Some Questions

Hi Tim,

Thank you for your compliments 😉 and your questions.

I registered all 140 subs together to get a common reference frame. When I tried to save the registered frames it would only save a few – never more than 40. I tried numerous times.

The most likely cause, for only saving part of the frames, is that not all frames were registered. What do you see in the bottom panel in the Registration RMS column? Do all frames have a RMS value and star pair count? If so, then all frames should be saved., if that doesn't happen, we need to investigate further.

I found the same when trying to save the normalized frames. Then I tried to integrate the Ha, O111 and S11 separately by leaving the check marks by the stack that I wanted to integrate and removing the others. I couldn't figure out how to tick and untick the boxes in groups. Doing it one by one is very slow and if you go too fast you get a java error. I was able to do it successfully by going slow. Is there a better way?

Same reason, for the normalized frames, probably part of the frames weren't registered?

If you want to combine the different channels for a compositie. The most easy workflow is to create the integrations/stacks of the separate channels first. Then register the integrations of the channels and save those. Then you can load them into the RGB combine tool to make a composite.

Your workflow should work however. If registration fails, try to increase the star count in 3) Analyse stars by lowering the detect above noise level kappa.

In 4) enable distortion correction, possibly increase the scale stop to 10. If all frames were shot with the same camera and optics (and if the frames weren't registered before) enable same camera and optics, otherwise disable it.

Let me know if you can register all frames, APP should really be able to do that, even in case of strong optical distortion.

I tried to figure out what the various settings did in “Analyse Stars”. I don't know what “Filter Star Profile” does but I often got Java errors when that was ticked. So, I just stopped using it. Reducing the “clip star Profile” and the Kappa setting increased the number of stars. However, running with the lowest Kappa (01) would often result in less subs being registered than using a Kappa of 03. So does reducing the Kappa do more than increase stars? I finally settled on a Kappa of 03 to 05 with the lowest clip star profile. This allowed all subs to be registered. Never had any problems with these settings. However, I did get other random Java errors when the Kappa was set to 01.

Yes, the filter star profile is not needed in most cases, and can give a java error on very small stars  (FWHM less than 2 pixels) which is likely the case with your narrowband data?

Indeed, the kappa value should normally not be lowered below 3. The faint stars that can be detected just above the noise floor are not very suited for accurate registration because the intensity profiles are strongly dominated by noise.

Decreasing the clip star profile increases the star count. Normally 0.1 is fine, only lower it if you need to detect more stars, but then it should still work without problems.

So yes, the detect above noise level kappa can be problematic if you set it too low (lower than 3), so is the filter star profile in case of very small (undersampled) stars.

When Registering images what does the “Flip Descriptors in X/Y” mean. It doesn't seem to be needed for meridian flips. Is it only for mosaics? Is there any advantage to using the “same camera and optics” setting? Does optics refer to the scope or filter as well? Can you use it for different nights and/or different sub lengths?

Flip descriptors in X/Y is needed in cases where you want to register data of different camera's/ota's which have some of the data flipped in either the x or y-axes. This can be the case when you combine data of an apo with a certain type of reflector. An extensive example of when you need this, is shown in:

https://www.astropixelprocessor.com/registration-normalization-integration-using-ddc-lnc-mbb/

So it's not a setting related to mosaics in particular.

The same camera and optics setting is usefull to have robuster registration in the case of stron optical distortion, but you should only enable it in the case where the frames that you register are from the same optical configuration. That means, the same camera and optics. Filters have no influence in this case 😉 But if the frames were registered before by APP or any other program, it's safer to disable the setting, because the registration engine actually is a camera technically speaking and can warp the data in all sort of ways depending on the particular registration engine that processed the data. And yes, you can have it enabled for frames that were shot during different nights and sub lenghts if you shot with the same camera and optics combination. Only be carefull in that case, if you are using optics that don't have steady collimation. So if you adjusted the collimation between sessions, it's safer to disable "same camera and optics".

The java errors that I mentioned above were easily cleared by hitting ok and continuing. I kept screen shots of these errors and sometimes the relative console. If you are interested – I can send them to you.

Please send them to me, so I knwo exactly what you saw and have this properly fixed 😉 You can send them to support@astropixelprocessor.com.

Thank you Tim for all your questions, I hope my reply answers most of your questions. Let me know if you have more questions.

Kind regards,

Mabula

Thanks Mabula,  for the explanations and suggestions. I'm sending screenshots of the the java errors to support. 

Thank you Tim 😉 I have received them.

I reran some of my testing and I found that APP will only save registered files up to the first file not registered. For example I just tried to register 20 subs.  Only sub 3 was not registered . When I saved the files APP only saved the first and second ones. It didn't continue with the forth and the other correctly registered subs. So, after you register your subs you have to make sure that any subs that were not registered are unticked before you try to save anything.  Then all your registered subs will be saved.  Not sure if this is obvious to everyone else but I thought it worth mentioning.

Mabula, On the java errors, I'm finding its relate to slider thresholds not being met to force it and any step that doesn't actually complete. Its easily reproducable in the analyse stars step, use subs with off round stars and set kappa to 1 (lowest) and you get a java error, click away error and step finishes with a "completed" message in the console.

BUT

when you look at the file list none of them have STARS next to the names so it didn't complete successfully after all. This happens on any step where it fails for some reason we get a finished successful message in the console after a java error popup and no tagging changes in the file list. So I guess its poor error handling that causes the java popup (first problem) and console reporting a successful finish of the step (second problem). Look around those, as i said easy to test and repeat.

Posted by: Tim

I reran some of my testing and I found that APP will only save registered files up to the first file not registered. For example I just tried to register 20 subs.  Only sub 3 was not registered . When I saved the files APP only saved the first and second ones. It didn't continue with the forth and the other correctly registered subs. So, after you register your subs you have to make sure that any subs that were not registered are unticked before you try to save anything.  Then all your registered subs will be saved.  Not sure if this is obvious to everyone else but I thought it worth mentioning.

Hi Tim,

Thank you for your feedback, I'll try to fix this as soon as possible, it's on my TODO list.

Mabula

Posted by: Steve

Mabula, On the java errors, I'm finding its relate to slider thresholds not being met to force it and any step that doesn't actually complete. Its easily reproducable in the analyse stars step, use subs with off round stars and set kappa to 1 (lowest) and you get a java error, click away error and step finishes with a "completed" message in the console.

BUT

when you look at the file list none of them have STARS next to the names so it didn't complete successfully after all. This happens on any step where it fails for some reason we get a finished successful message in the console after a java error popup and no tagging changes in the file list. So I guess its poor error handling that causes the java popup (first problem) and console reporting a successful finish of the step (second problem). Look around those, as i said easy to test and repeat.

Thank you Steve for your feedback. I'll try to have it handled properly.

Kind regards,

Mabula

Hi Tim and Steve,

I have located and fixed the Error: java.lang. ArrayIndexOutOfBoundsException: 0 in 3) ANALYSE STARS.

Working to solve the other reported problems now 😉

Mabula