90 replies to this topic

[Borodog]

Somebody recently recommended a different software package than DSS for stacking and stretching all in the same tool, but I can't find that thread. Was it APP?

I am really starting to dislike DSS. The stacked images look absolutely nothing like what I expect them too, which is an average of the lights corrected for vignetting by the flats (I don't expect to visually see much from the darks or biases from the test shots I've been making while waiting for clear dark skies). Instead I get an extremely dark image with greatly increased noise and crazy color balance written out in a tiff format that displays differently in every piece of software I open it in. Maybe it's user error, but I really prefer software that is intuitive to use.

 

Anyway, what is your favorite stacking software that is not DSS?

[JEPott]

SiriL is similar to DSS - runs both on PC and Mac

[DuncanM]

Somebody recently recommended a different software package than DSS for stacking and stretching all in the same tool, but I can't find that thread. Was it APP?

I am really starting to dislike DSS. The stacked images look absolutely nothing like what I expect them too, which is an average of the lights corrected for vignetting by the flats (I don't expect to visually see much from the darks or biases from the test shots I've been making while waiting for clear dark skies). Instead I get an extremely dark image with greatly increased noise and crazy color balance written out in a tiff format that displays differently in every piece of software I open it in. Maybe it's user error, but I really prefer software that is intuitive to use.

 

Anyway, what is your favorite stacking software that is not DSS?

Astroart 7 is a preprocessing, processing, image acquisition and complete automation package, with a built in star chart, multistar autoguider interface, integrated stand alone platesolving and multistar autofocus.

 

It has a 64bit preprocessing and processing engine, that is probably the fastest available, while also being extremely frugal with system resources. It has two 32bit image acquisition/processing engines for backward compatibility:

 

http://www.msb-astroart.com/

 

There is a free unlimited time demo to experiment with and to test hardware compatibility:

 

http://www.msb-astro...com/down_en.htm

 

It is an order of magnitude faster than APP, according to my testing and somewhat faster than DSS. It has a straightforward and easy to use interface, which IMHO is easier to use than DSS.

Edited by DuncanM, 01 January 2021 - 03:11 PM.

[Jim R]

Sequator decent results, simple to use, free.

[AZ Maverick]

For paid programs, I like and use APP or the WBPP in PixInsight.

For free and not DSS, I have ASTAP which has a stacking program but I have never used it.

[bobzeq25]

PixInsight is "the best".  An array of _extremely_ adjustable tools.  The downside is the time spent learning to use it well, and using it.  There are scripts to shortcut the process, but they throw away many of the possible adjustments.  I tried using them, but after fixing a number of problems I had with the scripts by using the individual tools, I no longer use them.  If you want simplicity, PI is not a good choice.  But...

 

A lot of people like Astro Pixel Processor.

 

Those two combine stacking and processing into one program, which is a real advantage.  Some prefer using multiple programs, my personal opinion is that's (_especially_ for beginners) an option more likely to cause problems than fix them.

 

Especially true since those two excellent comprehensive programs with distinct strengths exist.

 

The "free" alternatives are not free in terms of time, frustration, and image quality.

 

Other combination stacking/processing programs are pretty ancient.  <smile>

 

I just don't see a good reason for going outside those two alternatives.  Certainly not to save a bit of money, insignificant for almost everyone here in terms of total budget.

Edited by bobzeq25, 01 January 2021 - 02:05 PM.

[Scott Mitchell]

Somebody recently recommended a different software package than DSS for stacking and stretching all in the same tool, but I can't find that thread. Was it APP?

APP = Astro Pixel Processor

 

It will absolutely give you better results than DSS in a similar plug and play fashion. It has lots of options to allow you to tweak the process, but the "automatic" defaults are really quite remarkable in my view.

[choward94002]

PixInsight, DSS and APP will give the best results *but* for different objects ... PixInsight for galaxies and planetary nebulae, DSS for wide field objects like SNR's, APP for larger nebulae like Horsehead, Jellyfish and Swan ...

 

Like hammers, there is no "one tool to rule them all" ... a craftsman learns to use each one where appropriate, same with stackers ... 

[Borodog]

PixInsight, DSS and APP will give the best results *but* for different objects ... PixInsight for galaxies and planetary nebulae, DSS for wide field objects like SNR's, APP for larger nebulae like Horsehead, Jellyfish and Swan ...

 

Like hammers, there is no "one tool to rule them all" ... a craftsman learns to use each one where appropriate, same with stackers ... 

Can you comment on why this would be? The wide field test shots I did of Orion I stacked in DSS suffered the same frustrating darkness and crazy color balance problems.

 

I mean, is it too much to ask that the unstretched stack resemble the lights? I can't fathom why I have to put in a bunch fiddling just to get back to where I started. And the fact that the stack is SO much darker than the lights is just thrown away dynamic range, about an order of magnitude's worth. Why???

[wxcloud]

Been using astap myself for fit files since sequator didn't like them.

[Alex McConahay]

PixInsight, DSS and APP will give the best results *but* for different objects ... PixInsight for galaxies and planetary nebulae, DSS for wide field objects like SNR's, APP for larger nebulae like Horsehead, Jellyfish and Swan ...

 

Like hammers, there is no "one tool to rule them all" ... a craftsman learns to use each one where appropriate, same with stackers ... 

I'm with Borodog on this one. 

 

Just why does one or another of these programs have an advantage over the other. 

 

Alex

Edited by Alex McConahay, 01 January 2021 - 05:53 PM.

[dcaponeii]

I’ve been thrilled with APP since I started using it about a month ago. Huge improvement over DSS. Warning it will not process 8-bit images so make sure you’re camera software will produce at least 12-bit data.

[BrendanC]

APP. Tried it alongside DSS and without a doubt it produces better stacked images. 

[Borodog]

Ok, before I shell out the bucks . . . can you guys confirm that APP stacks at least passingly resemble the lights? Not radically dimmer or bizarre colors?

[bobzeq25]

Borodog, one thing to watch out for.  A proper unstretched (linear) _stack_ will look very dark.  Because your eyes don't see linear.   Example below.  The linear stack and the final image.  This is how things are _supposed_ to look.

 

Magic.  <smile>  Note that some software automatically stretches everything so you can see it.

 

 

Edited by bobzeq25, 01 January 2021 - 07:16 PM.

[klaussius]

Borodog, one thing to watch out for.  A proper unstretched (linear) _stack_ will look very dark.  Because your eyes don't see linear.

Let me correct that. It's because your tiff viewer software is not designed to display linear tiff files correctly.

 

Our eyes don't care about the file format, they only pick up the photons.

 

As for free alternatives, I'm writing my own (free) stacking/pre/postprocessing software. It's currently a commandline without a GUI so it's not very friendly, but it's quite easy to use and has given me good results. I'm planning to write a simple GUI for it soon as well.

 

I'm not sure how it compares to the established alternatives out there, I have never done a proper comparison. It's not particularly fast though, but that's never been a problem for me.

 

I think my sig has a link to it.

[TareqPhoto]

I can buy all those to use, but the best thing is when you choose one and focus all your skills with it.

 

I was thinking about which one, in the past, i was reading some threads here about it, and at the end i saw more votes goes to PixInsight, so i didn't want to waste time to use each software individually for different targets or test each time to time, i just focused to use PixInsight and i feel like i don't need anything else.

 

In fact what i found out is that, i can live for long time using PI only and nothing else, but i can't live for long at all using only DSS or only APP without PI, so it was for me like i will use PI the most more than anything else, and on Astrobin which i look at many times i see that 90% of my choices randomly images/results end up to be done with PixInsight mainly or only, so that tells me that i shouldn't waste time, it is like when someone ask for a telescope and we tell him go use 6 telescopes, each for certain targets, that is true, but he can use at least two or three anyway, or even one with more skills and accessories to be fine with different targets.

[Borodog]

Borodog, one thing to watch out for.  A proper unstretched (linear) _stack_ will look very dark.  Because your eyes don't see linear.   Example below.  The linear stack and the final image.  This is how things are _supposed_ to look.

 

Magic.  <smile>  Note that some software automatically stretches everything so you can see it.

 

Rosette(32l,f,b,d).jpg

 

RosetteV3.2.jpg

Bob,

 

I understand this. This issue I have is that the un-stretched stack looks absolutely nothing like the lights. That is not what I am expecting to see and it makes absolutely zero sense to me. The stack should represent a higher bit-depth version of the lights with the noise calibrated and averaged out. Instead what DSS gives me is some crazy nearly black image with more noise and crazy colors. If that is the intended functionality, I don't want it. I want something that behaves intuitively. I.e. the stack looks more or less like the lights.

[Stelios]

Seriously: Get Pixinsight. 

 

Pixinsight was a bear back when there were no books for it. Today, this is not the case. Get the Warren Keller book and just follow it slavishly (you can skip the deconvolution parts in the first go). You will get much better stacking *and* processing than you are likely to get with anything else. 

 

For me, getting APP rather than Pixinsight is like buying a Toyota Corolla because it's $200 when a Mercedes S-class is $300. 

[sbharrat]

Bob,

 

I understand this. This issue I have is that the un-stretched stack looks absolutely nothing like the lights. That is not what I am expecting to see and it makes absolutely zero sense to me. The stack should represent a higher bit-depth version of the lights with the noise calibrated and averaged out. Instead what DSS gives me is some crazy nearly black image with more noise and crazy colors. If that is the intended functionality, I don't want it. I want something that behaves intuitively. I.e. the stack looks more or less like the lights.

If you have background calibration enabled in DSS, especially if you have the calibrate channels separately, the final stacked image does have the colors shifted. For RGB calibration, this mostly shows as blacker background. For separate channel calibration, the overall color can be VERY different. 

 

See http://deepskystacke...al.htm#Stacking

[Borodog]

If you have background calibration enabled in DSS, especially if you have the calibrate channels separately, the final stacked image does have the colors shifted. For RGB calibration, this mostly shows as blacker background. For separate channel calibration, the overall color can be VERY different.

See http://deepskystacke...al.htm#Stacking

Thank you; I think that is what is happening. I would never, ever, ever want to use such a “feature.”

[sbharrat]

Thank you; I think that is what is happening. I would never, ever, ever want to use such a “feature.”

Note that if you do not use any background calibration, then you shouldn't use any of the kappa sigma rejection methods. (I am not sure if winsor is usable.) So if you have subs with significantly different background levels within the sub group, you do need some sort of background calibration. 

[bobzeq25]

Let me correct that. It's because your tiff viewer software is not designed to display linear tiff files correctly.

 

Our eyes don't care about the file format, they only pick up the photons.

 

As for free alternatives, I'm writing my own (free) stacking/pre/postprocessing software. It's currently a commandline without a GUI so it's not very friendly, but it's quite easy to use and has given me good results. I'm planning to write a simple GUI for it soon as well.

 

I'm not sure how it compares to the established alternatives out there, I have never done a proper comparison. It's not particularly fast though, but that's never been a problem for me.

 

I think my sig has a link to it.

My tiff viewer software is PixInsight.  I think it knows how to do tiffs.  <smile>

[bobzeq25]

Seriously: Get Pixinsight. 

 

Pixinsight was a bear back when there were no books for it. Today, this is not the case. Get the Warren Keller book and just follow it slavishly (you can skip the deconvolution parts in the first go). You will get much better stacking *and* processing than you are likely to get with anything else. 

 

For me, getting APP rather than Pixinsight is like buying a Toyota Corolla because it's $200 when a Mercedes S-class is $300. 

I agree that, for the much experienced you, that's the analogy.

 

For a beginner, it's a _considerably_ different story.  APP is simply a far better learning tool.  That doesn't take up time the beginner needs for other things.  The fact that you (as I do) is recommending a 450 page book is not irrelevant to the argument.  <smile>

 

The cost in either case, is utterly trivial, as I've said repeatedly.

 

Very minor point compared to the above.   Many people think APP stacks as well as PixInsight, if one uses the default options for both.

Edited by bobzeq25, 01 January 2021 - 11:04 PM.

[klaussius]

My tiff viewer software is PixInsight.  I think it knows how to do tiffs.  <smile>

Monitors expect sRGB data, which uses gamma=2.2 more or less. Some monitors deviate from the standard a bit but no monitor is linear. The operating system expects all graphics to already be gamma-encoded, because all assets everywhere are encoded like that, so it's on the application to do the encoding if it is reading linear data.

 

If PixInsight reads a linear tiff and displays it as-is without applying that gamma curve, then it's doing it wrong.

 

I know it's hard to think of PixInsight doing something wrong, but that's just it. Displays operate in nonlinear sRGB color space, giving them linear data is a bug.

 

For an appication that does this right, check out gimp. You can switch to a linear color space in gimp, but you won't see a difference. Because it displays the linear data correctly. The difference is seen only when editing, since editing in linear produces different results than editing in nonlinear color spaces, as we all know.

Edited by klaussius, 01 January 2021 - 11:05 PM.