My stars won't align. after integration

Are you sure DSS shows pointy stars, can you show a zoom on that result? These look a lot like tracking problems at first glance.

And I would leave the mode at Full and let APP decide the reference frame just by itself. First try the standard settings and change when you have specific needs or see a problem with those. Maybe do change the weights to be set at "quality".

Hello, nope, it's not tracking errors. I get a normal image from my telescope. I'll leave the mode full and try again later tonight. Thanks for responding.

This is direct from my telescope.

This is from Deep Sky Tracker with additional Photoshop processing

Maybe set the registration to "distortion correction" as well, that might help sometimes. If that all fails you can send me some of the frames later for me to look at the issue. Curious to know why the direct from telescope image is much clearer than the DSS processed one..

OK, I just tried combining 10 light frames using Full mode and Distortion correction and I'm seeing the same issue.

The image from the  Stellina telescope was integrated using their built-in software. I believe it's around 430 stacked frames at 10 sec exposures. They do provide FITS files but they are only the light frames. Maybe they do some post-processing that I'm not aware of. They weren't very specific about it.

I'm not sure why the DSS export is so poor. It's probably me. I just used their default settings. I loaded 430 light frames without any calibration frames since the telescope didn't generate any.

I'll try sending you a Google drive link with some of the FITS files. 

Thanks! Alan

Here's a Google Drive link to the FITS files. It's everything that was included in the output. 

https://drive.google.com/drive/folders/1o-zTVtXST0nRa0Jo9Cx1qYVs6yzQ6pNz?usp=sharing

Does look like there is some pretty big distortion going on, I'll have a look at the data thanks!
ps. I don't have a lot of alt-az experience, isn't it the case these cause field rotation? I guess that's why you use a de-rotator, but does that completely solve all issues?

Ok, downloaded the first 20 and they indeed seem ok as far as I can tell. Trying an integration now.. and yes, getting the same issue. They all rotate which is really weird. How can you make sure the de-rotator worked, does it say that anywhere in the setup? And I'm seeing mentions online it could also be done with software sometimes, do you know if they indeed have a hardware or software solution maybe?

I'm now suspecting it is done using software, the actual telescope doesn't have a derotator which means the lights won't be corrected for that. As far as I know APP doesn't have a derotation feature.

Thanks Vincent for the support.  They claim there is a built-in field de-rotator but like you said it must be done in software.

https://vaonis.com/stellina/faq#1507449641938-89b204b4-c9a0

It looks like I'll need to find an alternative to APP. I may try Pixinsight but it doesn't have the simple interface like APP. DSS by the looks of my results doesn't seem to do a good job at stacking compared to what it does in-system on my telescope. 

I just tried the live stacking functionality in Sharpcap and it is also able to stack the FITS files. The end result isn't that great but it's able to successfully align the images. It would be great if I could make it work with APP. I just watched a tutorial on Pixinsight. It's way overkill for my purposes and I'm not sure I want to invest the time to learn.

Right, so I'm investigating it a bit further now as it would be a bit strange if all those packages do it and APP doesn't, I'll have to look if rotation is actually the problem or not, to be sure. Will come back to it! Keep the Google drive share up for a little while please, thank you.

Sure, np. Thank you for investigating this. I'm really hoping that I can use APP. 

Hi Alan @soylentgreen66 and @vincent-mod,

Thank you Alan, I have downloaded your data 😉

I know what the issue is... it's definitely not field rotation, that will never be the problem with exposures of only 10 seconds 😉

The problem is that APP's star analysis thinks that the bad pixels in the images are stars... so it registers on the bad pixels and therefor the images are not registered at all....

I will try to fix this now off course 😉 as soon as possible.

Alan, do you have bias, dark of flat frames ? Are you supposed to create this with the Stellina? Or is that all automated as well normally? Do you know if the Stellina performs bad pixel correction internally ? The Sony Exmor sensor is good for sure, but also is full of bad pixels as are most of these sensors.

Kind regards,

Mabula

Hi Mabula, I really appreciate that you are working on solving this problem.  I captured this in Bortle 8 skies on my balcony across from an apartment complex, not the best place to be capturing DSOs. Unfortunately, they only provide the light frames, no dark, bias or flat frames. I asked them about it and they said because the exposures are so short they aren't needed. I'm not sure I agree if it's unnecessary. I don't know exactly how they process the data internally. It's a mystery to me. They stack the image internally and once processed it is sent to my iPad.  They also store the raw FITS files onto a usb flash drive that I shared with you.  My plan is to capture multiple sessions of the same DSO and process it externally in a program like APP. As you have noticed the files are a bit messy and I wanted to see how good of an image I could capture using this approach. Thanks, Alan

Hi Alan @soylentgreen66 & @vincent-mod,

You are most welcome 😉

Okay, thanks for sharing that. So the good news now is, that I have further improved star detection in APP, so the next release will be able to process your data:

https://www.astropixelprocessor.com/community/release-information/astro-pixel-processor-1-075-preparing-next-release/

• IMPROVED, STAR ANALYSIS, further improvements have been implemented in star detection to make it more robust on both data with small and large FWHM values. Especially discrimination between hot pixels and small stars has been further improved. Further improvements were triggered by this topic where registration was done on hot pixels instead of stars: https://www.astropixelprocessor.com/community/main-forum/my-stars-wont-align-after-integration/

I will run APP later today on all the frames that you sent me to see what APP can make of it 😉 so I will let you know later...

Kind regards,

Mabula

I'm excited to see the results of your test! 

Posted by: @soylentgreen66

I'm excited to see the results of your test! 

Hi Alan @soylentgreen66 & @vincent-mod,

I have run the complete data set in APP now, I have deleted maybe 4-5 frames that were very bad with respect to guiding. The rest wast processed, integrated, and post-processed with APP. This is what I got, i have created a similar composition as your results:

Kind regards,

Mabula

Thank you for making it work. I'm excited now because I can try integrating data from multiple sessions so I can get better SNR. Has the update been released? I plan on buying your software now. Thanks for the fast support. Alan

Not yet released, you will see the announcement on the main page when it does.

It's not possible to make bias or flats with stellina, but you can make darks: just shoot your target for some minutes, so you are sure everithing is going well. Then cover the telescope with a sqare box (eg GItzo box) for approx 9min, not too long (ie more than 10 min) because the Stellina could shut down, due to no sight. Write down the exact time and last stacket image on your smartphone ore tablet (jpeg). Then take of teh box, and wait till the Stellina stacking kicks in again, note down the time of box removal and the number of stacked frame on your smartphone or tablet.

Please repeat this at the near end of your session, since the bayer patern of your fits may swap from BGGR to RGGB druing the session, once your scope flips over the zenit. You will not notice this. You will see this only once you start reading out the Fits 😉

Please find in attachment a more elaborated version

Indeed a great amount of bad pixels are to be detected on the sensor ...