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

JetsonSky : software for Deep Sky videos

  • Please log in to reply
164 replies to this topic

#151 easybob95

easybob95

    Viking 1

  • -----
  • Vendors
  • topic starter
  • Posts: 703
  • Joined: 23 Apr 2014
  • Loc: France

Posted 08 December 2024 - 10:20 AM

Hello,

 

i have uploaded a new version of JetsonSky in my Github : V51_01RC

 

https://github.com/A...illou/JETSONSKY

 

I have also uploaded a brand new and up to date JETSONSKY_V51_01RC_INFOS.pdf file to explain changes in JetsonSky. I recommend the reading of this file.

 

This version of JetsonSky now support 16 bits video reading using SER format files (JetsonSky needs now Serfile library which is in my Github; thanks to Jean-Baptiste Butet who wrote this library ; i made some small modifications for JetsonSky).

 

16 bits mode will allow to recover small signal and to make HDR image with only 1 image even with post treatment (8 bits HDR mode needs several captures using different exposure times).

 

If you want to make some tests with a SER 16 bits file, you can find one here :

 

https://drive.google...iew?usp=sharing

 

This is a RAW capture so you will have to debayer it using JetsonSky with RGGB pattern.

 

I did not test this version of JetsonSky with the AGX Orin. I will test it ASAP.

 

For Windows version : i had some serious issue with CUDA SDK V12.5 and above. I had to get back to V12.1 to solve those issues (problem with nvcc compiler; impossible to compile CUPY kernels with V12.5 or above).

 

Alain


Edited by easybob95, 08 December 2024 - 10:35 AM.


#152 easybob95

easybob95

    Viking 1

  • -----
  • Vendors
  • topic starter
  • Posts: 703
  • Joined: 23 Apr 2014
  • Loc: France

Posted 30 December 2024 - 10:32 AM

Hello,

 

i made some changes to improve camera capture and i also get better HDR management with 16 bits capture or 16 bit SER files.

 

I also have 2 HDR modes :

- Mertens (slow)

- Mean (fast).

 

The new version is V52_03RC and it can be found on my Github :

 

https://github.com/A...illou/JETSONSKY

 

Alain



#153 easybob95

easybob95

    Viking 1

  • -----
  • Vendors
  • topic starter
  • Posts: 703
  • Joined: 23 Apr 2014
  • Loc: France

Posted 15 January 2025 - 12:58 PM

Hello,

 

still developing JetsonSky. V52_08RC is in debugging test.

 

Yesterday, no clouds (a kind of small miracle here) but very bad transparency and full Moon : worst conditions for deep sky objects.

 

I wanted to test 16bits captures and HDR on M45, M42 and M31. Potential is here but i need a good sky.

 

Captures and live treatments were possible but i discovered some small bugs i am working on.

 

The setup is always the same :
SkyWatcher AZ-GTi motorized mount
Sony IMX485MC sensor
Canon FF 50 F/D 1.4 & 135mm F/D 2.5 lens

 

JetsonSky was V52_07RC (with small bugs) :

 

The link to the video : https://youtu.be/pn5...?feature=shared

 

Alain



#154 easybob95

easybob95

    Viking 1

  • -----
  • Vendors
  • topic starter
  • Posts: 703
  • Joined: 23 Apr 2014
  • Loc: France

Posted 17 January 2025 - 11:30 AM

Hi,

 

i work on a new 3 frames noise removal filter which can be useful in some specific circonstances. This filter is also interesting for some atmospheric turbulence (quite slow). It can be combined with other filters like AANR or/and best frame reference.

 

An example on Jupiter :

 

https://youtu.be/fVV...?feature=shared

 

 

An example on the Moon :

 

https://youtu.be/QF4...?feature=shared

 

Alain



#155 easybob95

easybob95

    Viking 1

  • -----
  • Vendors
  • topic starter
  • Posts: 703
  • Joined: 23 Apr 2014
  • Loc: France

Posted 20 January 2025 - 12:50 PM

Some examples about atmospheric turbulence management with JetsonSky.

 

Jupiter :

 

https://youtu.be/tWF...?feature=shared

 

 

The Moon & Clavius crater :

 

https://youtu.be/cAj...?feature=shared

 

Alain



#156 easybob95

easybob95

    Viking 1

  • -----
  • Vendors
  • topic starter
  • Posts: 703
  • Joined: 23 Apr 2014
  • Loc: France

Posted 28 January 2025 - 12:30 PM

Hello,

 

a new test with JetsonSky about atmospheric turbulence management.

 

The target is Jupiter & Callisto.

 

The RAW capture was not good so impossible to retrieve interesting details, even with heavy treatments. When collimation is bad, no miracle to expect.

 

https://youtu.be/Mw_2_pOWc3M

 

Alain



#157 jcharron

jcharron

    Lift Off

  • -----
  • Posts: 16
  • Joined: 17 Apr 2023

Posted 28 January 2025 - 04:18 PM

Hi, that's interesting and I guess it is a kind of Temporal Denoising. It can be found in several ISP image post process.

This is not wide spread in astronomy, because the stacking is preferred. However if you want to make live astronomy video and reduce noise it is a good solution. 



#158 easybob95

easybob95

    Viking 1

  • -----
  • Vendors
  • topic starter
  • Posts: 703
  • Joined: 23 Apr 2014
  • Loc: France

Posted 29 January 2025 - 05:45 AM

Hi jcharron,

 

those filters are mainly used for noise management but they also work well with some atmospheric turbulence. They prevent some specific changes in each pixel value, as the perfect image is supposed to be constant. They work with N-1 and N images for AANR filter and N-2, N-1 and N image for 3FNR1 and 3FNR2 filters.

 

That is to say i consider only 2 or 3 frames (depending of the filter) instead of hundreds frames with classical technics (like AS3 and R6). And my software is effectively dedecated to video,

 

Alain



#159 jcharron

jcharron

    Lift Off

  • -----
  • Posts: 16
  • Joined: 17 Apr 2023

Posted 29 January 2025 - 07:38 AM

But the point is how do you define your reference image ?



#160 easybob95

easybob95

    Viking 1

  • -----
  • Vendors
  • topic starter
  • Posts: 703
  • Joined: 23 Apr 2014
  • Loc: France

Posted 29 January 2025 - 09:36 AM

For atmospheric turbulence visualization, you can choose any image to set the reference because you only have to know the difference between images and the reference. Or you can choose the best image (considering your quality estimator) to set the reference image.

 

But with my software, atmospheric turbulence visualization is not used to manage the turbulence. It is only to show what turbulence looks like.

 

My filters (AANR, 3FNR1 & 3FNR2) don't use a reference image.

 

I made an other filter which limit pixel changes considering a reference (during the video acquisition or the playback, the best image can change depending of the image quality estimator which is based of sobel variance). This image reference can be made on the RAW image or on the processed image.

 

Alain



#161 easybob95

easybob95

    Viking 1

  • -----
  • Vendors
  • topic starter
  • Posts: 703
  • Joined: 23 Apr 2014
  • Loc: France

Posted 29 January 2025 - 11:02 AM

Concerning performance treatments :

 

For the last example with Jupiter and Callisto, my laptop (i9-13980H & RTX4080) performs about 35 fps for post processing.

 

When i make live treatment with camera capture with 1260x960 resolution, i can manage about 55 fps including capture and treatment with an IMX178MC sensor with USB 3. That’s quite fast.

 

Alain



#162 easybob95

easybob95

    Viking 1

  • -----
  • Vendors
  • topic starter
  • Posts: 703
  • Joined: 23 Apr 2014
  • Loc: France

Posted 02 March 2025 - 08:58 AM

Hello,

 

a test yesterday, targeting Andromeda galaxy M31.

 

Treatment was applied during capture.

 

I used 16 bits capture, noise management and HDR (using a single 16 bits capture).

 

The result is not bad.

 

https://youtu.be/T5I...?feature=shared

 

Alain



#163 easybob95

easybob95

    Viking 1

  • -----
  • Vendors
  • topic starter
  • Posts: 703
  • Joined: 23 Apr 2014
  • Loc: France

Posted 09 March 2025 - 12:49 PM

Hello,

 

A new test with 16 bits capture.

 

https://youtu.be/UYl...?feature=shared

 

Alain



#164 decep

decep

    Apollo

  • *****
  • Posts: 1,418
  • Joined: 19 Aug 2018
  • Loc: Alabama, US

Posted 09 March 2025 - 01:20 PM

Honest question... Is there any real benefit to using the 16-bit capture?  Looking at the specifications for most sensors, once you reach higher levels of gain (usually 300 and above on ZWO cameras), the dynamic range is reduced to 8-9 stops.  There really should not be much difference between 8 and 16-bit data at that level.  Is the tradeoff in framerate worth it?



#165 easybob95

easybob95

    Viking 1

  • -----
  • Vendors
  • topic starter
  • Posts: 703
  • Joined: 23 Apr 2014
  • Loc: France

Posted 09 March 2025 - 02:40 PM

For my purpose, that is to say making videos instead of photos, using 16 bits allow me to retrieve some signal.

 

Everything is explained here :

 

https://forums.devel...-nano/76861/871

 

You can read in this topic the latest posts where I explain my work and tests.

 

This also allow me to make HDR with single capture. With one capture, I can make 4 or more images with different brightness and get HDR result using Mertens method for example.

 

So, yes, 16 bits capture is useful for me.

 

Alain




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