WBPP - Plate Solving with Bad Data Pleiades Astrophoto PixInsight · Jean-David Gadina · ... · 13 · 179 · 3

macmade 3.01
...
· 
·  1 like
·  Share link
I recently took a few pictures of M15.
The night was clear but pretty windy, so my guiding was terrible.

I decided to take pictures anyway.
As a complete beginner, I take every possible opportunity to experiment.
Even if I end up with bad data, I'm still learning things in the process.

I took about 60 pictures with one-minute exposure before it went off-center.
I stacked them using WBPP, and plate solving failed.

I tried to alter the star detection settings without success.
Then, I tried to use SuperPixel instead of VNG as a debayer method.
Plate solving was successful, but the result was garbage anyway.

In the end, I decided to stack them without plate-solving.
Then, I sharpened the stars with BlurXTerminator and ran the Image Solver, which was successful. I could then continue with my usual workflow (like SPCC, etc.).

Now, I'm not going to use the resulting image as it is really bad, and I'll try to shoot M15 again with better conditions.
But I'm still curious to see if I could have tried something else.

If I wanted to use this data for some reason, what could I have tried in WBPP to plate solve successfully?
Or is it OK to ignore plate solving in WBPP and do it after some processing?
And why was plate solving successful with SuperPixel?

Thanks for your insights!

For reference, here's the stacked version.
Pretty bad, indeed.

CS!

M15-Master.jpg
Like
messierman3000 7.22
...
· 
·  1 like
·  Share link
that's not bad

share the stack and I'll prove it
Edited ...
Like
macmade 3.01
Topic starter
...
· 
·  1 like
·  Share link
that's not bad

share the stack and I'll prove it

Thanks for the reply!
Here's the stacked version: https://drive.proton.me/urls/9GP3SEHRDW#8HLHxp3c8AVD

I could still get something out of it, but let's say I had better results with clusters...
It would be interesting to see what you can do with it, but as mentioned in my post, I'm more interested in the plate-solving aspect than in the final processing. : )

Have fun!

M15-Final.jpg
Like
messierman3000 7.22
...
· 
·  1 like
·  Share link
successfully plate solved with:

2.90um pixel size
1723.48mm focal length

Object Identifier: M15
Like
macmade 3.01
Topic starter
...
· 
·  1 like
·  Share link
successfully plate solved with:

2.90um pixel size
1723.48mm focal length

Object Identifier: M15

Damn!
I just tried that in PixInsight, and indeed, plate-solving works.
That's great because it means I'm going to learn something new today!
Thanks a lot!

2.9um is correct, I'm using an ASI585MC Pro.
But the focal length?

I'm using a Celestron NexStar Evolution 6, which should be 1500mm.
Is it something with my image train?
Like
Bookebster 0.00
...
· 
·  Share link
Adjusting the star detection threshold in WBPP (Weighted Batch Preprocessing Script) is crucial when dealing with windy conditions that might blur or distort stars. Lowering the threshold might help detect fainter stars or improve the accuracy. subway surfers
Edited ...
Like
messierman3000 7.22
...
· 
·  1 like
·  Share link
I haven't really developed a solid workflow for globs, but here is a try:

M15 glob.jpg

kinda hazy, I know; I tried removing the noise by duplicating the same image, multiplying the pixel values by 2x on the duplicate, then applying that as a mask to the original, inverting the mask, then applying convolution to the masked image; I just made this technique up at that moment, don't follow me if you think it's too hazy.

first steps were to neutralize background, correct gradient, BXT, image solve, do SPCC, and the rest was just stretching and what I did with convolution
Jean-David Gadina:
I'm using a Celestron NexStar Evolution 6, which should be 1500mm.
Is it something with my image train?

must be
Edited ...
Like
macmade 3.01
Topic starter
...
· 
·  Share link
I haven't really developed a solid workflow for globs, but here is a try:

kinda hazy, I know; I tried removing the noise by duplicating the same image, multiplying the pixel values by 2x on the duplicate, then applying that as a mask to the original, inverting the mask, then applying convolution to the masked image; I just made this technique up at that moment, don't follow me if you think it's too hazy.

first steps were to neutralize background, correct gradient, BXT, image solve, do SPCC, and the rest was just stretching and what I did with convolution
Jean-David Gadina:
I'm using a Celestron NexStar Evolution 6, which should be 1500mm.
Is it something with my image train?

must be

Wow, gorgeous result!
It's a bit hazy yes, but so is the original data.
I really like it, you made each star really pop.

I might try to reprocess this data, see if I can achieve something close.
That would be a great exercise for me!

I still don't know what's wrong with my setup.
I had some issues recently with my Canon R7, attached to the same image train:
https://www.astrobin.com/forum/c/astrophotography/solar-system/size-of-the-moon-in-recent-pictures/

It's a bit the opposite symptom, but it's most likely related.

By the way, how did you find out the correct focal length?

Thanks a lot anyway! I really appreciate your help.
Like
macmade 3.01
Topic starter
...
· 
·  Share link
Maybe it's starting to make some sense now…
My image train is most likely too long.

Without the reducer, as with M15, this create a longer focal length.
But with the reducer, as with my Moon pictures, because the light is in a cone shape, this creates the opposite effect, reducing the focal length…

Or maybe I'm making things up…
I should have paid more attention at school in optics classes…
Like
messierman3000 7.22
...
· 
·  1 like
·  Share link
Jean-David Gadina:
By the way, how did you find out the correct focal length?


image solver put in 1500mm FL, 2.9um pixel size automatically, idk why, maybe it's related to the FITS header, if there is info in there

then, after image solver worked with that, Process Console showed what FL the image solver used, which was the number I gave you
Like
macmade 3.01
Topic starter
...
· 
·  Share link
Jean-David Gadina:
By the way, how did you find out the correct focal length?


image solver put in 1500mm FL, 2.9um pixel size automatically, idk why, maybe it's related to the FITS header, if there is info in there

then, after image solver worked with that, Process Console showed what FL the image solver used, which was the number I gave you

Yes, 1500mm comes from the FITS header.
That's how I configured N.I.N.A, since this is supposed to be the focal length of my C6.
Like
messierman3000 7.22
...
· 
·  Share link
your backfocus length might be changing the focal length
Like
macmade 3.01
Topic starter
...
· 
·  1 like
·  Share link
your backfocus length might be changing the focal length

Yes, it seems completely logical.
If it's too long, it will increase the focal length.
However, using a reducer will have the opposite effect, as I discovered with my latest Moon pictures.

I read somewhere that backfocus distance does not affect the focal length.
But it makes complete sense that it does!

Celestron recommends a distance of 5'' or 127mm without the reducer.
That's what I did, using spacers, but it looks like it's wrong...
I'll have to experiment to find the correct value...
Like
cgrobi 7.16
...
· 
·  1 like
·  Share link
I don't know about the NexStar 6" Celestron scope, but the EdgeHD 800 is listed everywhere with 2032mm focal length, because this is the 8" (203.2mm) multiplied by the f/10 ratio. In reality, the focal length is around 2150. I can't remember the exact number, but I found this written in a document about the Edges on the celestron website. This number matches the focal lengths I get when plate solving my images. So I assume, the same is true for your scope. 6" (150mm) times 10 is 1500mm. But the real focal length is a bit different.
Edited ...
Like
 
Register or login to create to post a reply.