Forum
Hi Rogelio,
Congratulations on AdaptiveWorkflow – I ran Preprocessing-OSC on its release day with a real data set and got a clean stack in about half an hour. While doing it I found two problems and have one suggestion.
Setup: AdaptiveWorkflow 1.0.0 (trial), PixInsight 1.9.5 build 1706, Windows. Data: Player One Ares-C Pro (OSC), N.I.N.A. 3.2, 119 lights x 180 s, 40 darks 180 s, 50 flats, 50 flat-darks (IMAGETYP = 'DARK', ~11.2 s), no bias frames.
1) No bias -> run fails at "Master bias"
With no bias frames the "Master bias" integration step is correctly skipped, but the next Images step reads from {workdir}/masters, which does not exist yet, so the run stops with: "Master bias: The folder H:/.../masters does not exist." Creating the empty folder by hand fixes it. Perhaps the step could treat a missing folder as an empty list, or the folder could be created at the start of the run.
2) Flat-darks never matched (exposure rounding)
"Exposure of the flats" reads EXPTIME = 11.21243 from the header, but the Flat-darks rule EXPTIME|EXPOSURE == {fexpo} is evaluated as == 11.2124 (the log shows it that way). The flat-darks have 11.21243, so 0 of 259 files pass and the flats go uncalibrated ("Flats without bias"). It looks like {variables} inside rules are formatted with 6 significant digits, as in messages. Substituting the full value, or comparing exposures with a small tolerance, would fix it. My workaround was EXPTIME|EXPOSURE < 60.
3) Suggestion: rejection algorithm by frame count
The Integrate step proposes WinsorizedSigmaClip by default, while its own prompt says ESD is better from 25 frames up. With 119 frames I switched to ESD by hand. A Branch on the number of frames could choose the default.
Minor: Check reports "writes/replaces existing files outside the work folder" for steps whose output is {workdir}/..., which looks like a false positive since {workdir} is not expanded at check time.
I can send the .awf file and the full log if useful.
Also, I couldn’t copy text from the log panel to the clipboard (Ctrl+C / paste into Notepad did nothing).
Best regards,
Sergiusz Gobiecki
Congratulations on AdaptiveWorkflow – I ran Preprocessing-OSC on its release day with a real data set and got a clean stack in about half an hour. While doing it I found two problems and have one suggestion.
Hi Sergiusz,
Thanks for the detailed report. All reported issues have been fixed. They'll go on AdaptiveWorkflow 1.0.3.
RBA