PixInsight WBPP issue cropped up - Forums

Topic starter

Dec 23, 2023 · Edited: Dec 23, 2023

I did the latest upgrade, but I don't think this issue is related as I have processed several sets without issue. I received the following similar messages when stacking the last three sets of subs. See below:

FIRST FAILURE MESSAGE
* Loading subframe file: D:/Astronomy/MarkariansChain/debayered/Light_BIN-1_4144x2822_EXPOSURE-300.00s_FILTER-UVIR_CFA/Light_Markarians Chain_300.0s_Bin1_294MC_UVIR_gain120_20231223-051024_14.0F_0016_c_d.xisfLoading image: w=4144 h=2822 n=3 RGB Float323 image properties87 FITS keyword(s) extracted Measuring subframe: D:/Astronomy/MarkariansChain/debayered/Light_BIN-1_4144x2822_EXPOSURE-300.00s_FILTER-UVIR_CFA/Light_Markarians Chain_300.0s_Bin1_294MC_UVIR_gain120_20231223-051024_14.0F_0016_c_d.xisf** Warning: Noise estimates are not available in the image metadata and are being calculated from possibly non-raw or uncalibrated data. Image weights can be inaccurate.** Warning: PSF signal estimates are not available in the image metadata and are being calculated from possibly non-raw or uncalibrated data. Image weights can be inaccurate.** Warning: Noise scaling factors are not available in the image metadata and are being calculated from possibly non-raw or uncalibrated data. Image weights can be inaccurate.Stars detected: 30Valid PSF fits: 30 (100.00%)

SECOND FAILURE MESSAGE
D:/Astronomy/MarkariansChain/debayered/Light_BIN-1_4144x2822_EXPOSURE-300.00s_FILTER-UVIR_CFA/Light_Markarians Chain_300.0s_Bin1_294MC_UVIR_gain120_20231223-034858_14.0F_0001_c_d.xisfDescriptor failed:** Warning [162]: C:/Program Files/PixInsight/src/scripts/WeightedBatchPreprocessing/WeightedBatchPreprocessing-engine.js, line 6748: reference to undefined property descriptor.FWHMFWHM : undefinedeccentricity : undefinednumberOfStars : undefinedPSFSignalWeight : undefinedPSFSNR : undefinedSNR : undefinedmedian : undefinedmad : undefinedMstar : undefined** Warning: Failed to measure frame - image will be ignored.

Any ideas?

Jeff Rothstein

Dec 23, 2023

Hi David. Could you describe how you set up WBPP (use of bias, darks, flats), and any diagnostic messages you received before you clicked Run? The first failure message sounds to me like you were missing some calibration frames. The second could also follow from missing calibration frames. (I've seen the first - always due to missing cal frames- but not the second.)

Best,

Jeff

Well Written Insightful Respectful Concise

Gaillard Jérôme

Dec 23, 2023

Have you tried to reinstall Pixinsight 1.8.1.2 ? Sometimes, it's a just this kind of problem…

DavesView

Topic starter

Dec 23, 2023 · Edited: Dec 23, 2023

I may have found the problem. I ran the subs through Astro Pixel Processor and it shows the FWHM at above 12 on all the subs. I think I'm out of focus. I just started using a Clear Focusing filter and then switch to a UV/IR or L-eXtreme filter after focusing.

Does that sound plausible?

Well Written Insightful Respectful Concise Engaging

noon

Dec 23, 2023

That is plausible. Before I run WBPP I Blink all my raw files to make sure they are all in the same ballpark and delete (“move”) those dirty files to a Trash folder where I delete them from later. That avoids situations where subs are out of focus, invasive powerlines, or low flying aircraft spoil a sub and throw off WBPP

Helpful Concise

noon

Dec 23, 2023

Also I often get the “Warning: PSF signal estimates are not available in the image metadata and are being calculated from possibly non-raw or uncalibrated data” message. I’m not sure that’s anything to worry about

dmatson1

Dec 23, 2023

I was also getting that first error on many subs when I only used WBPP for calibration and debatering. I found that if I leave “Subframe Weighting” checked in the lights tab of WBPP I don't get that error in later steps of preprocessing. This seems to just write some extra metadata to the subs, and then Subframe Selector is much happier.

Well Written Helpful Concise

DavesView

Topic starter

Dec 23, 2023

For my settings I use all presets and choose Maximum Quality and have always (sounds longer than the reality) gone this route because I'm ignorant

(unknowing). I noticed on my flat frames that it shows them as <1 second, but shows my bias frames as 0.0 seconds. I always (again with the always) insure that the flats and bias frames match length. Also, out of processing maybe 50 sets of subs, I have never received the warning messages before these last three sets. For D/F/B I use 30 each.

That is plausible. Before I run WBPP I Blink all my raw files to make sure they are all in the same ballpark and delete (“move”) those dirty files to a Trash folder where I delete them from later. That avoids situations where subs are out of focus, invasive powerlines, or low flying aircraft spoil a sub and throw off WBPP

Me too, but I haven't really looked for star size... just the obviously abnormal.

Dan Kearl

Dec 23, 2023

If you have APP why don't you just use it.
WBPP does not stack as well for me at all.
I have both, stack each target in both and the APP result 90% of the time is better.

DavesView

Topic starter

Dec 23, 2023 · Edited: Dec 23, 2023

Jeff Rothstein:
Hi David. Could you describe how you set up WBPP (use of bias, darks, flats), and any diagnostic messages you received before you clicked Run? The first failure message sounds to me like you were missing some calibration frames. The second could also follow from missing calibration frames. (I've seen the first - always due to missing cal frames- but not the second.)

Best,

Jeff

Jeff nailed it! So I was receiving a message in the 'continue window' after choosing 'run' that stated I had Registration turned off. I went through the settings and everything was turned off, including registration. Turned on what I thought was appropriate and voila, it works. Not sure how it all got turned off.

Thanks folks!

DavesView

Topic starter

Dec 23, 2023 · Edited: Dec 23, 2023

Dan Kearl:
If you have APP why don't you just use it.
WBPP does not stack as well for me at all.
I have both, stack each target in both and the APP result 90% of the time is better.

I like APP and I definitely like the stacking 'sessions' that you can apply your subs and supporting frames to for multiple nights on the same target. PI probably has some way to do that via name association or some other method. I get a little lost when using the tools in APP and just haven't taken the time to go to Youtube University for that particular course. I really like PI though and like all of my tools being under one roof. I'm 11 months into this hobby and I've never had any kind of photography related learning, so... But hey, I don't have to know it since you guys do.

Dan Kearl

Dec 23, 2023 · Edited: Dec 23, 2023

I use no tools in APP, just stack and take the fits file to Pixinsight. Way more simple than WBPP.
I use Pixinsight for a bunch of stuff and move on to photoshop and Capture one.
Pixinsight has a ways to go to reach Photoshop type processing IMO, for final processing.
It has tools that Pixinsight has no answer for.

noon

Dec 23, 2023 · Edited: Dec 24, 2023

I do the same thing as Dan as far as Photoshop goes; although I use WBPP in PI instead of APP. Photoshop has its own leaning curve though. I’ve been using it since ‘96 but I imagine it would be a little daunting starting from scratch at this point. That being said there are plenty of videos and tutorials online to help these days.

DavesView

Topic starter

Dec 23, 2023

Dan Kearl:
I use no tools in APP, just stack and take the fits file to Pixinsight. Way more simple than WBPP.
I use Pixinsight for a bunch of stuff and move on to photoshop and Capture one.
Pixinsight has a ways to go to reach Photoshop type processing IMO, for final processing.
It has tools that Pixinsight has no answer for.

When I stack in APP and without using any tools, just take the fits file and process it in PI, I find that the astrometric data is missing. I then use Image Solver to write the data to the file so as to use SPCC.

scotty38

Dec 26, 2023

DavesView:
When I stack in APP and without using any tools, just take the fits file and process it in PI, I find that the astrometric data is missing. I then use Image Solver to write the data to the file so as to use SPCC.

Is this still an issue as I reported it to the devs maybe 2 years ago and at the time they said the fix was coming......

Timothy Prospero

Dec 26, 2023

DavesView:
When I stack in APP and without using any tools, just take the fits file and process it in PI, I find that the astrometric data is missing. I then use Image Solver to write the data to the file so as to use SPCC.

Is this still an issue as I reported it to the devs maybe 2 years ago and at the time they said the fix was coming......

Using WBPP will solve this issue, it performs plate solving during the pre-processing. Therefore, after you have obtained the astrometric solution, the only issue arises when combining images (ie RGB or NB) using pixel math, as it can result in the loss of the astrometry solution. I'm not sure if I missed something ther.

But nowdays after running WBPP, the astrometry solution is there.

DavesView

Topic starter

Dec 26, 2023 · Edited: Dec 26, 2023

Timothy Prospero:
DavesView:
When I stack in APP and without using any tools, just take the fits file and process it in PI, I find that the astrometric data is missing. I then use Image Solver to write the data to the file so as to use SPCC.

Is this still an issue as I reported it to the devs maybe 2 years ago and at the time they said the fix was coming......

Using WBPP will solve this issue, it performs plate solving during the pre-processing. Therefore, after you have obtained the astrometric solution, the only issue arises when combining images (ie RGB or NB) using pixel math, as it can result in the loss of the astrometry solution. I'm not sure if I missed something ther.

But nowdays after running WBPP, the astrometry solution is there.

Yes! I use WBPP for that reason. I use APP as a backup. I also use APP to stack if I have multiple nights with different flats and bias subs because of the 'multiple sessions' feature. Haven't figured out how to do it in PI.

EDIT: Let me rephrase that. I haven't taken the time to figure it out or research how to do it in PI.
research = ask the gurus

Concise

DavesView

Topic starter

Dec 26, 2023 · Edited: Dec 26, 2023

DavesView:
When I stack in APP and without using any tools, just take the fits file and process it in PI, I find that the astrometric data is missing. I then use Image Solver to write the data to the file so as to use SPCC.

Is this still an issue as I reported it to the devs maybe 2 years ago and at the time they said the fix was coming......

It is an issue with my copy of APP. I'm running 2.0.0-beta22 on a Win11 machine.

Well Written

Timothy Prospero

Dec 26, 2023

DavesView:
Timothy Prospero:
DavesView:
When I stack in APP and without using any tools, just take the fits file and process it in PI, I find that the astrometric data is missing. I then use Image Solver to write the data to the file so as to use SPCC.

Is this still an issue as I reported it to the devs maybe 2 years ago and at the time they said the fix was coming......

Using WBPP will solve this issue, it performs plate solving during the pre-processing. Therefore, after you have obtained the astrometric solution, the only issue arises when combining images (ie RGB or NB) using pixel math, as it can result in the loss of the astrometry solution. I'm not sure if I missed something ther.

But nowdays after running WBPP, the astrometry solution is there.

Yes! I use WBPP for that reason. I use APP as a backup. I also use APP to stack if I have multiple nights with different flats and bias subs because of the 'multiple sessions' feature. Haven't figured out how to do it in PI.

EDIT: Let me rephrase that. I haven't taken the time to figure it out or research how to do it in PI.
research = ask the gurus

Hi Dave, it's actually quite straightforward. Here's how I do it: You can store your different nights in separate folders, for example, Night-1, Night-2, and so on (including, flats, darkflat). In WBPP, create a grouping keyword like 'Night (pre-processing only)'. Then, simply select your files using the 'Add +Directory' option and choose the folder where all the night folders are located. That's all there is to it!

Well Written Helpful

scotty38

Dec 26, 2023

DavesView:
DavesView:
When I stack in APP and without using any tools, just take the fits file and process it in PI, I find that the astrometric data is missing. I then use Image Solver to write the data to the file so as to use SPCC.

Is this still an issue as I reported it to the devs maybe 2 years ago and at the time they said the fix was coming......

It is an issue with my copy of APP. I'm running 2.0.0-beta22 on a Win11 machine.

I'm not sure what version of the 2.0 beta was that I used before I let my licence lapse. I got fed up of waiting for the headers fix plus the documentation........

Terry Doran

Jan 2, 2024

DavesView:
I may have found the problem. I ran the subs through Astro Pixel Processor and it shows the FWHM at above 12 on all the subs. I think I'm out of focus. I just started using a Clear Focusing filter and then switch to a UV/IR or L-eXtreme filter after focusing.

Does that sound plausible?

Two rules I never break; 1 never focus with no filter and then add a filter and start shooting. 2 never focus with one filter then switch filters and start shooting. Filters can change focus pretty significantly.

Helpful Concise

DavesView

Topic starter

Jan 2, 2024

Terry Doran:
Two rules I never break; 1 never focus with no filter and then add a filter and start shooting. 2 never focus with one filter then switch filters and start shooting. Filters can change focus pretty significantly.

1) Me either.
2) I wasn't, but had trouble seeing the diffraction spikes when using the L-eXtreme filter. Did some research and was assured that using a focus filter of the same thickness as the intended filter was exactly what this focus filter is for. I've been checking it after I switch back and it looks good, but only as far as I can tell. Truth is, I'm about half blind. Maybe the glass I need should be resting on my nose.

Terry Doran

Jan 2, 2024

DavesView:
Terry Doran:
Two rules I never break; 1 never focus with no filter and then add a filter and start shooting. 2 never focus with one filter then switch filters and start shooting. Filters can change focus pretty significantly.

1) Me either.
2) I wasn't, but had trouble seeing the diffraction spikes when using the L-eXtreme filter. Did some research and was assured that using a focus filter of the same thickness as the intended filter was exactly what this focus filter is for. I've been checking it after I switch back and it looks good, but only as far as I can tell. Truth is, I'm about half blind. Maybe the glass I need should be resting on my nose.

Sounds like maybe it’s time for an auto focused.

DavesView

Topic starter

Jan 3, 2024

Terry Doran:

Sounds like maybe it’s time for an auto focused.

Arrives tomorrow.