Compose module needs alignment

[Splutnik Andrew]

Hi,
I've tried creating an LRGB composite in COMPOSE, but the calibrated colour frames are not aligned and I can't find any way to do it, and nothing in the tutorials or documentation that I've been able to find. I've attached a close-up of the brightest star to show what I mean.
This means I have to do the alignment in another program, so I might as well do the compositing in the other program too.
I gather many people use dithering these days, so the colour frames will very often not be aligned before compositing. It would be great if Star Tools could do the alignment.
Thanks,
Andrew

[Startrek]

Startools is not Stacking software and I’m glad it’s not. There are loads of stacking programs out there to prepare your data set for Startools.

I’ve started using Siril , then DSS for about 5 years and now use ASTAP which I think is the best stacker around ( as good as Pixinsight )

ASTAP runs on any platform, it can plate solve , stack and align your data with excellent results.

You may have missed some important information in the Startools documentation regarding preparing your data set ready for Startools.

https://www.startools.org/links--tutori ... -and-donts

I prefer to use my own individual programs in Astrophotography ( including acquisition software , I don’t use asiAir ) so Startools is the icing on the cake at the end of the Astro daisy chain.

Give ASTAP a try and then load your calibrated , stacked and aligned data set into Startools , I’m sure you will enjoy its simplicity and powerful algorithms.
I have a procedure for stacking and aligning Mono data sets in ASTAP if you decide to pursue this path.

Clear Skies
Martin

[Carles]

Startools is not Stacking software and I’m glad it’s not. There are loads of stacking programs out there to prepare your data set for Startools.

I’ve started using Siril , then DSS for about 5 years and now use ASTAP which I think is the best stacker around ( as good as Pixinsight )

ASTAP runs on any platform, it can plate solve , stack and align your data with excellent results.

You may have missed some important information in the Startools documentation regarding preparing your data set ready for Startools.

https://www.startools.org/links--tutori ... -and-donts

I prefer to use my own individual programs in Astrophotography ( including acquisition software , I don’t use asiAir ) so Startools is the icing on the cake at the end of the Astro daisy chain.

Give ASTAP a try and then load your calibrated , stacked and aligned data set into Startools , I’m sure you will enjoy its simplicity and powerful algorithms.
I have a procedure for stacking and aligning Mono data sets in ASTAP if you decide to pursue this path.

Clear Skies
Martin

Hi,

i do have stacking issues when aligning with ASTAP for subsequent use on NBAccent for example.

Align artifact.png (640.38 KiB) Viewed 17033 times

On the right is the original Lextreme file, on the left, is the aligned with ASTAP. You can see the "broken" star along with the "junction" artifact.
Have you encountered those?

Regards,

Carles.

[Startrek]

I’ve mostly used ASTAP with Mono data sets inclusive of all filters ( L, R, G, B , Ha , Oiii and Sii ) and never had an issue with alignment using ST Compose. This includes an LRGBHa composed data set.

Heres my basic workflow for stacking in ASTAP ….,
I visually check my all lights ( fits ) in ASTAP for satellite trails plus zoom in to check poor star shape caused by atmospheric turbulence ie: guiding hiccups. Discard bad subs
Then just load all my lights , Flats , Flat darks and Darks for a specific filter
Analyse my lights and remove any super low value / low score subs
Create a master dark
Create master Flat calibrated with matching flat dark
Stack using Sigma Clip
Save
Repeat above for each filter
Then load completed stacked files ( each filter ) into lights
Select Calibration and Alignment only
Save fully stacked and aligned data set

Load data set into ST Compose

Maybe Han from ASTAP can offer an explanation or solution , possibly something to do with OSC Dual-band filters ?
Do you split your Ha and Oiii from the Dualband data and then align separately in ASTAP using Calibration and Alignment algorithm prior to loading into Compose.
Bear in mind Im still using ST V 1.8 ( not V1.9 )

Ivo from ST and Han from ASTAP would probably know the issue and have a solution.

https://sourceforge.net/p/astap-program/discussion/


Clear Skies
Martin

[Splutnik Andrew]

Hi, thanks for the insights Martin & Carles.

I've given ASTAP a go, and it improved matters, but not perfectly yet. Attached is the latest star close-up.
I expect there's a solution to this, I'll keep trying.

[decay]

Hi Andrew,

I would guess, that this is caused by atmospheric refraction:
https://en.wikipedia.org/wiki/Atmospheric_refraction

This star seems to be nicely round, so this is most probably no problem which is caused by stacking. Another possibility is chromatic aberration of your optics (if glass / lenses are involved when using a refractor scope or a field flattener), especially if this star is located towards the edges of your image / FOV.

If the alignment of this colour pattern is equal throughout your whole image (independent of the location of the stars) it's most likely due to atmospheric refraction (or a strong tilt).

If the pattern is concentric to the optical axis (e.g. diagonal green, white, red/purple) at the upper right corner and the other way round (red/purple, white, green) at the lower left corner, it's chromatic aberration, I would say.

Best regards, Dietmar.

[Startrek]

Andrew,
I concur with Dietmar,
At first glance it looks like chromatic aberration of the optics or possible aberration caused by imaging at low altitudes lower than 30deg.
Chromatic aberration usually shows blue/ greens on the bottom and red hues on the top of the star across the image field.
When you loaded your LRGB data set into Compose what Luminance , Color did you set ? Also did you set the exposure sliders based on your total time captured for each filter or did you leave them as default 3600sec 1hr for each filter ?
In regard to aberrations…..,
I’ve experienced a similar phenomenon with my Newtonian’s as well on the medium to larger Stars probably caused by stray converging light rays from the OTA spider vanes and internal flocking not even throughout the tube.
You can correct this or minimise its affect in post processing with Startools as follows -

Some useful tools in Startools to colour correct annoying star colours and star halos are as follows -

Colour Module
You set full mask or specific Star mask
“Highlight Repair”can help to colour correct abnormal star colour but works best with Narrowband data sets
Increase from 6 pixels to see the effect or repair


Shrink Module
Star Colour correction using Shrink Module
Create a good Star Mask and ensure all star cores and halos that require colour correction ( including diffraction spikes ) are sampled or green. Grow mask as required

Open Shrink
Adjust iterations to 0
Change Mode to Tighten
Color Taming increase to between 4 and 10 pixels as required.
Note: Be careful not to increase Color Taming too much as this can cause Star cores to bleach out colour or “white out”. Start off with 2 pixels , then 4 pixels and so on until Star colour is reasonably uniform in the core and halo area.
Halo Extend increase to 1 or 2 pixels
Un glow Kernel leave off
Abnormal Stars core and halo colour ( and diffraction spikes ) should morph or blend into a uniform colour closest to adjoining colour of star core.

Filter Module
Filter module can minimise unsightly and abnormal colour around halos (medium to large stars)
Create a good Star Mask , stars green or sampled
Select “Fringe Killer” and 3x3 Sampling method and adjust mask fuzz accordingly
Click on the stars halo and/or main core area, after each click colour will change closer to star colour.

Hope the above is helpful

Clear Skies
Martin

PS: Maybe Ivo if he’s browsing can review this post as well

[admin]

Late to the party, but indeed, the latest could be some sort of chromatic aberration.
All great suggestions (thanks Martin). The Filter module's Fringe Killer option may be a good solution here.