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

any real difference between stacking in DSS, and some other real stacking software?

  • Please log in to reply
27 replies to this topic

#1 adamphillips

adamphillips

    Ranger 4

  • ****-
  • topic starter
  • Posts: 349
  • Joined: 21 Sep 2014

Posted 15 October 2018 - 06:00 PM

ive always stuck with DSS just because it seems to work really well, ive always wondered if stacking with ccdstack or maxim would make a difference. I do have maxim DL, but I haven't taken the time to learn how to stack with it.

 

anybody have some good advice for me? I do use maxim to register my pictures, I don't really see the point of buying registar.

 

I just looked and ccdstack has a 60 day trial period so maybe I should try it out. im a little curious about the deconvolution/sharpening.


Edited by adamphillips, 15 October 2018 - 06:12 PM.


#2 Goofi

Goofi

    Cosmos

  • *****
  • Posts: 8137
  • Joined: 03 May 2013
  • Loc: Coastal Southern California

Posted 15 October 2018 - 06:39 PM

I started out using CCDStack, it does a better job than DSS in my opinion. But, I switched to Pixinsight a few years ago, and now I use it or AstroPixelProcessor. Both do an outstanding job at all aspects of pre-processing.

 

Note: I don't think it is bad or wrong to use DSS, just there are tools that give better results. Are they worth it to you? That's a subjective call.



#3 bobzeq25

bobzeq25

    Hubble

  • *****
  • Posts: 16435
  • Joined: 27 Oct 2014

Posted 15 October 2018 - 07:15 PM

ive always stuck with DSS just because it seems to work really well, ive always wondered if stacking with ccdstack or maxim would make a difference. I do have maxim DL, but I haven't taken the time to learn how to stack with it.

 

anybody have some good advice for me? I do use maxim to register my pictures, I don't really see the point of buying registar.

 

I just looked and ccdstack has a 60 day trial period so maybe I should try it out. im a little curious about the deconvolution/sharpening.

Expansion on Goofi, directly related to your question.  Here's what needs to be "worth it" to you.  In's not money.

 

PI has an enormous (not an exaggeration, honest, when you consider the almost endless possibilities of combining them) number of ways to tweak the stacking process.  Those tweaks can, sometimes, make a significant difference in quality, compared to DSS relatively few adjustments.

 

The rub is that, in order to make full use of the PI advantage, you need to spend a lot of time figuring out what those adjustments do.  I have hundreds of hours in PI, and I surely don't understand many of them.  PI does _not_ make things easy for you.  I am familiar with a few that I find most useful.

 

I'm not that familiar with APP (used it once), but I'd make an educated guess that it provides some of the benefits of PI much more easily. 

 

PI is the "best", but it demands a lot out of those wanting to get all the advantages it offers.  The upside is that, even used simply, with mostly the defaults, it still is no worse then DSS, and perhaps better. 


Edited by bobzeq25, 15 October 2018 - 07:17 PM.

  • jdupton likes this

#4 APshooter

APshooter

    Soyuz

  • *****
  • Posts: 3653
  • Joined: 14 Oct 2013
  • Loc: Camby, In.

Posted 15 October 2018 - 07:32 PM

I've only just started using APP and find it almost as easy to use as Pi's batch preprocessing script.  There are 6 steps you step thru, fairly logical in their setup.  I processed the same image through both this morning.  Pi's was faster, and being more familiar with it produced a cleaner image.  BUT, I don't know how to apply yet the cosmetic correction in APP yet, so most likely it's a wash.  I agree with Bob, though, PI is a bear to learn at first.  I bought the CD set, looked at online videos for weeks and bought the Warren Keller Book.  PI is deep, and takes a long time to get your head around, at least for me  :)



#5 adamphillips

adamphillips

    Ranger 4

  • ****-
  • topic starter
  • Posts: 349
  • Joined: 21 Sep 2014

Posted 15 October 2018 - 07:34 PM

I know Pixinsight is a great program, but I don't think ill make the jump to PI. ive come so far in photoshop. one thing I struggle with is accurate color on my RGB, I know pixinsight has color calibration tools that would benefit me, and I don't know of any in photoshop. that's one area that would definitely help me out.

 

im more just curious about the stacking aspect. I wonder if people even still use ccd stack


Edited by adamphillips, 15 October 2018 - 07:36 PM.


#6 Scott Mitchell

Scott Mitchell

    Viking 1

  • -----
  • Posts: 859
  • Joined: 28 Jul 2007
  • Loc: Atlanta-ish, GA

Posted 15 October 2018 - 07:57 PM

Definitely give AstroPixelProcessor a shot. It is pretty interesting with a lot of adjustments you can make to each stage in the process. My first impression was that it produced a cleaner stacked image that DSS.


  • AhBok likes this

#7 AhBok

AhBok

    Vanguard

  • *****
  • Posts: 2217
  • Joined: 02 Dec 2010
  • Loc: Lakeland, TN

Posted 15 October 2018 - 08:45 PM

I agree that APP gives a cleaner stacked FITS. I still only use APP for pre-processing, but it does that very well. Takes a bit more processing power than DSS.

#8 mikeyL

mikeyL

    Viking 1

  • *****
  • Posts: 732
  • Joined: 17 Dec 2007
  • Loc: Longmont, CO, USA

Posted 15 October 2018 - 09:28 PM

I only had to try PixInsight once, on the very same data set as I had processed previously with DSS, to decide there was a big difference. Now granted, this data is not very great in the first place, but the 2 images shown below are the exact same data, just processed in DSS (1st version) and then a few years later, re-processed in PI (2nd version.) Furthermore, at the time I was running PI for this, I knew virtually nothing about the program at all, so was using the batch processing script with pretty much all default values. So the PI result could possibly be even a little better (not that it is worth the trouble to go back and process again at this point.)

 

DSS Version:

p958433196-4.jpg

 

PI Version (same data re-processed a year or 2 later...)

p1400347260-4.jpg

 

 

I bought PI on the spot and have never regretted it.

 

ML


Edited by mikeyL, 15 October 2018 - 09:30 PM.

  • mumbles likes this

#9 adamphillips

adamphillips

    Ranger 4

  • ****-
  • topic starter
  • Posts: 349
  • Joined: 21 Sep 2014

Posted 15 October 2018 - 11:11 PM

I only had to try PixInsight once, on the very same data set as I had processed previously with DSS, to decide there was a big difference. Now granted, this data is not very great in the first place, but the 2 images shown below are the exact same data, just processed in DSS (1st version) and then a few years later, re-processed in PI (2nd version.) Furthermore, at the time I was running PI for this, I knew virtually nothing about the program at all, so was using the batch processing script with pretty much all default values. So the PI result could possibly be even a little better (not that it is worth the trouble to go back and process again at this point.)

 

DSS Version:

p958433196-4.jpg

 

PI Version (same data re-processed a year or 2 later...)

p1400347260-4.jpg

 

 

I bought PI on the spot and have never regretted it.

 

ML

well it looks like there is a lot of post processing done in your example. is that correct? I don't use DSS for post processing, its not made for that.


Edited by adamphillips, 15 October 2018 - 11:16 PM.


#10 einarin

einarin

    Surveyor 1

  • *****
  • Posts: 1875
  • Joined: 23 Dec 2016

Posted 15 October 2018 - 11:19 PM

Yeah, the DSS is really just for stacking and you shouldn't do anything after stacking.



#11 jsmoraes

jsmoraes

    Apollo

  • -----
  • Posts: 1196
  • Joined: 20 Nov 2014
  • Loc: Vilatur, Saquarema, Rio de Janeiro, Brazil

Posted 16 October 2018 - 07:51 AM

The original question is very interesting: difference in stacking between DSS and others. Others processing aren't involved.

 

I know that with color files, DSS can vanish stars' color. Many people complain about it.

What we can say about mono files ? What is the difference in shape and intensity of signal after stacking on stars ? And faint areas of nebulaes ?

 

I have only DSS. I am reading tutorials about the others. But as I haven't them and never worked with them ... get the trial version, and do test by myself,  can result in wrong conclusion.

 

DSS is free. It would be very usefull if any fellow could lose time doing a simple test with both. The same group of files. Without any other processing ressource. Only stacking.


Edited by jsmoraes, 16 October 2018 - 07:52 AM.

  • happylimpet likes this

#12 Mert

Mert

    Fly Me to the Moon

  • *****
  • Posts: 6176
  • Joined: 31 Aug 2005
  • Loc: Spain, Pamplona

Posted 16 October 2018 - 07:56 AM

Can't really understand the point, but........... I started using DSS and sort of "liked"

it when beginning.

Once images are fainter ( less gain ) DSS starts to have problems finding stars.

Batchpreprocessing in Pix sometimes gave me surprises when using binning other

than 1x1 having to choose other debayer scheme then my camera uses in order

to get the colors right!

Once I got really sick of DSS I opted for buying APP because of the scientific

aproach behind it.

Never regretted it.

And for me, after comparing stacks done with DSS, done with Pix and done

with APP, I really stick with APP, easy to use and with great results.

Of course the pre-processing has to be ok, which is:  good Darks, good Flats and Bias

and after that all is going by it's own.

Can't recommend it enough ( and I don't have any personal or any other

interest wink.gif )

 

Regards,

Mert


  • Scott Mitchell and TimN like this

#13 happylimpet

happylimpet

    Soyuz

  • *****
  • Posts: 3822
  • Joined: 29 Sep 2013
  • Loc: Southampton, UK

Posted 16 October 2018 - 08:05 AM

The only time I saw a straight comparison of stacking with PI and DSS (it was on CN) the results were virtually indistinguishable. I seem to recall that DSS had the edge as regards sharpness, but this was very very small. Of course, perhaps the parameters werent tweaked to their fullest.



#14 Dan Crowson

Dan Crowson

    Vanguard

  • *****
  • Moderators
  • Posts: 2146
  • Joined: 08 Oct 2010
  • Loc: Dardenne Prairie, MO

Posted 16 October 2018 - 09:39 AM

I know Pixinsight is a great program, but I don't think ill make the jump to PI. ive come so far in photoshop. one thing I struggle with is accurate color on my RGB, I know pixinsight has color calibration tools that would benefit me, and I don't know of any in photoshop. that's one area that would definitely help me out.

 

im more just curious about the stacking aspect. I wonder if people even still use ccd stack

Adam,

 

I'm probably in the same boat as you here since I don't use PI. I determine filter ratios on every image by using Bob Franke's eXcalibrator tool - http://bf-astro.com/...xcalibrator.htm.

 

Dan



#15 adamphillips

adamphillips

    Ranger 4

  • ****-
  • topic starter
  • Posts: 349
  • Joined: 21 Sep 2014

Posted 16 October 2018 - 12:27 PM

Adam,

 

I'm probably in the same boat as you here since I don't use PI. I determine filter ratios on every image by using Bob Franke's eXcalibrator tool - http://bf-astro.com/...xcalibrator.htm.

 

Dan

hey dan, ive been a fan of yours for a while. I looked at excalibrators page. it looks to me as I expected, it gives you those ratios to use on RGB. is that only for stacking in maxim? I don't know if DSS does that.

when I do RGB, this might sound bad to some, I stack the red, then green, then blue separately then register them. but my experience shows me that the color is not perfect, I would like to find a way to calibrate the color better. it looks like that program is free, ill have to try it out.

 

my next target is going to be ngc772 inspired by that guys awesome picture of it recently. so hopefully I can try something new



#16 A. Viegas

A. Viegas

    Mercury-Atlas

  • *****
  • Posts: 2992
  • Joined: 05 Mar 2012
  • Loc: New York City/ CT

Posted 16 October 2018 - 01:51 PM

The new 64 bit version of DSS vastly improves speed. One of the problems I used to have was combining large sets of short exposure files like 100 30s subs. New 64bit DSS is faster than PI at just simple stacking. Overall I agree with what most have said here, PI is very powerful but requires a lot of time to master or actually even learn how to use. I tried APP as well a couple of times and while the interface is easy and intuitve it also has some features that require learning while DSS is basically load and click. Also free is free. Since I got PI I have spent a huge amount of time trying to figure it out. Are my images better for it? Yes... how much better than photoshop + DSS well I would say a bit sometimes to a lot sometimes... but basically I would ballpark t at 50% better to give a number. The $ spent on PI really has not been the issue for me, rather it's been the time. And today I do not think I am much better at using it than I was a year ago when I started, it really is a super detailed and comprehensive program. I used to have more free time so I did not mind sinking time into it to get 50% better results. But in the last 4 months I have had very very little time to play with astro images. I have 20-40gb of data sitting around... truth be told I think if I had never gone to PI I probably would have processed it by now.... but with PI it just takes so much more time. And that is the real issue. My advice... if you have a lot of time get PI. If not stick with DSS

Al
  • happylimpet likes this

#17 terry59

terry59

    Cosmos

  • *****
  • Posts: 8965
  • Joined: 18 Jul 2011
  • Loc: Colorado, USA

Posted 16 October 2018 - 02:16 PM

The new 64 bit version of DSS vastly improves speed. One of the problems I used to have was combining large sets of short exposure files like 100 30s subs. New 64bit DSS is faster than PI at just simple stacking. Overall I agree with what most have said here, PI is very powerful but requires a lot of time to master or actually even learn how to use. I tried APP as well a couple of times and while the interface is easy and intuitve it also has some features that require learning while DSS is basically load and click. Also free is free. Since I got PI I have spent a huge amount of time trying to figure it out. Are my images better for it? Yes... how much better than photoshop + DSS well I would say a bit sometimes to a lot sometimes... but basically I would ballpark t at 50% better to give a number. The $ spent on PI really has not been the issue for me, rather it's been the time. And today I do not think I am much better at using it than I was a year ago when I started, it really is a super detailed and comprehensive program. I used to have more free time so I did not mind sinking time into it to get 50% better results. But in the last 4 months I have had very very little time to play with astro images. I have 20-40gb of data sitting around... truth be told I think if I had never gone to PI I probably would have processed it by now.... but with PI it just takes so much more time. And that is the real issue. My advice... if you have a lot of time get PI. If not stick with DSS

Al

"The new 64 bit version of DSS vastly improves speed. One of the problems I used to have was combining large sets of short exposure files like 100 30s subs. New 64bit DSS is faster than PI at just simple stacking. Overall I agree with what most have said here, PI is very powerful but requires a lot of time to master or actually even learn how to use.......while DSS is basically load and click."

 

Using BPP is pretty much the same. The default settings aren't the best but a one time adjustment improves it tremendously. I use settings recommended by David Ault.

 

I know BPP isn't for everyone but it serves me fine


Edited by terry59, 16 October 2018 - 02:17 PM.


#18 Dan Crowson

Dan Crowson

    Vanguard

  • *****
  • Moderators
  • Posts: 2146
  • Joined: 08 Oct 2010
  • Loc: Dardenne Prairie, MO

Posted 16 October 2018 - 02:18 PM

hey dan, ive been a fan of yours for a while. I looked at excalibrators page. it looks to me as I expected, it gives you those ratios to use on RGB. is that only for stacking in maxim? I don't know if DSS does that.
when I do RGB, this might sound bad to some, I stack the red, then green, then blue separately then register them. but my experience shows me that the color is not perfect, I would like to find a way to calibrate the color better. it looks like that program is free, ill have to try it out.

After registering the R, G and B, you would need to plate solve one of them and load into eXcalibrator. It would then give you ratios. You'll find that they usually aren't the same with each image because there are other factors involved like how high or low the object is. Anyway, once you have them, I would think you could go into PhotoShop and combine them using the ratios. Off the top of my head, maybe you paste them into each channel (red into red, etc) and then you could change the opacity to 'match' the ratios.
 
Programs like CCDStack, Maxim, PixInsight and others allow you to do the combine within them. If someone was starting today, I'd definitely say just buy PixInsight and go from there but, having learned on PhotoShop and really only spending 5-10 minutes at most processing an image, I haven't moved to PI even though I bought it 4-5 years ago.

Dan



#19 choward94002

choward94002

    Viking 1

  • *****
  • Posts: 580
  • Joined: 25 Jul 2016
  • Loc: Bay Area, Central AZ

Posted 16 October 2018 - 11:28 PM

There are two tautologies that come to mind reading this thread, "if the only tool in your belt is a hammer, then every problem is a nail" and "remember the spork" ...

 

Just like there are many ways to gather images (SCT, Reducer, HyperStar, Refractor, Dob using DSLR, CMOS, CCD guided with nothing, guide scope, OAG, ONAG) there are just as many ways to process them (PI, APP, Maxim, PhotoShop, GIMP) and stack them (Registax, PI, DSS, CCDStack) ... if you restrict yourself to just one method (say SCT with Guidescope, or SCT with reducer and OAG) then you're going to be spending a lot more time pounding on the bolt when just using a wrench would do.  By the same token, a "spork" makes a pretty crummy fork and a very frustrating spoon ...

 

Each stacking software (and processing software) has it's strengths and weaknesses; my tagline lists five different stacking programs and that's not because I like typing -- each has strengths for different kinds of objects, each has weaknesses ... the trick is knowing when to use a screwdriver for screws, a wrench for bolts and the hammer when I find a nail (and, of course, not trying for a "one thing does it all 'spork'").  I constantly change stacking programs and techniques as they change and get better/ worse ... because they ALL have real, obvious differences just like hammers are different than wrenches and screwdrivers ...

 

And, realizing that I've spent thousands on equipment, untold hours setting things up, getting things collimated, more hours processing things, needing to pony up $200 or $400 for a program is like buying an expensive sports car and then pulling into the cheapest gas station I can find for gas ...

 

Great before/ later example, BTW! 



#20 Jon Rista

Jon Rista

    ISS

  • *****
  • Posts: 23352
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 16 October 2018 - 11:34 PM

Pre-processing results were the first thing that really sucked me into PI. With the most naive approach in PI, I always got sharper results with it than with DSS. With more advanced usage of PI, DSS was never able to compete. DSS is great for it's very simple workflow, which is one of the easiest out there. If you want to maximize the potential of your data, PI has the tools and the low level power to allow you to do it. It CAN and often WILL take more effort, but that effort can be worth it.

 

APP does a good job as well, and comes in at a balance point between PI and DSS. It is pretty easy, but more configurable than DSS. The think that has generally kept me with PI is APP is so much slower. PI used to be quite slow, but once its slowest pre-processing tools received the multi-threading treatment a year and a half ago or so, PI got a major boost in overall pre-processing performance. It just screams now, and is actually quite fast even with very large stacks (and my computer by todays standards is rather modest...doesn't come close to the kind of threadripper builds people are making for PI these days.)



#21 Goofi

Goofi

    Cosmos

  • *****
  • Posts: 8137
  • Joined: 03 May 2013
  • Loc: Coastal Southern California

Posted 16 October 2018 - 11:45 PM

@APShooter ... create a BadPixelMap in APP, and when you use it that takes care of your cosmetic correction.

 

For all ... One thing I'll comment on in terms of preprocessing is support for newer CMOS cameras. PI doesn't have any strong built-in tools for dealing with dark flats. I've got the customized version of BPP that swaps dark flats into the bias slot, but it's still not an official part of the program ... so it can break on an upgrade. APP has built in support for dark flats, and it works very well.

 

I'm not sure where DSS is in this ...  I haven't used DSS for a few years now.



#22 terry59

terry59

    Cosmos

  • *****
  • Posts: 8965
  • Joined: 18 Jul 2011
  • Loc: Colorado, USA

Posted 17 October 2018 - 08:13 AM

@APShooter ... create a BadPixelMap in APP, and when you use it that takes care of your cosmetic correction.

 

For all ... One thing I'll comment on in terms of preprocessing is support for newer CMOS cameras. PI doesn't have any strong built-in tools for dealing with dark flats. I've got the customized version of BPP that swaps dark flats into the bias slot, but it's still not an official part of the program ... so it can break on an upgrade. APP has built in support for dark flats, and it works very well.

 

I'm not sure where DSS is in this ...  I haven't used DSS for a few years now.

 

I've not used APP in a while and wasn't aware of the dark flat support....thanks Goofi


  • Goofi likes this

#23 mexbound

mexbound

    Sputnik

  • -----
  • Posts: 33
  • Joined: 29 Jun 2016

Posted 17 October 2018 - 11:04 AM

Curious I haven't seen this in PI yet.  How were you able to obtain it. I'm working with Dark Flats instead of Bia frames now do I am interested in this modified BPP.



#24 mexbound

mexbound

    Sputnik

  • -----
  • Posts: 33
  • Joined: 29 Jun 2016

Posted 17 October 2018 - 02:28 PM

I am able to answer my own question.  On the PI forum I found the new PI script named BatchPreProcessingFD  (FD for Flat Dark). It can be found at https://pixinsight.c...p?topic=11992.0 

 

You need a login to the PI forum to get this.


  • bobzeq25 likes this

#25 Goofi

Goofi

    Cosmos

  • *****
  • Posts: 8137
  • Joined: 03 May 2013
  • Loc: Coastal Southern California

Posted 17 October 2018 - 06:05 PM

@Terry ... they're releasing regular improvements to APP. It's worth checking out.

 

@mexbond .. that's what I use, that script you linked to.




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