WBPP fails on the final process and does not create master light

Richard Sauerbrun

Hello Adam,

I am a FastTrack user and have been following the videos and updating my WBPP configuration choices. I have used PI and WBPP for a couple years, and am seeing some improvements in the master lights, but also have experienced some major issues.

On May 29th I processed 90 120s lights with accompanying flats and dark flats. I used an ASI2600MC Pro which has almost zero amp glow so I skipped including any darks. I shot the images with two different filters (Askar C1 & C2 - 45 subs each) so I used the FILTER keyword and has WBPP set up to create two master lights.

I am attaching a ZIP with several files that hopefully will help you see what happened. Reviewing the log, I see at the 15:16:58 mark that WBPP complete the Debayer step with 90 succeeded, 0 failed, and 0 skipped.

What I see next in the log I believe is the Local Normalization, but I see no errors as the cache was created successfully. At 15:16:59 I see "End demosaicing of light frames" and then nothing more.

One of the screen shots shows the process monitor with the message "Abort Requested" ... but I had not touched the PC to make such request. In fact, I let the machine run for several hours and saw this when I returned. The Windows task manager showed PixInsight as active but using 0% CPU.

Any help would be very much appreciated ... thanks,

Richard

BlockHead

Hmmm... I have no idea. I have not seen an unsolicited abort before. 

I can't even guess why this would happen- never seen anyone with this issue.

It is curious the logfile ends *exactly* after debayering. 

-the Blockhead

Richard Sauerbrun

Measurements is next in the pipeline ... I wonder if that killed it?

I am running the entire WBPP process again with the exact same input data and configuration choices. I will share the results.

Thank you for your excellent timeliness and support.

Richard Sauerbrun

This time it worked, and I got two master lights (C1 & C2). It took over 5 hours, and I ended up with 34 subs in the C1 image and 10 subs in the C2 image.

Now I can give try at extracting the channels and then combining for a final image.

Very interesting that these files failed once, but worked the second time.

BlockHead

Well... the possible issue in this case is hardware. (Memory)

The numbers are the same..and algorithms are the same and most important is that computers are deterministic machines. 

-the Blockhead