Strange Integration Result

@mfranks_160 Hi Mike, the red and blue wobbly lines are caused by hot pixels that walk across the image due to dithering or tracking imperfections. how often do you use the camera? It seems unlikely that there suddenly are much more hot pixels than before. You wrote that the camera settings haven't changed so can you please confirm that the darks and bias have been shot at -5ºC as well? Unfortunately APP doesn't store that info in the FITS headers of the masters.

I'll have a look at your data now.

@mfranks_160 Hi Michael,

It looks like the flats haven't been taken correctly. When I look at a light, I see the following image plus histogram:

The histogram shows that all three color channels have their peaks well detached from the left. Red and blue have their peaks more or less at the same place and green further to the right. The image looks green but that's expected because of the double amount of green pixels w.r.t. red and blue in an OSC camera.

The flats however look like this

The red channel is not detached from the left at all. Blue is and green is nicely to the right. This means that in the flats red is underexposed.

What light source do you use for the flats? Have you been using this always? How do previous flats look? do they have similar histograms as these flats or do they have the red peak detached from the left?

I tried integrating your data using the MD and MB you provided and got a warning that image calibration couldn't be performed correctly and that many pixels clip to black. Did you get that warning too? It suggests that your lights have a problem and are underexposed, which seems odd given the 300 sec of exposure time. Did you use any filter like L-eXtreme? There is no filter data in the FITS header.

Wouter

@wvreeven Thanks for looking into this!  I'm taking flats as I have always done.  I put a white cloth around the telescope and aim for blue sky, making sure direct sunlight isn't hitting the front of the scope and using a setting of 17000 ADU.  I use the camera whenever I possibly can, it's been a lot here lately due to clear skies.  I'm using the L extreme.  My guiding has been good with PHD giving me an error of around or just under 1.  I look at the histogram for the flats on the FITS viewer and it seems that all channels are lining up.  I don't understand why it says my lights are underexposed as I'm at 200 gain with an exposure of 300 sec, but I never got that warning.  In your first response to me, you mentioned hot pixels.  Like I said before, I have been using the camera a lot.  I live in Texas, and we are currently experiencing higher temperature than normal during the day (Over 100).  I leave my all my gear outside (it's been two weeks), making sure I cover both my imaging camera and guiding camera with a white t-shirt to protect them from the sun.  Could I be "frying" camera with the heat?  Could the heat be causing hot pixels to develop, making my dark library obsolete? 

Posted by: @mfranks_160

Could I be "frying" camera with the heat?  Could the heat be causing hot pixels to develop, making my dark library obsolete? 

It certainly is a possibility. Perhaps you can try to take new darks and bias and see if that solves the issues of the hot pixels.

Posted by: @mfranks_160

I use the camera whenever I possibly can, it's been a lot here lately due to clear skies.  I'm using the L extreme. 

Did you use the L-eXtreme filter for your previous images as well?

@wvreeven yes I used the l extreme for the previous images also.  I started taking new calibration frames today.  We'll see what happens!

@wvreeven well I THINK I figured it out.  I reinstalled the camera drivers and everything seems to be working now.  I also believe I have a light leak in my train, but I can get that sorted out.  Thanks for looking into this.

Hi Michael,

Excuse me for reviving this post, which is now almost 1 year old!  It's just that, for no obvious reason that I can determine, I also suddenly have exactly these same 'cross' artifacts appear on stars as you seemed to have, on my more recent couple of imaging sessions over the last few days.

Kit last used around 5 weeks ago when all was well, and nothing has been changed, moved or adjusted in the meantime.  Clear night, temp around 12 deg. C here in UK.

It's only evident on red final image after integration - individual red subs all look fine as far as I can tell; other integrations (L,G,B) are fine, only red is affected.  Using same masters for flats, darks, dark flats and bad pixel map from library as used many times before on this same set up.

Using ASI183mm Pro on WO GT81 'scope and not seen this issue before!

In your final post you mention you ended up reinstalling your camera drivers - did that definitely cure the issue?  If so, I'll try doing the same.

Alternative speculation was maybe bad tracking/dithering on my mount (HEQ5 Pro + Rowan belt mod) given the last couple of targets (NGC4236 and NGC4244) which have shown this anomaly are both currently very high for me, so 'scope was almost vertical at times!  However, never had a problem before, and how come the other subs integrations are OK?

If you see this post, would appreciate your clarification for what resolved this for you, either way!

Many thanks

Richard

Sorry - to clarify, air temp was ~12 deg. C but camera was set to -10 deg. C ......!

Hi @richbandit &  @mfranks_160,

Michael, do you recall what exactly was the problem?

Richard, which APP version are you using? Did you try the latest version from our download page?

Mabula

Hi Mabula,

Many thanks for your response.

I'm using 1.083 .... I note I last updated it on 13th January 2022, so not sure if that was the 'stable' version or one of the subsequent updates (.2, .3 or .4)?  (downloaded file just shows 1.083).  Didn't update to a later version 2.0.0xx yet, as they all seem to still be 'beta' versions and I've tended to wait until you advise the version is 'stable' before doing so!

Do you think a v.2.0.0 version might make a difference? - but it's very odd that I've never seen this issue before, though.  Still, what the heck I'll give it a go with latest version and report back ...

Stacked image as below (slight stretch, otherwise untouched after stacking):

Thanks again,

Richard

Hi Richard @richbandit, I think i know what is happening and what Michael's problem was. The data is being demosaiced/debayered because the capture software put a Bayer pattern in the file headers probably... Can you check what happens if in 0) RAW/FITS you set the debayer algorithm to no interpolation and then process the data ? When you load the frames, you can see the CFA column, is there a bayer pattern reported?

Hi Mabula,

Thanks for your advice - but I'm using an ASI183mm Pro camera (mono) with RGB filters, so I always thought the 0) menu options are not required and/or have no effect anyway .... is that not the case, or have I completely misunderstood this function for the last 4 years!!?

Regards capture software - I'm using Voyager to acquire the subs and, as mentioned before, I've not had any problem before using this (also for the last few years).

I've downloaded latest version of APP and re-run the process with all the same parameters, and it's still exactly the same result.

HOWEVER - I think I've identified what my problem is, although at the moment I don't know what caused it in the first place!  I've now noticed that about 60-65% of my red subs have terrible stats according to APP's analytical graphs, which finally gave the game away.  Viewing each sub individually it seemed to look fine, by eye, but the graphs clearly showed up the problems.

I've now re-run the process using the ~30%-odd red subs that seem to be OK (according to the graphs), and the artifacts around the stars have now disappeared!  Yippee!

So it seems that my 'problem' was nothing more than simply bad data ..... all I have to do now is figure out why only the red subs were affected, and only ~60% of them.

Meanwhile, thank you so much Mabula for your advice and guidance anyway - perhaps you would still be so kind as to clarify my use, or need (or not?) of the 0) menu please!

Regards,

Richard

Posted by: @richbandit

↑

Hi Mabula,

Thanks for your advice - but I'm using an ASI183mm Pro camera (mono) with RGB filters, so I always thought the 0) menu options are not required and/or have no effect anyway .... is that not the case, or have I completely misunderstood this function for the last 4 years!!?

Regards capture software - I'm using Voyager to acquire the subs and, as mentioned before, I've not had any problem before using this (also for the last few years).

I've downloaded latest version of APP and re-run the process with all the same parameters, and it's still exactly the same result.

HOWEVER - I think I've identified what my problem is, although at the moment I don't know what caused it in the first place!  I've now noticed that about 60-65% of my red subs have terrible stats according to APP's analytical graphs, which finally gave the game away.  Viewing each sub individually it seemed to look fine, by eye, but the graphs clearly showed up the problems.

I've now re-run the process using the ~30%-odd red subs that seem to be OK (according to the graphs), and the artifacts around the stars have now disappeared!  Yippee!

So it seems that my 'problem' was nothing more than simply bad data ..... all I have to do now is figure out why only the red subs were affected, and only ~60% of them.

Meanwhile, thank you so much Mabula for your advice and guidance anyway - perhaps you would still be so kind as to clarify my use, or need (or not?) of the 0) menu please!

Regards,

Richard

Dear Richard @richbandit,

Okay, that sounds better 🙂 There must be a reason why those particular frames are that bad, do you have any clue at the moment?

If you use a monochrome camera, then yes, the menu 0) Raw/FITS can be ignored, except when the capture software is misconfigured. If you tell the capture software that is a color camera with a bayer pattern, then you need to use 0)RAW/FITS with the no itnerpolation demosaic algorithm to treat is as monochrome data. The best solution of course is to properly configure the capture software in that case 😉 and tell it is monochrome data.

Mabula

Hi Mabula,

Thanks for the clarification regards 0) menu .....

I could understand if all the red subs had been affected for some reason (faulty/dirty/misaligned filter, etc.) but why just some of them I wonder?

I started the sequence off with the red filter first, and I note the 'good' ones were the first 17 acquired, after which the issue appeared.  The odd thing is that when the sequence subsequently finished the reds and moved on to the green, blue and finally lum filters they all seem to have come out just fine.

Looks like we may have a few clear nights here, so I'll do some experiments/checking and report back here if/when I find something.  You never know, it may help someone else who finds this odd problem.

Best regards,

Richard

Hi Mabula,

I haven't had chance for another imaging session yet, but having checked again I'm unable to find anything obviously wrong with my equipment set up.  However, I have a theory as to what may have caused my 'issue' - to some extent unproven, but the evidence suggests it's a strong candidate.

Some 'Googling' has revealed that the Jet Stream was slap bang over the UK on the night(s) I experienced this problem, in particular it seems it was passing right on top of my location!  So could it simply be the case that, by chance, its (high level) wind speed during my affected period (about 1 hour between 22:30 and 23:30) picked up enough to disrupt the seeing and thereby affect my subs?  All other subs before, and after, that period were fine.

I wonder if anyone else had similar issues around then?

In the absence of any other information this sounds plausible to me - I'll plan to conduct exactly the same imaging sequence, mount position, start time, etc. again on a night when the Jet Stream is definitely not in evidence and see what transpires ....

Regards,

Richard

Hi Richard @richbandit,

This issue is weird for sure. I would not think that the Jet Stream would cause a problem like this. Please keep us updated if it happens again though !

Mabula