

Emil, I guess I meant clients or users or......  Anyway thanks for getting back to me.

 

Grant, I am encouraged by your response.  Thanks

 

Dave


  • RedLionNJ likes this

#54 yock1960

yock1960

    Aurora

  • *****
  • Posts: 4642
  • Joined: 22 Jun 2008
  • Loc: (Mostly Cloudy) Ohio, USA

Posted 30 March 2017 - 04:17 PM


@zAmbonii

You should indeed force a bayer pattern. It is the only way to be sure of what you get. Guessing the bayer pattern from pixel statistics only is difficult and sometimes fails. If you record in SER format, the bayer pattern should be encoded in the header file, which means it should default to using what the header says the bayer pattern is.There may be a bug there if this is not working...

 

In my experience, AS is pretty good at getting it right, >95%. Perhaps it has something to do with capture settings.

 

Steve


  • zAmbonii likes this

#55 MvZ

MvZ

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1872
  • Joined: 03 Apr 2007
  • Loc: The Netherlands

Posted 30 March 2017 - 05:22 PM

>Perhaps it has something to do with capture settings.

Yes; things like color balance will have a big effect for example. You have to realize that to AS! - without any additional meta-data - a raw bayer pattern is just a bunch of grayscale pixels that sometimes looks to have a bit of a grid like appearance, meaning that on average every pixel on every odd row and ever odd column has a slightly different brightness and or noise characteristics then other odd/even combinations. With all traditional bayer patterns you have two green pixels, and one red and one blue for every 2x2 block in the image, guessing the two green pixels is most of the time not too difficult - they should have VERY similar noise AND intensity characteristics, but the red and blue are more likely to result in mistakes.

 

So either use the SER format that stores this bayer pattern in the metadata of the file, or force the correct bayer pattern. 

 

Few more bugs fixed, including one important one for those wanting to stack tifs or pngs or jpgs in some funny formats (there was a memory leak when converting some formats):

http://www.astrokraa...ware/latest.php

 

>Emil, I guess I meant clients or users or......  Anyway thanks for getting back to me.

Ooh, clients sound interesting too ;)

 

(I'll be decreasing the update time from now on, but please keep sending bug reports)


Edited by MvZ, 30 March 2017 - 05:24 PM.


#56 zAmbonii

zAmbonii

    Mercury-Atlas

  • -----
  • Posts: 2516
  • Joined: 19 Jan 2008
  • Loc: AZ

Posted 30 March 2017 - 11:15 PM

>Perhaps it has something to do with capture settings.

Yes; things like color balance will have a big effect for example. You have to realize that to AS! - without any additional meta-data - a raw bayer pattern is just a bunch of grayscale pixels that sometimes looks to have a bit of a grid like appearance, meaning that on average every pixel on every odd row and ever odd column has a slightly different brightness and or noise characteristics then other odd/even combinations. With all traditional bayer patterns you have two green pixels, and one red and one blue for every 2x2 block in the image, guessing the two green pixels is most of the time not too difficult - they should have VERY similar noise AND intensity characteristics, but the red and blue are more likely to result in mistakes.

 

So either use the SER format that stores this bayer pattern in the metadata of the file, or force the correct bayer pattern. 

 

Few more bugs fixed, including one important one for those wanting to stack tifs or pngs or jpgs in some funny formats (there was a memory leak when converting some formats):

http://www.astrokraa...ware/latest.php

 

>Emil, I guess I meant clients or users or......  Anyway thanks for getting back to me.

Ooh, clients sound interesting too wink.gif

 

(I'll be decreasing the update time from now on, but please keep sending bug reports)

Thx!  I didn't realize that it was a "best guess" on the matrix pattern.  I usually force GRBG anyways, just passing along the info.  LIke yock said, AS! is usually pretty accurate in determining the right color pattern.

 

I noticed a problem earlier today with the beta.  My system was quite a bit sluggish and I opened up the task manager and AS!3 didn't unload properly when I had shut it down, had to force exit it from the manager.  Not sure which version it was that did it though since I was using both 3.0.11 and 3.0.12 last night.  Will try out the new beta.



#57 MvZ

MvZ

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1872
  • Joined: 03 Apr 2007
  • Loc: The Netherlands

Posted 31 March 2017 - 05:16 AM

>AS!3 didn't unload properly when I had shut it down,

You mean there was an autostakkert.exe still running (but no window?), weird. If I had to guess there was an unexpected error in one of the processing threads (it should not happen of course, at least it should force close all running threads..). I have not seen this myself. If you find a away to replicate this, let me know.



#58 zAmbonii

zAmbonii

    Mercury-Atlas

  • -----
  • Posts: 2516
  • Joined: 19 Jan 2008
  • Loc: AZ

Posted 31 March 2017 - 05:40 AM

Yea, it was chewing through CPU cycles and was causing other programs some grief.  It was not that high in memory usage, only about 1.5GB, not outside the normal usage.  I don't remember seeing an error window or anything wrong with the program when running though.  Weird.



#59 maxmir

maxmir

    Mariner 2

  • *----
  • Posts: 213
  • Joined: 03 Oct 2005

Posted 31 March 2017 - 09:06 AM

Emil I have a question.  Would Autostakkert be suitable for stacking images of Binary Stars?  I do lucky imaging of binary stars with the ultimate goal of measurement of the pairs.  The camera is an ASI 224 MC.....

 

Have any of your customers, that you know of, used Autostakkert in this way?

 

Or, has anyone reading this thread stacked double star videos with Autostakkert?

 

Thanks,

 

Dave

Can Autostakket3 be used for lucky deepsky imaging?

Seems like there should be something aimed at this process specifically since the images are much dimmer.



#60 Quaternion

Quaternion

    Messenger

  • -----
  • Posts: 407
  • Joined: 08 Dec 2011
  • Loc: USA

Posted 31 March 2017 - 01:39 PM

Emil,

Would your AS!3 be able to stack and sharpen planetary nebula, like the ring or helix? I'm thinking of trying to image these with short (250ms - 500ms) exposures, which my homemade EQ platform might be steady enough to handle. Would the automatic selection of alignment points work for planetary nebula or other DSOs?

Thanks. Sorry to bother you.

Tom P.

#61 Sunspot

Sunspot

    Cosmos

  • *****
  • Posts: 8456
  • Joined: 15 Mar 2005
  • Loc: Surprise, AZ

Posted 31 March 2017 - 01:42 PM

Emil,

Would your AS!3 be able to stack and sharpen planetary nebula, like the ring or helix? I'm thinking of trying to image these with short (250ms - 500ms) exposures, which my homemade EQ platform might be steady enough to handle. Would the automatic selection of alignment points work for planetary nebula or other DSOs?

Thanks. Sorry to bother you.

Tom P.

I stacked images of the Orion Nebula with AS!2 a while back. Best thing is to try and see and if you are successful you can become a resource for others. bow.gifbow.gif


  • bunyon and eekeek77 like this

#62 bunyon

bunyon

    Aurora

  • *****
  • Posts: 4580
  • Joined: 23 Oct 2010
  • Loc: Winston-Salem, NC

Posted 31 March 2017 - 07:41 PM

In addition to all the official planets except Earth, the Moon and the Sun I have used AS!2 on multiple stars, planetary nebulae and other DSOs.

 

It's very good on planetaries and the few times I've tried it with multiple stars it works quite well.  I highly recommend anyone with a bunch of frames of these objects give AS!3 a try.

 

I haven't used AS!3 yet not because I don't want to but because I'm on an extended trip and have no data to use.  But I look forward to it.


  • RedLionNJ likes this

#63 Stargazer3236

Stargazer3236

    Gemini

  • *****
  • Posts: 3004
  • Joined: 07 Aug 2010
  • Loc: Waltham, MA

Posted 01 April 2017 - 08:42 AM

I have used AS!2 extensively with my captured images of the planets and planetary nebula and emission nebula and a few globular clusters too. I even used it on the Mercury transit last May when it transited the Sun. That is a very robust image stacking software suite. I can throw numerous frames at it, somewhere upwards of 14000 frames and it just laughs at me and spits out a nice image.

 

I wonder how good AS!3 will be?

 

On another note, I tried out Firecapture 2.5 and did not like it. I erased it off my hard rive the other day in favor of Firecapture 2.4.



#64 wargrafix

wargrafix

    Gemini

  • -----
  • Posts: 3321
  • Joined: 10 Apr 2013
  • Loc: Trinidad

Posted 02 April 2017 - 01:43 PM

As2! Has varied success with deepsky, but generally it handles it well

#65 GeorgeInDallas

GeorgeInDallas

    Viking 1

  • -----
  • Posts: 581
  • Joined: 15 Apr 2012
  • Loc: Dallas, Texas USA

Posted 02 April 2017 - 04:06 PM

Emil,

Would your AS!3 be able to stack and sharpen planetary nebula, like the ring or helix? I'm thinking of trying to image these with short (250ms - 500ms) exposures, which my homemade EQ platform might be steady enough to handle. Would the automatic selection of alignment points work for planetary nebula or other DSOs?

Thanks. Sorry to bother you.

Tom P.

I believe Emil has numerous examples of deep sky images with 1 to 4 sec. exposures stacked with autostakkert on his web site.

 

http://www.astrokraa...?t=y&category=7

 

George


  • RedLionNJ likes this

#66 kbev

kbev

    Viking 1

  • -----
  • Posts: 887
  • Joined: 29 Dec 2010
  • Loc: Far, far east Mesa

Posted 05 April 2017 - 05:26 AM

Emil, tonight I was processing some videos of Jupiter and came across an IR capture that had doubled/split frames.  Since it seemed like something the new beta with improved hard edge detection should be able to handle I downloaded v3.0.14 and tried it.  Here is a screen capture of the result:

 

double_frame.jpg

 

Not sure if you'll be able to get the code to where it will detect frames like this but I thought you would want to know.  I can still hit spacebar to lock the frame out of processing but as mentioned earlier if there are multiple frames like this it's a bit of a pain to hunt them all down and remove them from the queue. 

 

I was going to add a dropbox link to a copy of the video file but adding the file puts me over the 2GB limit for my dropbox account.  I'll see if I can add it to my Google Drive storage and link from there.


Edited by kbev, 05 April 2017 - 05:32 AM.


#67 MvZ

MvZ

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1872
  • Joined: 03 Apr 2007
  • Loc: The Netherlands

Posted 05 April 2017 - 06:35 AM

Send me a link sometime and I'll have a look.



#68 happylimpet

happylimpet

    Mercury-Atlas

  • *****
  • Posts: 2945
  • Joined: 29 Sep 2013
  • Loc: Southampton, UK

Posted 05 April 2017 - 07:41 AM

Emil I have a question.  Would Autostakkert be suitable for stacking images of Binary Stars?  I do lucky imaging of binary stars with the ultimate goal of measurement of the pairs.  The camera is an ASI 224 MC.....

 

Have any of your customers, that you know of, used Autostakkert in this way?

 

Or, has anyone reading this thread stacked double star videos with Autostakkert?

 

Thanks,

 

Dave

The trick is the quality selection as the standard methods (as were - gradient and edge) often 'reverse order' the quality with the worst images first, probably because they contain a wealth of high contrast speckles.  After much trial and error I found a setting of gradient=8 seemed to produce the best ordering.



#69 Stargazer3236

Stargazer3236

    Gemini

  • *****
  • Posts: 3004
  • Joined: 07 Aug 2010
  • Loc: Waltham, MA

Posted 05 April 2017 - 12:11 PM

I have used Autostakkert!2 for stacking images of double stars. It works pretty good!



#70 EdwardMH

EdwardMH

    Lift Off

  • -----
  • Posts: 22
  • Joined: 12 Jun 2011
  • Loc: Battlefield, Mo

Posted 06 April 2017 - 02:17 PM

Does Autostakkert rely more on Faster CPU or GPU? At this time I mostly am learning Solar imaging I Will be building a new system and I am thinking

i7 7700k 4.2ghz,

Nvidia 1080Ti,

32GB ram and  dedicating one of the m.2 slots to Astro imaging

Samsung 960 PRO M.2 SSD Specifications

Interface: PCIe 3.0 x 4, NVMe 1.2 (Partial)
Capacities: 512GB, 1TB (1,024GB), 2TB (2,048GB)
DRAM Cache Memory: 512MB LPDDR3, 1GB LPDDR3, 2GB LPDDR3
Form Factor: M.2(2280)
Performance
Sequential read: 3,500MB/s
Sequential write: 2,100MB/s

should never have to worry about bottlenecks between storage and camera. Then move finished processed images to a mechanical drive.



#71 RedLionNJ

RedLionNJ

    Mercury-Atlas

  • *****
  • Posts: 2964
  • Joined: 29 Dec 2009
  • Loc: Red Lion, NJ, USA

Posted 06 April 2017 - 04:11 PM

I'd be really, really surprised if Autostakkert cares about the graphics processing unit. Nothing it does is display performance relevant.

 

CPU and RAM is where it's all at.


  • WarmWeatherGuy likes this

#72 MvZ

MvZ

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1872
  • Joined: 03 Apr 2007
  • Loc: The Netherlands

Posted 06 April 2017 - 04:56 PM

AS!3 indeed doesn't care about GPU. Go for lots of ram, fast storage, and many cpu cores (I'll wait a bit for testing the new AMD Ryzen CPU, but I have high hopes for the performance of this with AS! to be honest)

 

I personally also quite like to put a couple of cheap(er) SSD drives in raid 0. It's not as fast as the NVMe stuff, but getting +- 1000 MB/s on read/write for half the price (or double the capacity) is still quite nice. 


  • R Botero and RedLionNJ like this

#73 bratislav

bratislav

    Viking 1

  • *****
  • Posts: 640
  • Joined: 07 Sep 2006
  • Loc: -

Posted 06 April 2017 - 09:20 PM

Another round of applause to Emil for a continuous development of already fabulous product. 

 

Planetary imaging would not be where it is today (and we are experiencing golden era of imaging where any decent seeing delivers all that telescope is capable of, and I mean ALL) without Autostakkert.

 

My hat off to you sir (again)! bow.gif

 

Bratislav


  • R Botero, happylimpet and RBChris like this

#74 MvZ

MvZ

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1872
  • Joined: 03 Apr 2007
  • Loc: The Netherlands

Posted 07 April 2017 - 04:13 AM

Compliments make me very lazy, so if you want me to stop working, please continue ;)



#75 Sunspot

Sunspot

    Cosmos

  • *****
  • Posts: 8456
  • Joined: 15 Mar 2005
  • Loc: Surprise, AZ

Posted 07 April 2017 - 09:23 AM

Compliments make me very lazy, so if you want me to stop working, please continue wink.gif

Right then, this software is a piece of ...lol.gif

 

So, get back to work my friend! waytogo.gif


  • kbev 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





Also tagged with one or more of these keywords: astrophotography



Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics