Bug in file lists: Deep sky stacker can not have same dark or bias files in several groups

[Hansi123456789]

Hi there, I have a dso object photographed with two different gain values (this corresponds to different iso settings) but the same exposures with a monochrome camera in LRGB. This means I must use several groups, say Gain1 Luminance, Gain 2 Luminance, Gain 1 Red, Gain 2 Red and so on... Now it is clear that I need two different sets of dark files for this, namely a set of darks for gain 1 and a different set of darks for gain 2.

And here the problem arises: If I put all darks into the main group, dss will stack them all together as it does only recognize different exposures but not different isos...

So the obvious way would be to use the dark 1 set in the Luminance gain 1 group and the dark 2 set in the luminance 2 group.

Unfortunately, once I have added the dark 1 set in the luminance gain 1 group, I can not add the same darks into the red gain 1 group again.

Apparently, DSS accepts the same file only either in one group only or in the main group where it is then used with all files...

This is some stupid bug.... please can you fix this? Apparently, this was done to prevent users to add the same lights for multiple times into several groups. This may be reasonable but it should not prevent you to add the same dark and bias files into several different groups

[perdrix52]

Of course you can't use the same darks in groups with different gains set.

If you have a set of lights taken with gain 1 then you need a matching set of darks taken with gain 1, both would go in one group (not the main group),

If you also want to process another set of lights taken with gain 2, then you need a different set of darks also taken with gain 2 and these would go in another group.

This sort of thing should be talked about on the DeepSkyStacker group at groups.io before you create an issue here (which you shouldn't have done because it's not a bug).

[Hansi123456789]

You have not understood...

In an LRGB setup, when I have LRGB sets with different gains, then I must add the same darkfile with gain 1 to Luminance with gain 1 and to red with gain 1 and so on.... So my single dark with gain 1 needs to go to several groups.

But because a deep sky stacker bug, I can not do this.

I can also not put this darkfile in the main group, since it should not be mixed to Luminance gain 2... do you understand this?

Currently, deep sky stacker prevents me to use the same dark in different groups if I do not want to use it in the main group, even if that would be the mathematically correct thing to do.....

And this is a bug....

Do you understand? In order to make it even simpler:

I have the following files

Luminance Gain1 Red Gain1 Green Gain1 Blue Gain1

and

Luminance Gain2 Red Gain2 Green Gain2 Blue Gain2

all with the same exposure and I have two dark files Dark Gain1 and Dark Gain2 with the correct exposure.

Clearly, it is not correct to put the darks into the main group since they would be added and averaged together.

So what I need is this grouping:

L G1, Dark G1 R G1, Dark G1 G G1, Dark G1 B G1, Dark G1

L G2, Dark G2 R G2, Dark G2 G G2, Dark G2 B G2, Dark G2

and that grouping is prevented by a bug in deep sky stacker....

[perdrix52]

No actually it's not a bug - it's a limitation of DSS which was designed to process the output of One Shot Colour cameras. and is not designed to process LRGB. Tony Cook is currtently thinking about how to add support for LRGB and Narrow Band but that's going to take a while.

In the meantime, I suggest that you process as follows:

L G1, Dark G1 R G1 G B1 B B1

to create one stack

and L G2, Dark G2 R G2 G G2 B G2

and then merge in PS or Gimp. David

[tonk777]

... or duplicate the calibration files and add the duplicates to the other groups.

OK this limitation is noted for future LRGB/NB workflows. I wont be able to start on this until around April

Tony

---

From: David C. Partridge notifications@github.com Sent: 05 January 2019 10:45:38 To: LucCoiffier/DSS Cc: Subscribed Subject: Re: [LucCoiffier/DSS] Bug in file lists: Deep sky stacker can not have same dark or bias files in several groups (#55)

No actually it's not a bug - it's a limitation of DSS which was designed to process the output of One Shot Colour cameras. and is not designed to process LRGB. Tony Cook is currtently thinking about how to add support for LRGB and Narrow Band but that's going to take a while.

In the meantime, I suggest that you process as follows:

L G1, Dark G1 R G1 G B1 B B1

to create one stack

and L G2, Dark G2 R G2 G G2 B G2

and then merge in PS or Gimp. David

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHubhttps://github.com/LucCoiffier/DSS/issues/55#issuecomment-451645164, or mute the threadhttps://github.com/notifications/unsubscribe-auth/ADr9xHoNJdwrTtjj7-hrF0z9d-QmUQ-Dks5vAIJSgaJpZM4ZlBlY.

[Hansi123456789]

In fact, even in an LRGB workflow stacking L R G B together would, in my opinion make not much sense since the L are already there. I personally just wanted to have all files in the dss ui and then select them later easily for stacking...

However, some people really create synthetic luminances with R,G and B channels. They stack all the R,G,B files together in order to get a monochrome luminance for better S/N ratios during color combination.

https://www.cloudynights.com/topic/510027-pixinsight-synthetic-luminance-question/

So in that case, if one has two sets of R G B monochrome channels with different gains and wants to create a synthetic luminance, one should be allowed to add the same bias and dark files into several groups....

The point is that adding a single dark or bias file into several groups is not complicated for a programmer implement.

DSS just needs not to do a check more precisely.

Currently, DSS checks that no file appears twice in the list. This check must simply be changed such that DSS checks tests whether darks or bias or flats do not appear twice in a single group...

There is actually not much image processing involved in this, it is just a filename check that has to be changed....

[Hansi123456789]

LRGB/NB workflows are quite hard by the way...

If you are interested in the mathematics for NB RGB combination:

Even Pixinsight has some problems with that. Its main ui features a quite lousy script...

Recently I had some success with this pixinsight plugin "emission line integration" from an astronomer: http://www.skypixels.at/pixinsight_scripts.html

http://www.skypixels.at/downloads/PixInsight/scripts/EmissionLineIntegration/EmissionLineIntegration.pdf

That script does the job of enhancing the color and luminance channel of LRGB data with narrowband quite good... (just enhancing the Color does not suffice.. the difficulty is usually to add the data to the luminance channel too...)

So you may simply copy the math from there...