16 replies to this topic

[HMS_Endeavour]

I am trying to calibrate my narrowband images, and for the first time I am using WBPP

 

I have my per filter (SHO) lights and flats selected, plus regular darks.

 

But when I try to add my per filter dark flats, they all get lumped together. So it looks like I have a bunch of darks at one exposure, and a section set of darks at another exposure. 

 

Flats get separated by filter (then by exposure) but my dark flats dont. 

 

am I doing something wrong? Are my dark flats associated with the flats or not?

 

 

Attached Thumbnails

• 
• 
• 

[afd33]

In the darks tab there should be a spot for exposure tolerence. Set it to zero and see if that fixes it.

[HMS_Endeavour]

Whats concerning me is on the pipeline page, integration of the 180 dark frames says NoFilter

 

 

 

Attached Thumbnails

• 

[HMS_Endeavour]

In the darks tab there should be a spot for exposure tolerence. Set it to zero and see if that fixes it.

Thanks

that does separate the 180 into the 3x60 groups of different exposures, but its still not recognizing them as dark flats, rather than darks

[afd33]

Whats concerning me is on the pipeline page, integration of the 180 dark frames says NoFilter

 

In that case you'll probably want to use the script batchfitskeyword edit. Then you'll change edit the fits keyword in each of the flat frames to have a filter associated with it.

 

Pixinsight isn't going to call them darkflats. They're darks. If they're not properly being assigned to the flats, you'll either have to do what the above, or you can manually assign them to each group of flats in the calibration page.

Edited by afd33, 19 June 2025 - 04:26 PM.

[Alex McConahay]

Thanks

that does separate the 180 into the 3x60 groups of different exposures, but its still not recognizing them as dark flats, rather than darks

As AFD33 points out in the post #5, there is no such thing as "dark flats."  This common misnomer is improperly applied to "flat darks." 

 

The noun is "dark." The type of dark that is being described is one used to process flats. But it is a dark. The word "flat" in "flat darks" modifies "darks."

 

A flat is taken with light hitting the sensor. With a dark, no light is supposed to hit the sensor. A dark records only camera noise (generally speaking). A Flat records both camera noise and the optical pattern of a flat light source. Manipulate the flat with a dark, and you can see how the optical system affects the light. Knowing that, you can properly calibrate the Light frames taken of your celestial targets. 

 

Alex

[HMS_Endeavour]

As AFD33 points out in the post #5, there is no such thing as "dark flats."  This common misnomer is improperly applied to "flat darks." 

 

The noun is "dark." The type of dark that is being described is one used to process flats. But it is a dark. The word "flat" in "flat darks" modifies "darks."

 

A flat is taken with light hitting the sensor. With a dark, no light is supposed to hit the sensor. A dark records only camera noise (generally speaking). A Flat records both camera noise and the optical pattern of a flat light source. Manipulate the flat with a dark, and you can see how the optical system affects the light. Knowing that, you can properly calibrate the Light frames taken of your celestial targets. 

 

Alex

yeah, lets get bogged down in pedantry, when you know exactly what I mean, and what I am trying to do

lol

Edited by HMS_Endeavour, 19 June 2025 - 05:58 PM.

[HMS_Endeavour]

 

 

Pixinsight isn't going to call them darkflats. They're darks.

Thanks, you've been helpful - I got them working

 

 

NINA flatwizard calls them darks - ie

 

DARK_Ha_0.23s_139g_50off_0004.fit

Edited by HMS_Endeavour, 19 June 2025 - 06:00 PM.

[bignerdguy]

The dark flats need to have the filter keyword embedded in them to be able to be automatically assigned. It looks like from the images you posted the darks are all without the filter keyword.  PI matches the darks, flats, and lights by several variables, the Filter keyword is one of them.

[HMS_Endeavour]

The dark flats need to have the filter keyword embedded in them to be able to be automatically assigned. It looks like from the images you posted the darks are all without the filter keyword.  PI matches the darks, flats, and lights by several variables, the Filter keyword is one of them.

according to the FITS Keyword script. it has the filter in there:

Attached Thumbnails

• 

[KuiperBeltKing]

If the NINA flatwizard just went back to creating a separate folder for the flat darks a lot of this confusion would be avoided.

[PIEJr]

If the NINA flatwizard just went back to creating a separate folder for the flat darks a lot of this confusion would be avoided.

Confusion is good though. It keeps people talking.

 

 

I simply run a set with NINA Flats Wizard (Flats, and Darks) Then while the telescope is capped anyway, fire off a set of BIAS.

 

Whooska-Bushka, good for a month or more of knockin out images.

I've been using 30 as my magic number (Bias, Flats, Darks) and letting ASI Studio do the blending with whatever number of Lights I load up.

 

Very simple, and the outcome pleases me.

[bignerdguy]

according to the FITS Keyword script. it has the filter in there:

Well the only reason it would not pair the flat-dark with the flat is if either the filter, exposure, iso, or some other settings don't fully match.  I have used flat darks and flats with different filters and as long as everything has matched and i do not use Bias frames it matches the proper flat-dark with the correct flats every time.

[bignerdguy]

I just tried to see if i could replicate it and initially it wasn't doing it.  i realized i had a keyword grouping selected and it filtered it.  when i removed that it grouped all similar file types together.  try this, add a keyword to the far right side of WBPP that says FILTER and set it for PRE not POST and it should separate them.

 

Without keyword:

 

 

With keyword

 

 

I should note here i used ISOSPEED since all my filters were the same .  you should use the keyword FILTER instead to do what you are wanting.

[afd33]

I didn’t even think of a keyword doing it but that makes a lot of sense. Glad you got it figured out.

[dswtan]

I am trying to calibrate my narrowband images, and for the first time I am using WBPP

...

am I doing something wrong? Are my dark flats associated with the flats or not?

Maybe OP is happy now they solved the "problem", but note there was no problem in the first place.

 

First, darks are dark. No light. So a "filter" is irrelevant. This is why I have believed PI just notes "NoFilter" in the Pipeline view of WBPP. This is what it shows in my WBPP stacks too, for ALL my darks, even though the flat darks have a filter noted in the FITS header. It's just not used and not relevant. 

 

Likewise, you don't always need to be exact in matching flat darks, since the exposure time for flats can be relatively short (OP's are especially short). PI/WBPP takes this into account with the "Exposure Tolerance" in the Darks tab, as has been noted in the thread. 2 is the default, which works for common situations.

 

With 2 as a default, it will group exposures like your 0.21-0.35 sec. This is because there is practically no difference in the signal (noise) level with modern sensors within that range of exposures. So PI/WBPP just groups them all together labeled as the longest duration one. 

 

Mine similarly, I have 0.09-0.4sec flat darks (from my LRGB filters). It just bundles them to 0.4. It works fine. 

 

If you have much longer duration flat darks, e.g. for a narrowband filter (I appreciate this didn't happen for OP with their short durations), then it will split them out and match them to the flats if they are above the tolerance setting. But you don't have to worry about any bundling going on normally. 

 

My longer duration NB filters are not bundled, since their exposures are longer duration than the tolerance (13, 7, 4 secs).

[bignerdguy]

First, darks are dark. No light. So a "filter" is irrelevant. This is why I have believed PI just notes "NoFilter" in the Pipeline view of WBPP. This is what it shows in my WBPP stacks too, for ALL my darks, even though the flat darks have a filter noted in the FITS header. It's just not used and not relevant. 

Pi doesn't list any actual filters in WBPP for darks, but that doesn't mean that it'll simply apply the correct dark for the correct light or flat by default. In order for the flats and the lights to get the correct dark file, you actually have to filter with a key word, such as, for example, FILTER or ISOSPEED or some other name that's listed in the Fits header. Otherwise, like the OP showed, it will just group all the dark files together into one single file that are below the exposure tolerance setting.  This won't work correctly if you have multiple lights and multiple flats with different time scales, especially if they are below the minimum setting, like in the OP's example. Flat darks do need to have the same exposure time and other settings as the actual flat in order for it to correctly apply to the flat automatically. You also need to make sure that you are not using a bias file as if there is a bias file, WBPP will default to using that instead of a flat dark.