Problem with flats mono CCD

[Thalestris24]

I had a quick look at some of them - seem ok. I'm wondering if the flats problem is specific to the above dataset? You could maybe try taking some more flats with the exposure shifted eg to, say, 35000 adu, and see if that's better. Maybe also try stacking in dss just to compare.

Louise

[tooth_dr]

14 minutes ago, Thalestris24 said:

I had a quick look at some of them - seem ok. I'm wondering if the flats problem is specific to the above dataset? You could maybe try taking some more flats with the exposure shifted eg to, say, 35000 adu, and see if that's better. Maybe also try stacking in dss just to compare.

Louise

Thanks Louise. I’ll try that tomorrow, it’s really stormy here so I’ll stay inside at the fire. 

[Thalestris24]

26 minutes ago, tooth_dr said:

Thanks Louise. I’ll try that tomorrow, it’s really stormy here so I’ll stay inside at the fire. 

Storm Erik is spoiling all our fun! Windy washout weekend here

[Xiga]

2 hours ago, tooth_dr said:

@vlaiv

Here are the single fits, the flats and dark flats were taken on a few days ago, the darks on Christmas night (not sure what I was doing taking darks then lol)

 

 

 

 

DF_2019-02-03_22-55-06_Bin1x1_8s__-15C.fit

F_2019-02-03_20-22-36_Bin1x1_8s__-16C.fit

D_2018-12-25_20-31-51_Bin1x1_300s__-15C.fit

Not being a CCD user, i'm not sure if any of this is material or not, but i notice in the Fits header info that the Gain for the Darks is 0, but it's -1 for the Flats and Dark Flats. Also, the temp of the Flat is about -1 degree C colder than the other two. 

Could either of these be the culprit i wonder? I would at least try and get the temp of the Flats a bit closer, just in case that's it. 

I don't see any info regarding Offset. Are you absolutely sure all of the files (both calibration and lights) have been taken with the same Offset? 

Weirdly, only the Dark has an entry for 'Telescope' in the fits header info. The other two are missing this option entirely. 

It's a strange one alright. But right now my hunch is that there's nothing wrong with your optical train or gear, and that the most likely reason is just something silly that's gone wrong during acquisition of some of the subs. 

[tooth_dr]

23 minutes ago, Xiga said:

Not being a CCD user, i'm not sure if any of this is material or not, but i notice in the Fits header info that the Gain for the Darks is 0, but it's -1 for the Flats and Dark Flats. Also, the temp of the Flat is about -1 degree C colder than the other two. 

Could either of these be the culprit i wonder? I would at least try and get the temp of the Flats a bit closer, just in case that's it. 

I don't see any info regarding Offset. Are you absolutely sure all of the files (both calibration and lights) have been taken with the same Offset? 

Weirdly, only the Dark has an entry for 'Telescope' in the fits header info. The other two are missing this option entirely. 

It's a strange one alright. But right now my hunch is that there's nothing wrong with your optical train or gear, and that the most likely reason is just something silly that's gone wrong during acquisition of some of the subs. 

Hi Ciaran

Thanks for taking a look at this. I’ve retaken my darks and bias, the fits headers may different now? APT 3.60 was used for the F and DF so I wonder if the new gain feature did that? As for temp, some of the lights are at -16, the camera fluctuates +/- 1 deg

Edited February 8, 2019 by tooth_dr

[vlaiv]

Something is very odd with those dark subs.

Here are my findings:

Average background value for first and second dark at 300s is very much different in intensity - it should be roughly the same, if you take two darks only difference between them should be random noise and their average value should be about the same.

I subtracted bias from both and got ~910.4 ADU in first, while only ~494.8 ADU in second dark.

We can use short dark (flat dark) to do cross check which one might be the right value, if any. We can also look at the specs for your camera and see what should be expected dark current value at this temperature and exposure.

Short dark (8s ) after bias removal has average value of ~6.65 ADU.

Dark current should be linear in time, so if 8s dark produces 6.65 average ADU value, 300s one should produce 300/8 * 6.65 = 249.375 ADU

None of the above seem to have that value - first one has more than 3 times that amount, while second has about two times that amount - this is very strange and should not happen!

Looking around for QHY9 specs, I can't seem to find system e/ADU (gain) value anywhere? Nor can I find dark current at -15C? This is something manufacturers of cameras provide. But I did stumble upon something interesting - there is debate over gain and offset settings for that camera. I was under impression that CCD cameras don't have this - system gain and offset should be factory set with no user intervention.

Do you have these values, and can you change gain and offset? And most importantly did you change those values?

 

 

[vlaiv]

According to this thread, there are in fact offset and gain values:

What values do you have set on your camera, and more importantly did you change them at any point?

[Xiga]

8 minutes ago, tooth_dr said:

Hi Ciaran

Thanks for taking a look at this. I’ve retaken my darks and bias, the fits headers may different now? APT 3.60 was used for the F and DF so I wonder if the new gain feature did that? As for temp, some of the lights are at -16, the camera fluctuates +/- 1 deg

Did you use APT for all of the subs? Lights and calibration files? 

It sounds like Gain is a recent addition to the Fits header info. Clutching at straws a bit here, but you could maybe check your APT Log file with a fine tooth (no pun intended! lol) comb. Look specifically for Gain and Offset values for all the calibration files and Lights and make sure they match. 

Or if you'd like i'd be happy to take a look at the data for you, just to see if i get the same outcome. Even just a subset of the full data should suffice. Say 10 subs of each would do. 

[tooth_dr]

8 minutes ago, Xiga said:

Did you use APT for all of the subs? Lights and calibration files? 

It sounds like Gain is a recent addition to the Fits header info. Clutching at straws a bit here, but you could maybe check your APT Log file with a fine tooth (no pun intended! lol) comb. Look specifically for Gain and Offset values for all the calibration files and Lights and make sure they match. 

Or if you'd like i'd be happy to take a look at the data for you, just to see if i get the same outcome. Even just a subset of the full data should suffice. Say 10 subs of each would do. 

Thanks Ciarán. I’ve quickly checked the new darks and bias and they have that value of -1 matching the flats. It must be the new APT settings. I only use APT for everything for data collection. I’m currently stacking 182 lights from 2 nights  so will report back tomorrow (depending on laptop speed!)

[tooth_dr]

I thought I would come back to this thread and update it.

Previously I had just checked the first few flats, they all looked the same, so I thought nothing more of it.  Tonight I went through them all one by one, and I found 8 flats that looked differently exposed to the others.  I have removed these and currently am stacking the data.  Hopefully this will resolve the issue.  I've attached examples if the outliers.

 

 

[mcbbcn]

I don’t know if this will help, but for my QHY9M, I use offset 115 and Gain 15.

Best,

Miquel

[martin_h]

I use offset 100 and gain 25%

[tooth_dr]

Thanks guys. Everything has taken a back seat over the summer. The values are I think specific to your camera. You manually work them out using a process I found online.