My dark and bias median ADU is higher than my lights, pedestal not working.

Lead_weight

Member
Jul 9, 2019
5
0
This is for sure a challenge for me. My bias and dark are about 2000 median ADU above my lights. What sort of trickery do I need to do to get these calibrated? I can add 1000 ADU with the Pedestal function, but still seems to not be enough to fix the issue.

Here's a link to 10 each of my calibration and light frames. Please let me know if you're able to work this issue, and how to do it.
 

bulrichl

PTeam Member
Nov 2, 2016
938
81
La Palma, Canary Islands
The FITS headers of a bias, a dark and a light frame tells:
Code:
Bias:  COMMENT                            "Generated by Indi"
Dark:  COMMENT                            "Generated by Indi"
Light: SWCREATE    'N.I.N.A. 1.10.1.90'   "Software that created this file"
Flat:  SWCREATE    'N.I.N.A. 1.10.1.90'   "Software that created this file"
Bias and dark frames were captured with Indi, light and flat frames were captured with NINA. Maybe that Indi and NINA use different camera drivers (ASCOM / native) as well - I don't know.

As you already stated, the bias offsets of bias frames (median: 3200 ADU) and dark frames (median: 3228 ADU) do not match the bias offset of the light frames (median 1692 ADU).

The light and flat frames (captured with NINA) are not compatible with the bias and dark frames (captured with Indi) although according to the FITS header all frames were captured at -10 °C, at gain 120 and offset 50. The dark and bias frames on hand cannot be used to correctly calibrate the light and flat frames.

Generally you will have to capture light frames and ALL calibration frames with the same software and the same camera driver.

Bernd
 

Lead_weight

Member
Jul 9, 2019
5
0
Hi Bernd, thank you for taking a look. I think I have solved this issue by using a program that let me put in a custom pedestal value on the lights of 1800 ADU. I tried manually adding a pedestal value to my fits headers, but it seemed PixInsight didn't like this, or maybe I did something wrong there. But I could add a custom value in Astro Pixel Processor, and got the images to process there.

Incidentally I reshot all my master images using NINA, and the results were the same when processing. So despite the drivers being different, they had the same result.

Some astute observers on Cloudy Nights think I put my offset too high, and I'm going to use a lower value while I examine a histogram to make sure it doesn't offset the signal too far off the left side. A little more flexibility with PixInsight's pedestal setting and I would have been able to fix them there.
 

bulrichl

PTeam Member
Nov 2, 2016
938
81
La Palma, Canary Islands
In this case, applying a pedestal is not a valid solution of your calibration issues, that is messing around with not compatible calibration frames. You will not achieve a correct flat field correction either with this approach.

I cannot imagine at all that - with matching calibration frames - such a large pedestal would be needed. It is on purpose that an output pedestal cannot exceed 1000 ADU in PixInsight, this value is sufficient for any regular project data with matching calibration files. Please upload one of the bias and dark frames captured with NINA, I would like to take a look.

The offset which was set in the camera driver is not the culprit at all IF the offset was the same for all frames. According to the FITS headers that I looked at, this condition was met.

Bernd
 

Lead_weight

Member
Jul 9, 2019
5
0
Here is 10 images Light, Dark, Bias, Flat from the sequence. These are the redone master images all taken with NINA. They did calibrate out an LRGB set and a previous narrowband set of lights with no issue, but for some reason these HA frames wouldn't calibrate out. The median ADU values of the lights was a lot lower than the bias and dark. I agree, the pedestal value isn't a great solution, but it did fully correct out, and you can see the final image here. I'm interested to see what you think of these frames and why there's such a large discrepancy. I'm in a bortle 7/8 zone, which makes the lights median ADU value even more baffling.
 

bulrichl

PTeam Member
Nov 2, 2016
938
81
La Palma, Canary Islands
You are right, bias and dark frames captured with both Indi and NINA (at - 10 °C, at gain 120 and offset 50) have the same high median of about 3200 ADU. So the problematic frames are the light frames (and potentially the flat frames).

I have no explanation why the Ha frames, reportedly captured at the same settings, should have a median of 1692 ADU. It looks as if a lower offset value than that documented in the FITS header (offset 50) was actually set in the camera. This is very strange.

Bernd
 

RodrigoQ

Member
Jul 11, 2020
10
1
I don't know If I'm facing the same problem as you @Lead_weight. I got my very first mono camera ( ASI294MM Pro) and I shoot almost one hour on Ha for The Rossete Nebula. I follow the procedure and for some reason is impossible to me using the WPBB. If I just try to calibrate my light frames the result (Light Calibrated) are pitch black images, If configure the whole WPBB tool it get cancelled in the middle of the process. I'm using N.I.N.A to acquire the images.

Following the data information
Lights : 20 x 180 s - Gain 180 - Offset 60
Darks : 15 x 180 s - Gain 180 - Offset 60
Flats : 30 x 3 s - Gain 180 - Offset 60
Dark Flats: 30 x 3 s - Gain 180 - Offset 60
Bias : 73 - Gain 180 - Offset 60

The images attached have STF applied.

Also right here is the link with the images:


I'm wondering as well if I miss some special configuration in N.I.N.A and I got bad data at the beginning of the process.

Let me know your thoughts!

Best Regards.

Rodrigo Quiroga
 

Attachments

bulrichl

PTeam Member
Nov 2, 2016
938
81
La Palma, Canary Islands
Presumably your dark frames suffer from light leakage.

Bernd

P.S.
These are the statistics of a light frame and a dark frame:
Code:
            Light        Dark
count (%)   100.00000   100.00000
count (px)  11694368    11694368
mean        3984.8      4341.4
median      3972.0      4340.0
variance    118623.0    10377.7
stdDev      344.4       101.9
avgDev      69.8        90.2
MAD         65.2        89.0
minimum     3480.0      3776.0
maximum     65532.0     65532.0
The frames are captured at the same gain (180), offset (60) and exposure time (180 s). However, the median of the dark frame is greater than the median of the light frame by 368 ADU. The STF Auto-stretched dark frame (appended) shows nonuniform exposure to light.

Dark.JPG
 
Last edited: