Siril 0.9.2

[lock042]

Hello, we are glad and proud to announce that a new version of Siril has just been released and is now available to all users: Siril 0.9.2
1 - Presentation:
Siril is meant to be Iris for Linux (sirI-L). It is an astronomical image processing tool, able to convert, pre-process images, help aligning them automatically or manually, stack them and enhance final images.

2 - How to install the release:
Each new version is produced with several packages for most OS we handled. We try to build maximum of packages and cheer you to check if a package does exist for your OS at this page: http://free-astro.org/index.php/Siril:0.9.2
Remember that Siril is a software for UNIX-like system, in consequence it does not work on Windows.

 

3 - New features:

Siril 0.9.2 is a stability and minor improvements release: it contains several bug fixes and at least 15 improvements over to the previous version, Siril v0.9.1.

 

• Provide a better planetary registration algorithm and a better planetary quality algorithm.
• Parallelize Registration.
• AVI to SER conversion.
• Animated GIF output.
• Hot pixel removal feature.
• Allow a reference magnitude to be set to get absolute magnitude instead of relative.
• Add support for dark optimization.
• New sequence selection range command.
• Vertical banding noire removal tool.
• Clarify the use of demoisaicing.
• Improve star detection and therefore global star alignment.
• Allow the user to manually select background samples.
• Add autostretch viewer mode.
• Improved RGB compositing tool.
• Many bug fixes and stability improvements: This release comes with a large amount of bug fixes thanks to many reports of our beta-tester. Feel free to report any bugs you could find by your own. We’ll do our best to fix them.

 

We sincerely hope you will enjoy using Siril and you will produce nice shots.

The Siril development team

 

[lock042]

Hello, because of a serious bug, we updated the version to 0.9.3

Downloads here : http://free-astro.org/index.php/Siril:0.9.3

 

Thanks.

[GuLinux]

Hi,

First of all, thank you for your application! I think it fills a gap in the GNU/Linux ecosystem, as it's quite hard to find a similar application for images preprocessing!

 

I have a suggestion/request: I've been using siril also for stars spectra preprocessing, but the "quality detection" is often wrong, probably because the object itself is very different from the usual objects (I tried mostly the "planetary" quality detection).

Would you think it's possible to implement a separated algorithm for spectra?

I'm a software developer too, so I might help at least in discussing it in detail

[lock042]

Hello.

 

Sorry for the late answer, but I want to thank you for your message. You have to know that spectra processing is a feature we want in Siril for further versions. But for now .... lack of time.

For your problem, it is normal that planetary quality algorithm gives wrong result: it is not dedicated for these purposes. Having a good quality algorithm for spectra images requires to know which quality criteria we want to apply. Could you post a good and a bad image ?

 

> I'm a software developer too, so I might help at least in discussing it in detail

Siril is free and opensource, so fell free to contribute :).

 

Thanks a lot,

Best regards

[GuLinux]

I guess the "good image" would be one with very well defined lines, and, when using a non slit spectrograph like a StarAnalyser (the one I'm using), a very sharp and thin star line.

I'm attaching a sample, in this particular case there's no huge difference in the extracted profile, but I guess you can understand what I mean at least.

I wrote a python application for spectra analysis (from image to calibrated and corrected profiles), and I tried to implement an algorithm for frames quality detection; I didn't go too far, but the algorithm I tried to apply involved comparison of the spectrum profile (plot along the x axis) from a user defined reference image.

So the steps would be:

• rotate all the frames (in this case you can notice it is not perfectly aligned with the x axis)
• show some kind of dialog with two frames: one with the image, another with the profile
• ask the user to set the reference frame
• sort all the other frames using a correlation index (I tried Pearson)

I'm not too good with C, I use mostly C++, and a litle bit of python... Python itself is quite good at these things, so maybe I could start with a separate prototype helper for this particular task, and see how that goes...

Still the huge issue right now will be finding the right time to do that

 

[lock042]

Hello,

Siril is a mix of C and C++ code: the code is in C but some pieces of code are in c++ (mainly for the use of opencv and a new avi export part (same code than used in PIPP/SER Player)).

So, if you want to develop some code in C++ I don't think it will be an issue. However, start with a separate prototype is a good idea.

Regards,

Cyril