Access violation when trying LRGB stacking

[Gary H]

Hi Han, hope you're well.
Thanks for the excellent software.
When tying to stack RG and B subs I get the above message.
It will stack just R or G or B fine but fails when trying to stack them at once.
Any suggestions.
This happens on latest version and also the developer version
Win 10 - 64 bit - 256 SSD - 8 gb ram

[han.k]

Hi Gary,

Difficult to trace without having the files involved. Can you prepare a test set to test it here using the debugger?

Note I just released version 2023-06-17 but that will likely not fix it.

Cheers, Han

[Gary H]

cheers Han, not sure what you mean ?
where is the debugger ?

[han.k]

When I run the program here in de compiler debugger, it will tell me exactly where the error occurs in the code. To generate the error here I have to use the same files as you used. So if you can share the files and send me a link to your files or a subset,then I can test it here and find the cause of the error.

[Gary H]

oh i see, ill put them in google drive later.
thanks Han.

[Gary H]

https://drive.google.com/drive/folders/1EjDuUId6IUBWdUZo9qBYpDahaGNH5M2e?usp=drive_link
I've only put a few subs in each folder as the same error happens no matter how many subs.

[han.k]

I found the cause. It tries to save the interim file with a name including the filter name. The forward slash in G/Oiii is causing problems. I will build a filter in the filter name handling. It will be ready in the next hours (for Windows).

Han

[Gary H]

Ah of course, well found 👍🏻
Vielen Dank

[han.k]

This version works with your files:

http://www.hnsky.org/astap_setup.exe

It will remove these characters from the filter-name: '.', '\', '/', '*', '"', ':', '|', '<', '>'

Cheers, Han

[Gary H]

Thanks Han, if id have thought about it i could have kust changed the file names 🤣

[han.k]

Feedback helps. ASTAP is now a tiny amount more robust.

[Gary H]

Thanks Han, that version works !
Excellent software and amazing support.

[Gary H]

Is there a way to choose the reference sub for aligning in LRGB mode ?
Im trying to combine my subs with a freinds but astap finds my R and G files to be best and my freinds B file to be best.
When it combines the 3 files at the end the colours are misaligned.
thanks again Han.

[han.k]

For each colour is selects the best file as reference. This result in the combined R, G, B and L files. Then in the final step it will use luminance to combine LRGB and red for RGB.

In some cases you could use astrometric alignment to fix problems.

There is no manual override. Share the 3 or 4 files where is goes wrong and I will have a look. No problem to investigate. A good test case.

[Gary H]

Thanks Han, i've put the files in ...
https://drive.google.com/drive/folders/1EjDuUId6IUBWdUZo9qBYpDahaGNH5M2e?usp=drive_link
4 of mine and 4 of freinds each colour.

I'm wondering if it's because they are poor quality subs ?

I'm just trying to get the hang of stacking / combining 2 different data sets as we are going to try a collaboration.
Both use ZWO 1600mm pro, both similar focal length (350mm) both 30 seconds 139 gain.

I've tried settings as below

[han.k]

It hangs in the black spot filter. Should not happen. I will investigate.

[han.k]

The binning you better set at 2 instead of 1.

Black spot filter:
It hangs because in the corners the green filter is different orientated. The green is missing in the corners (=purple) and ASTAP is looking around for green substitute which is not there. I will limit the search to avoid that. Otherwise it takes ages.

[han.k]

This image:

[han.k]

I have reduced the search range. So it will work in the next version. Attached is what I got out of it (cropped). I have used only half of the red files since there was white diagonal disturbance in them:

The skies in northern Europe are now during the summer nights too bright for faint objects like M101.

Han

[Gary H]

Thanks Han, you really do like a challenge :-)
Like I said, I'm just preparing / researching ways to stack from different cams.
In all honesty the skies where I live are always poor :-(

So, when we decide on our project, we will have to try to ensure both scopes are orientated similarly, this will help the software deal with the different angle and slightly different image size.