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

Siril a better DSS alternative?

  • Please log in to reply
55 replies to this topic

#1 Traveler

Traveler

    Vanguard

  • *****
  • topic starter
  • Posts: 2252
  • Joined: 19 Aug 2007
  • Loc: The Netherlands

Posted 13 October 2017 - 12:10 AM

Siril a better DSS alternative?

 

The last couple of days I played with Siril for Windows. Siril brought out last September for Windows. It is free, runs on several OS platforms and I am trying to figure out if Siril is an alternative for DSS under Windows. With DSS I have a love-hate relation, it is just an older and non 64b application.

 

If you want to look for yourself I provide some Siril links.

 

 

As said, I am playing with Siril and I must say: lots of options can be used. At this time I am not tried everything Siril can do but my feeling is that the stacks that comes out of Siril, are someway “better” then the stack results DSS produces. Better because processing the Siril stacks give me a faster adequate result when processed with Photoshop CS2.
Anybody have the same feeling?

 

In general I wonder what your experience is with Siril? Also, it will be good when there would be a short list of best Siril options. When you have such a short list I would like to read it.


Edited by Traveler, 13 October 2017 - 12:12 AM.

  • OleCuss likes this

#2 OleCuss

OleCuss

    Apollo

  • *****
  • Posts: 1207
  • Joined: 22 Nov 2010

Posted 13 October 2017 - 05:18 AM

Thank you for the heads-up!  I'd not heard of Siril.  I really like that it is 64-bit.

 

Not quite as happy when I note that the Windows version will always be in beta.



#3 poobie

poobie

    Ranger 4

  • -----
  • Posts: 331
  • Joined: 30 Dec 2013
  • Loc: Tuscaloosa, AL

Posted 13 October 2017 - 02:22 PM

Siril is great!  Been using it under MacOS and Linux for a year or so, and the Windows version in the last month.  

 

my workflow looks like this:

import calibration frames

stack calibration frames (winsorized sigma clipping)

import lights

calibrate lights

debayer lights

register lights

stack lights (again, winsorized sigma clipping)

 

Then I do a background extraction, color calibration, and histogram mod.  The background extraction works really, really well, as far as I'm concerned.  

 

It's a little more hands-on than DSS, but I've been a lot happier with the results.  It's also under active development, which is a huge bonus.


  • Traveler, nubycakes, whwang and 1 other like this

#4 whwang

whwang

    Surveyor 1

  • *****
  • Posts: 1889
  • Joined: 20 Mar 2013

Posted 13 October 2017 - 09:40 PM

import calibration frames

stack calibration frames (winsorized sigma clipping)

import lights

calibrate lights

debayer lights

register lights

stack lights (again, winsorized sigma clipping)

Can all these been done in one step?  I really hate to manually go through all the steps.  Packing all these into one step will allow me to work on other stuff while the software works through all these calibration and stacking.  I can't afford sitting in front of a computer and click at something every few minutes.  It's inefficient.

 

I browsed through the PDF manual of Siril, and did not find an indication about drizzle (both normal drizzle and Bayer drizzle).  I might have missed it.  However, if Sirial really doesn't have drizzle capability, it is a game stopper for me.

 

If it does drizzle, and if it can make the calibration/stacking a one-stop shopping, hack, I don't even mind paying for it.  DSS is free, which is great for many beginners.  But I also believe the fact that it's free prevents it from getting better in long run.



#5 poobie

poobie

    Ranger 4

  • -----
  • Posts: 331
  • Joined: 30 Dec 2013
  • Loc: Tuscaloosa, AL

Posted 13 October 2017 - 10:15 PM

Can all these been done in one step?  I really hate to manually go through all the steps.  Packing all these into one step will allow me to work on other stuff while the software works through all these calibration and stacking.  I can't afford sitting in front of a computer and click at something every few minutes.  It's inefficient.

 

 

 

I browsed through the PDF manual of Siril, and did not find an indication about drizzle (both normal drizzle and Bayer drizzle).  I might have missed it.  However, if Sirial really doesn't have drizzle capability, it is a game stopper for me.

 

If it does drizzle, and if it can make the calibration/stacking a one-stop shopping, hack, I don't even mind paying for it.  DSS is free, which is great for many beginners.  But I also believe the fact that it's free prevents it from getting better in long run.

 

The GUI doesn't allow for batching like that.  should be scriptable from the command line though.  As far as I can tell, drizzle (and bayer drizzle) are on the roadmap for next year, but not yet implemented.  

 

If you do find yourself with a little spare time, I think it's worth your time to check it out; there are definitely things it does better than DSS, at the cost of being a little less automated.  it handles the massive stacks I've made with my ASI178 significantly better than DSS.


  • xiando likes this

#6 lock042

lock042

    Siril astronomical image processing software

  • -----
  • Vendors
  • Posts: 78
  • Joined: 03 Nov 2015

Posted 14 October 2017 - 01:13 AM

Doing every step automatically is not a good idea IMHO.

Indeed, one of the reason why Siril may give better results than DSS is one can control what we are doing.

-> We can control rejection while stacking masters

-> We can control cosmectic correction

-> .....

 

About Drizzle (and Bayer Drizzle): it is on the roadmap it is true. But that is not the priority as many features are more important. Bayer Drizzle is maybe more important and will be implemented first, but no rush.

Of course, Siril is free and opensource, so if someone wants to code this function: no problem.

 

Have a good day,

Cheers



#7 Tonk

Tonk

    Cosmos

  • *****
  • Posts: 8521
  • Joined: 19 Aug 2004
  • Loc: Leeds, UK, 54N

Posted 14 October 2017 - 04:31 AM

Im interested in any plans to process 32 bit images. Currently i see its only 16 bit capable



#8 whwang

whwang

    Surveyor 1

  • *****
  • Posts: 1889
  • Joined: 20 Mar 2013

Posted 14 October 2017 - 07:17 AM

Doing every step automatically is not a good idea IMHO.

Indeed, one of the reason why Siril may give better results than DSS is one can control what we are doing.

-> We can control rejection while stacking masters

-> We can control cosmectic correction

-> .....

I am not quite sure what you meant here.  Rejection is indeed important.  Cosmetic correction is useful too.  I completely agree with this. However, whether to do these or not and the associated parameters can all be predetermined before the processing begins.  So there is nothing stopping you from doing all these at once.  PixInsight's batch preprocessing script is a good example.


  • Tonk and pfile like this

#9 lock042

lock042

    Siril astronomical image processing software

  • -----
  • Vendors
  • Posts: 78
  • Joined: 03 Nov 2015

Posted 14 October 2017 - 11:05 AM

I do not exactly agree. Batch processing is always possible but what if you need to adapt parameters because too much rejection in masters ?

#10 xiando

xiando

    Aurora

  • *****
  • Posts: 4919
  • Joined: 27 May 2015
  • Loc: Cloudy NEOhio

Posted 14 October 2017 - 11:47 AM

Some quick thoughts from my first use. Probably irrelevant in the grand scheme of things but nonetheless I'll  post them while they're still fresh in my mind...

 

1) I like the console window feedback. in comparison to DSS... DSS is a bit nerve-wracking sometimes. It also tends to lock the processors. The console feedback in Siril gives me heart that something actually happening, and I can actually review the progress for more than  a tenth of a second in a small modal window. It also doesn't seize the computer. Kudos.

 

2) it's quick (again, in comparison to DSS)

 

3) directions are a little muddled, but creating new master darks, flats, and bias to conform to Siril was relatively easy. It did not accept my DSS-created masters, presumably due to some file format issue. Still a bit puzzling but no big deal in the end.

 

4) Picky stupid observation regarding the icon on my laptops's quick bar (I use the tiny icon view to maintain work-space size)  the icon is a bit subdued and difficult to see in the quick-bar. suggestion:recommendation. Provide alternate icon tailored to "tiny" mode to improve visibility in quick-bar. Fluff...I know...

 

5) Perhaps I'm missing something. When you've already converted a file set, but forgotten to stack, it appears you have to go through the whole thing again from scratch (ie import files, convert, etc before you can remedy the missing stack operation (encountered when creating cal files) flats

 

6) while I realize it represents additional documentation work (issues a face-palmed groan from every overtaxed programmer who reads this, myself included wink.gif )  and obviously, there are only so many sufficiently intimate/knowledgeable and skilled enough to do that work... It would be nice if there was a "quickstart" first run paint-by-numbers procedure (I believe the trendy phrase is workflow) to get the novice over that first major hurdle of achieving an initial "final". I had several fits and starts in processing my first image due to the uncertainty in what needed to be done. Edit: It appears I got ahead of myself...There are tutorials, just at the end, not where I expected them to be. I'll shut up on that for now until I got through them. 

 

I really like it so far, but as noted, I've processed all of one photo, and that's just a crude first shot. I really have no idea what I'm doing, just cobbling my way through. I have a long way to go to master it, but it's very promising. My "final" image is much tighter (smaller stars) and color channel registration seems much better than with DSS, (more clarity in the galaxy I'm playing with NGC 6946), since using the same data for both DSS and Siril, Siril produces a less grainy output with higher detail and  less star bloat...again, with only the most rudimentary understanding of how to "make it go"

 

overall, I give it a big +1

 

more to come later as I figure out more.


Edited by xiando, 14 October 2017 - 10:21 PM.

  • Traveler and OleCuss like this

#11 whwang

whwang

    Surveyor 1

  • *****
  • Posts: 1889
  • Joined: 20 Mar 2013

Posted 14 October 2017 - 12:41 PM

I do not exactly agree. Batch processing is always possible but what if you need to adapt parameters because too much rejection in masters ?

 

Are you talking about master flat, bias, dark etc?  The stacking of them is pretty routine.  There is no need to change the rejection parameters all the time.  Once you find a set of parameters that work, you can use them pretty much without any changes in future works, unless you change your imaging workflow/equipment substantially.

 

Even for stacking object exposures, I find my standard set of rejection method/parameters work nearly all the time.  I also know how to adjust the parameters based on how many images I have (e.g., more aggressive rejection parameters when having less exposures).  I can adjust these before I initiate the processing.  I rarely find myself needing to going back and forth and change this and change that in PixInsight.

 

To be fair, PixInsight's stacking module is not that sophisticated in the batch preprocessing, comparing to its standalone integration function.  So I still do it in two steps, calibration (including stacking the calibration frames) and integration.  However, this is still way more efficient than what Matt described earlier.

 

Really, when you have tons of images to process, automation is absolutely essential.

 

Don't get me wrong, as a Mac user, I love to see a free stacking/processing program for the Mac platform.  I just hope what I say here can help the development team to know what may substantially improve their program.

 

Cheers,

Wei-Hao



#12 TimN

TimN

    Mercury-Atlas

  • -----
  • Posts: 2884
  • Joined: 20 Apr 2008
  • Loc: Ontario, Canada

Posted 14 October 2017 - 12:55 PM

The best replacement I've found for DSS is Astro Pixel Processor. APP works with Windows, Mac or Linux. Unfortunately, its not free but the improvement may be worth it to you. https://www.astropixelprocessor.com/


  • whwang likes this

#13 scopenitout

scopenitout

    Apollo

  • -----
  • Posts: 1366
  • Joined: 24 Aug 2013
  • Loc: Mt. Belzoni

Posted 14 October 2017 - 05:57 PM

DSS = Abandonware

#14 xiando

xiando

    Aurora

  • *****
  • Posts: 4919
  • Joined: 27 May 2015
  • Loc: Cloudy NEOhio

Posted 14 October 2017 - 06:31 PM

DSS = Abandonware

Well,. let's not misrepresent it. DSS == very useful, simple, and free.


  • Tonk and whwang like this

#15 whwang

whwang

    Surveyor 1

  • *****
  • Posts: 1889
  • Joined: 20 Mar 2013

Posted 14 October 2017 - 10:56 PM

There are still lots of mileage left in DSS.  It's no longer the best, but it has all the basic functions needed for stacking deep-sky images.  With regular updates to support new cameras, it will remain useful for beginners and intermediate level photographers for quite a few more years.

 

But I agree, it is a dying product, just dying very slowly.  I hate to say this.


  • Tonk and xiando like this

#16 xiando

xiando

    Aurora

  • *****
  • Posts: 4919
  • Joined: 27 May 2015
  • Loc: Cloudy NEOhio

Posted 14 October 2017 - 11:10 PM

I'll be keeping it on my toolbar for the forseeable future. Even if it doesn't produce the highest quality, it has its own elegance of sorts.



#17 Tonk

Tonk

    Cosmos

  • *****
  • Posts: 8521
  • Joined: 19 Aug 2004
  • Loc: Leeds, UK, 54N

Posted 15 October 2017 - 11:50 AM

DSS is the only stacking program that i know of that does time interpolated alignment of comets. You mark comet in first and last image of the set and DSS works the rest out from the image time stamps. It gives perfect positioning. Average stacking shows this well as star trails are perfect straight lines. Compare to say Images Plus where you need to mark the comet in every frame and because of its centroid alignment, comets passing very close to bright stars cause a kink in the alignment path. I'm still holding out using DDS because of this beautifully implemented solution.


  • nubycakes and xiando like this

#18 xiando

xiando

    Aurora

  • *****
  • Posts: 4919
  • Joined: 27 May 2015
  • Loc: Cloudy NEOhio

Posted 15 October 2017 - 12:10 PM

DSS is the only stacking program that i know of that does time interpolated alignment of comets. You mark comet in first and last image of the set and DSS works the rest out from the image time stamps. It gives perfect positioning. Average stacking shows this well as star trails are perfect straight lines. Compare to say Images Plus where you need to mark the comet in every frame and because of its centroid alignment, comets passing very close to bright stars cause a kink in the alignment path. I'm still holding out using DDS because of this beautifully implemented solution.

Another positive is the minimal number of intermediary files it produces. Using Siril is showing me one flaw in Sirils' design, in that each major step results in another duplication of source files. (converted raws plus cosmetically corrected raws plus registered raws... ) For someone like me, who uses up to many hundred raws per target, this can produce excessive harddrive use.

 

DSS also has groups. Siril does not. I have no idea how I'm going to handle my multinight captures...

 

Not bashing Siril on that, just buffering the abandonware claim.


Edited by xiando, 15 October 2017 - 12:10 PM.

  • Tonk likes this

#19 Orionis91

Orionis91

    Mariner 2

  • *****
  • Posts: 262
  • Joined: 01 Mar 2010
  • Loc: Maryland USA

Posted 15 October 2017 - 02:44 PM

Thanks for pointing this out, I had not hear of Siril before. I plan on giving it a try with some old images when I have some free time. 

 

Dave



#20 poobie

poobie

    Ranger 4

  • -----
  • Posts: 331
  • Joined: 30 Dec 2013
  • Loc: Tuscaloosa, AL

Posted 15 October 2017 - 08:37 PM

make sure y'all try the processing functionality too - I've been really happy with that part of it.  Heck, I routinely use it to process images I've stacked with other apps (usually Sequator, since nothing much else will handle Fuji raws.)



#21 Tonk

Tonk

    Cosmos

  • *****
  • Posts: 8521
  • Joined: 19 Aug 2004
  • Loc: Leeds, UK, 54N

Posted 16 October 2017 - 07:38 AM

I've looked over the Siril source code this weekend - its smack in my experience domain (C/C++, OpenCV).

Other than DSS's near perfect comet stacking not many stacking/calibration programs support comet aficionados like me well. I'm thinking that this might be a route for me to adapt the inner algorithms Siril uses to help me create a tool better suited to comet processing. Some specialized statistical stacking and sieving would help me out as well as the recreation of the time stamp alignment that DSS does if and when DSS fades too far away. The (manual - and tedious) sieving/stacking technique of mine is documented here - https://www.astrobin.com/151422/ - it could do with automating and its needs to use the image time stamps to work.

Yo! I have a cloudy night hobby to look forward too (once my current project - a remote robotic telescope rig - is off my hands)  ;)


  • Traveler, whwang and xiando like this

#22 Traveler

Traveler

    Vanguard

  • *****
  • topic starter
  • Posts: 2252
  • Joined: 19 Aug 2007
  • Loc: The Netherlands

Posted 16 October 2017 - 10:40 AM

Great intention Tonk!



#23 ImNewHere

ImNewHere

    Apollo

  • *****
  • Posts: 1130
  • Joined: 29 Mar 2017
  • Loc: Nashville, IN

Posted 16 October 2017 - 10:49 AM

I have not checked the link yet, but does it handle native .dng and .pef formats? My cameras shoot in either one as their native raw format, though I generally always have it in .dng for my raws.



#24 scopenitout

scopenitout

    Apollo

  • -----
  • Posts: 1366
  • Joined: 24 Aug 2013
  • Loc: Mt. Belzoni

Posted 16 October 2017 - 08:41 PM


DSS = Abandonware

Well,. let's not misrepresent it. DSS == very useful, simple, and free.

Not a misrepresentation at all.
When was DSS last updated?
Is the Developer still alive?

#25 xiando

xiando

    Aurora

  • *****
  • Posts: 4919
  • Joined: 27 May 2015
  • Loc: Cloudy NEOhio

Posted 16 October 2017 - 08:52 PM

 

 

DSS = Abandonware

Well,. let's not misrepresent it. DSS == very useful, simple, and free.

Not a misrepresentation at all.
When was DSS last updated?
Is the Developer still alive?

 

The general "street" implication of abandonware is "Do not use" or "this software is garbage". 

 

You did not qualify your post in any way to indicate that while it is now unsupported officially, it is still a useful piece of software, so I interpreted your post to imply the negative. However, it's still quite useful, as both Tonk and I noted in followup posts to your statement.




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.







Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics