- This topic has 43 replies, 6 voices, and was last updated 9 years ago by
.
-
Posts
-
I think something similar has happen to me in the past (master FLAT not applied whileI was under the impression it was); e.g. see issue reported in the first posting in the thread https://www.starry-night.nl/forums/topic/m3-bolhoop.
So do I got it right when I follow the following flow for mono-subs using color filters?
1) Create master BIAS (app. 100 -250 subs), clear all frames
2) Create master DARK (app. 50 subs) while loading master BIAS as well, clear all frames
3) Create master FLAT (app. 50 subs) while loading master BIAS and master DARK, clear all frames
4) Create Bad Pixel Map while loader master BIAS, master DARK and master FLAT
5) Calibrate subs while loading, Bad Pixel Map master BIAS, master DARK and master FLAT, clear all frames
6a) For additional colour subs only: Load reference frame and perform star analyses on all frames and unselect reference frame and select it under Registration as reference frame
6b) For first colour subs only: perform Star Analyses and remember selected reference frame
7) Select Windsor Clipping in Integrate and press Stack to Register, Normalise and Integrate frames of this colour
8) Repeat step 3, 4, 5, 6 and 7 for each color filter applied
9) When needed correct light pollution, vignetting and combine colours to a single RGB stack
10) Optimise photo using preferred tooling…
Let me know if I understood it correctly and if not, where I go wrong…
in my first reaction to @mauricetoet‘s question in this post I explained everything I think. As everyone could see, Maurice was able to perfectly calibrate his f/2.8 DSLR data with flat & light frames which had different iso values
Your explanation was perfectly clear to me, so I don’t know why others are struggling. Before actually stacking the data, you can get instant feedback from APP to check if the calibration files were probably prepared. I cannot imagine a clearer warning signal.
BTW. If you use darks, I won’t bother using a masterbias to correct the lights and darks (as the bias offset is both in the lights and darks). To keep things simple and clear, just bias-correct the flats.
Hoi @musquetier,
Your multi-channel workflow is correct I would think, although, I have 3 remarks:
1) when you have applied flats, the correct vignetting step is not needed, since you should have corrected the vignetting of your data with the flats ? So the correct vignetting function is only needed when the user doesn’t have adequate flats at that moment in time. For instance, the tool can be used for a quick review of your data if you haven’t made flats yet. Or if you are struggling with proper flat correction with really short focal lengths ( less than 200mm).
2) if you need to apply dark frame calibration/correction on your lights due to Fixed Pattern Noise in the dark current (which you have with the asi1600), don’t use bias, the bias is always in the darks, like @mauricetoet indicates as well.
3) Regarding the flats, either subtract a masterbias or subtract a master darkflat (or is it flatdark…) of the same offset and gain as the flats. ?
2) and 3) are both recommended versus applying both bias and dark subtraction.
If you do subtract both bias and dark masters, then always convince yourself that the bias is subtracted from the darks in the masterdark creation, otherwise you subtract twice the bias pedestal which you don’t want to do.
Thank you @mauricetoet, yes, I feel the same way. Using the l-calibrated image viewer option immediately tells you if calibration is working correctly. I always check it before I continue to star analysis.
On the other hand, while creating the master frames, it will be most helpfull to give warnings to the user if they follow a calibration workflow in which the bias correction is not done properly for both light and flat frames. I really feel the bias pedestal is the root cause of most calibration issues, especially proper flat-field correction. So like @keesscherer recommended, I will try to create a warning system in the calibration engine which could be shut of as well.
1) Indeed, so far vignetting hasn’t been a problem when I applied the flats!
2 & 3) So if I understand correctly I should load a Master BIAS when I create the Master DARK’s. And I should not load a Master BIAS when I create the Master FLAT’s, right?
1) great
2) no just the other way round actually ?
“So if I understand correctly I should load a Master BIAS when I create the Master DARK’s”
No, preferably not ? if you are going to use darks, it’s best to not use a master bias. So don’t load the master bias when you create the master dark. And don’t load the master bias when you calibrate the lights then.
And if you are going to make a master flat, always use a master bias or a master darkflat/flatdark ? to calibrate your flats.
Workflow for multi-color mono subs correct with Mabula’s remarks:
1) Create master BIAS (app. 100 -250 subs), clear all frames
2) Create master DARK (app. 50 subs), clear all frames
3) Create master FLAT (app. 50 subs) while loading master BIAS, clear all frames
4) Create Bad Pixel Map while loading master DARK and master FLAT, clear all frames
5) Calibrate subs while loading Bad Pixel Map, master DARK and master FLAT
6a) For additional colour subs only: Load reference frame and perform star analyses on all frames and unselect reference frame and select it under Registration as reference frame
6b) For first colour subs only: perform Star Analyses and remember selected reference frame
7) Select Windsor Clipping in Integrate and press Stack to Register, Normalise and Integrate frames of this colour
8) Repeat step 3, 4, 5, 6 and 7 for each color filter applied
9) When needed correct light pollution, vignetting (when no FLAT’s were used) and combine colours to a single RGB stack
10) Optimise photo using preferred tooling…
Or, if I’m right, you can use 150 darks with the bias in it which is what I do just to have a step less. But it takes a while to get those ofcourse.
You can either do:
[light – n * dark] / [m * (flat – masterbias)]
or
[(light – masterbias) – n * (dark – masterbias)] / [m * (flat – masterbias)]
where:
n * dark = a masterdark with the bias offset
n * (dark – masterbias) = a masterdark without the bias offsetAlmost perfect @musquetier, one final remark, at 4) create Bad Pixel Map. you don’t need to load the master bias. The master bias is not used in detection of bad pixels.
The bad pixel map can be created from master frames, like you are describing in this workflow. You can also create the bad pixel map from individual flat and dark frames as well ?
Or, if I’m right, you can use 150 darks with the bias in it which is what I do just to have a step less. But it takes a while to get those ofcourse.
For light frame calibration, yes, only use darks and flats then ?
But for the calibration of your flats, you’ll need to use either bias or darkflats/flatdarks (which is it?).
(Personally, I rarely use bias frames. Normally, I calibrate my light frames with darks and BPM and flats. And I calibrate my flat frames with flat darks and BPM.)
You can either do: [light – n * dark] / [m * (flat – masterbias)] or [(light – masterbias) – n * (dark – masterbias)] / [m * (flat – masterbias)] where: n * dark = a masterdark with the bias offset n * (dark – masterbias) = a masterdark without the bias offset
Indeed, excellent @mauricetoet ;-)
Almost perfect @musquetier, one final remark, at 4) create Bad Pixel Map. you don’t need to load the master bias. The master bias is not used in detection of bad pixels. The bad pixel map can be created from master frames, like you are describing in this workflow. You can also create the bad pixel map from individual flat and dark frames as well ?
ok, corrected now in the posting above…
(which is it?). (Personally, I rarely use bias frames. Normally, I calibrate my light frames with darks and BPM and flats. And I calibrate my flat frames with flat darks and BPM.)
Indeed, for flats I use flat-darks and the BPM, I do it like you describe.
- You must be logged in to reply to this topic.