Aligning Channels in Preparing for L,RGB

[SBRToolMichigan]

Hello All!

I'm rolling right along with getting some great results from ST already but I have question. I'm wondering what everyone is using to allign/register their L,RGB channels before loading in to ST? When I chose the recommended DSS settings, I find that I'm no longer able to use the 'Align' function on the result in Nebulosity as I've done in the past. Not sure why it won't work any more...the white balance option unticked in DSS maybe?

Anyway, I've seen some posts from a few years ago suggesting to use a reference frame in DSS when stacking the separate channels. Have there been any new/better options for registering since then? Is an "align" function being considered as a feature of ST in the future Ivo?

Having an absolute blast with StarTools!

Thanks...
-Craig

[admin]

Hi Craig,

Great to hear you and StarTools are getting along well!

It is very important to align all channels during stacking. That's because aligning individual frames will always be more precise than aligning just 3 channels after the fact. This increased precision is particularly important for things like deconvolution (making sure that any resulting aggregate luminance frame has a constant PSF that does not vary in different parts of the image), but also for coloring (making sure that color fringing does not occur).

These benefits are particularly important/noticeable for stackers that implement distortion correction of the individual frames before stacking.

Hope this makes sense and explains adequately why such functionality would be sub-optimal at the time of post-processing.

[SBRToolMichigan]

Thanks Ivo!

Is your reply supporting the use of the reference frame method in DSS to align the channels there as I read from the older posts?

For a little background, here's my basic workflow when processing narrowband subs:

- All Ha subs are stacked in DSS and saved as FITS to
a file.
- All OIII subs are stacked in DSS and saved as FITS to
a separate file.
- I then open ST in the Compose mode, L,RGB.
- I then load say the Red channel with the saved Ha file
- However, when I attempt to load say the Green channel
with the saved OIII file, I either see that the star color is
not lining up or I just get the Failed message "The
dimension of the bitmap you're trying to load..."

Before changing my settings in DSS from saving files as 16 bit to 32 bit integer as recommended, I was able to align the Ha and OIII saved files using Nebulosity before further processing. Everything worked fine in loading and processing in ST. However, it doesn't appear that Nebulosity can work with 32 bit integer files so aligning in that program is no longer an option(?!)

If I understand your reply correctly, this post-stacking alignment in another program isn't the best method to produce quality results anyway?

So, I guess I'm trying to find out the best (or only way?) to get the L, RGB function to work with properly aligned channels in ST. If it's the reference frame or other method in DSS I'll be happy to give it a try.

Thanks so much Ivo!
-Craig

[admin]

Is your reply supporting the use of the reference frame method in DSS to align the channels there as I read from the older posts?

That's correct!

For a little background, here's my basic workflow when processing narrowband subs:

- All Ha subs are stacked in DSS and saved as FITS to
a file.
- All OIII subs are stacked in DSS and saved as FITS to
a separate file.
- I then open ST in the Compose mode, L,RGB.
- I then load say the Red channel with the saved Ha file
- However, when I attempt to load say the Green channel
with the saved OIII file, I either see that the star color is
not lining up or I just get the Failed message "The
dimension of the bitmap you're trying to load..."

Indeed, this is where you will want to use one stack (for example the Ha stack) as a reference for when your are stacking the other stacks (for example the O-III frames.

Before changing my settings in DSS from saving files as 16 bit to 32 bit integer as recommended, I was able to align the Ha and OIII saved files using Nebulosity before further processing. Everything worked fine in loading and processing in ST. However, it doesn't appear that Nebulosity can work with 32 bit integer files so aligning in that program is no longer an option(?!)

Depending on how deep your data is, 16-bit may be enough.

If I understand your reply correctly, this post-stacking alignment in another program isn't the best method to produce quality results anyway?

Correct, it is very much sub-optimal!

So, I guess I'm trying to find out the best (or only way?) to get the L, RGB function to work with properly aligned channels in ST. If it's the reference frame or other method in DSS I'll be happy to give it a try.

That's it! Do let us know how you get on Craig!

[SBRToolMichigan]

Ok! I'll forge ahead and report back. Thank you SO much Ivo!

Craig

[philhilo]

Hi Ivo and Chris,

Hope you had a good Christmas, and thanks Ivo for the big update on ST, one day I hope to get the hang of it all!

My issue is the same as Chris, i.e aligning channels for input into ST. I have read this thread and the help file and seen the alluded to 'align in DSS method', but what is it? I have tried adding the output from the Lum stack as the reference frame for the next stack in DSS and checked it so it got used (which feels wrong - putting the lum result in the red stack). Result DSS wouldn't stack, 'The checked pictures are not compatible......'.

I tried again with the lum result as the reference frame but this time didn't check it so it didn't get added to the result. This time DSS ran but the result wasnt accepted in ST - wrong size etc.

What is the method of alignment?

Cheers,

Phil

[Stefan B]

...I have tried adding the output from the Lum stack as the reference frame for the next stack in DSS and checked it so it got used (which feels wrong - putting the lum result in the red stack). Result DSS wouldn't stack, 'The checked pictures are not compatible......'.

I think you should not check it, since then DSS tries to use it as a single frame and stack it. Which is not possible and not what you want. You should probably uncheck it and tell DSS to use it as a reference frame nonetheless. At least that's what I do (additionally I don't use a stack as a reference but a single frame from - in you case - the luminance subs).

Regards
Stefan

[happy-kat]

When I use a batch of images but stack in groups I add the reference frame to each separate DSS stack. I mark this frame as the reference frame but only include it (ticked) on the one stack that I want, it is referenced in the later stacks and marked as the reference frame but not ticked.
I did this for a comet stack because I used three stacks, each stack had every third frame for example.

[philhilo]

Thanks for the replies folks. Its 5.15 and I have finally figured it. It was the stacking method. The guide recommends intersection mode, which is always going to produce a file size dependent on the alignment of the files processed. If the files were really offset then there would only be a small intersection, if perfectly aligned then an intersection the size of the files. So I swapped to standard mode which sets the reference frame as the size so all the stacks will be the same size and it works. Another small step on the journey, another late night!
Cheers.

[happy-kat]

That's a good spot as I over missed that I always use standard mode as I much prefer to control what I crop off the edges in post processing.