15th Feb 2024: Astro Pixel Processor 2.0.0-beta29 released - macOS native File Chooser, macOS CMD-Q fixed, read-only Fits on network fixed and other bug fixes
7th December 2023: added payment option Alipay to purchase Astro Pixel Processor from China, Hong Kong, Macau, Taiwan, Korea, Japan and other countries where Alipay is used.
My problem is that when I use a narrowband demosaic e.g. Ha-OIII extract OIII, calibration fails.
My setup
Target NGC2237 101x60sec 125gain
Capture RedCat 51. Optolong L-eNhance Dual Narrowband. ZWO ASI183MC Pro
Processing MacOS Mojave. APP 1.083 or 2.0.0-beta12
I will illustrate the problem with some images with maximum stretch (30% BG, 2 sigma, 0,0 base) and neutralize-BG
I fire up APP (either 1.083 or 2.0.0-beta12) and do minimal configuration.
1) Load: no multi channel or session
2) Calibrate: scale MasterDark
Next I load up my masters (I have a full set BPM,MF,MD,MDF,MB) and finally my 101 lights.
Looking at the first light it clearly needs calibrating
So I view l-calibrated
much flatter and without amp glow. If I integrate with this I get a satisfactory result.
However I want to play with false colour so need to extract Ha and OIII channels.
In 0) RAW/FITS I select the algorithm Ha-OIII extract OIII
and again look at linear
and l-calibrated data
everything looks good so I integrate but get this
Something is terribly wrong but if I check a single l-calibrated frame it still looks O.K.
After going round in circles for several days I discovered that having selected Ha-OIII extract OIII if I go to 2) Calibrate and (re-)assign Masters to Lights and again look at the l-calibrated frame I get this
The amp glow is back again although it is still flat. It's almost as if the dark isn't being subtracted but it is more subtle than that as if in 2) Calibrate I untick scale MasterDark I get this
which makes sense to me as my MD are 300sec and if unscaled would massively over compensate my 60sec lights.
I have experimented with both versions of APP and if I select any algorithm from Hydrogen beta onwards, reassign my masters and then display l-calibrated data the fault reappears and remains even if I switch algorithm again.
So I conjecture that even though everything looked O.K. when I originally checked my l-calibrated OIII, that once I let APP process is did it's own master reassignment.
Either that or I am doing something dumb, but in any case I need help.
Dear @nickwest,
The issue is that dark scaling should always be disabled in case of severe amp-glow like yours.
You really need to provide matching darks in terms of sensor gain + offset & temperature & exposure time to be able to get rid of amp-glow the correct way.
The reason is that apm-glow is a non-linear signal, it does not double if you expose 2x longer 😉 So using dark scaling (a linear correction of the dark current signal) will never work nicely with clear amp-glow present.
Mabula
Hi Mabula,
Thanks for explaining that the issue is that dark scaling doesn't work for me as my amp glow is not linear, but I'm more confused than ever. Recently I took some M42 with my Optolong L-eNhance Dual Narrowband. I used 15sec subs as I did not want to burn out the core so, from what you have told me, the scaling of my 300sec darks will be even worse. However, if I debayer using the Adaptive Airy Disk then I get this.
Again maximum over stretch to expose calibration flaws. It doesn't look bad to me.
Of course, looking at the built-in help, I am told that Ha-OIII is better with my filter, but with that I get
Now the dark scaling has broken down. Naively, I thought that as dark noise isn't "coloured", it has not been subjected to any optical filter, so the noise subtracted would be the same regardless of the debayering method. Clearly I am mistaken, but I should really like to know how I appear to get away with scaling with AAD but not when I use the correct Ha-OIII.
Thanks,
Nick
Hi Mabula,
I know you are very busy but I really would like to understand the demosaic narrow band data better.
Thanks,
Nick
Dear Nick @nickwest
Sorry for the delay in my response.
Did youtry to provide matching darks for your lights frames? Was the result better?
I have no clue why AAD is not showing the amp-glow versus Ha-O3 extract O3 showing it without seeing the data myself. But as I said, any issue relating the amp-glow will be solved if you would supply 100% matching darks for the lights and disable dark-scaling.
Hi Mabula,
I have been out in India for 2 months so haven't yet had a chance to match exposures with my darks. I'll do that when I get home. I have seen this strange behaviour with two data sets now, is there really nothing in the way debayering influences calibration that could give us a clue? Of course fixing the problem is the priority but understanding the technology is part of the fun. If I continue to see this effect would you like me to upload a data set for you to study?
Thanks,
Nick
