What could the reason be for this extreme noise in one night that isn't there in the second? - Forums

Topic starter

Oct 10, 2023

This is the first time I wanted to do a project over several nights and realized that after stacking the image got noisier when stacking both nights than just stacking one. So I stacked each night seperatly and found this huuge difference in the data of the first night. what could the reason behind that be?
It was taken on my balcony, same setup same everything, just out of some reason much worse.

My gear: modified canon 600D
Skywatcher Staradventurer GTi
Zenithstar 61
L-Enhance filter
Asi Air

First night: just with auto stretch in siril

second night: just with auto stretch in siril

I did bump up the dither settings from 5 pixels to 10 pixels but I don't think that makes that big of a difference.

HR_Maurer

IOTD Staff

Oct 10, 2023 · Edited: Oct 10, 2023

Hi,
did you use the same calibration files on those two images? Or - are the darks or flats different?
Maybe you used the same one and implied that the difference must come from the light frames.

One of the first things that would come to my mind is temperature. I dont know the thermal behavior of the 600D, but at my 100D it was really awful. In a hot summer night the dark noise was huge.

One image seems to have much more banding, too. What about the ISO settings?

When it somes to signal/noise, noise is only one part of the equation. Can you pick a star that is not saturated, and measure its integral (sum of pixel values) after removing the background? Is this same star integral comparable in both images?

CS Horst

Helpful Insightful Respectful Engaging

Blaine Gibby

Oct 10, 2023

Could be many different things. Different temperature from one night to the next, could be a thin layer of clouds that reduces your signal to noise ratio or could be different seeing conditions. Could be any one or combination of the above, just to name a few things

Oscar

Oct 10, 2023 · Edited: Oct 11, 2023

Calibration frames might be the culprit

EDIT:
to those 19 people who saw my original mistake in the post: for some reason I confused dithering with drizzling, I deleted that part of the post.

Sean Mc

Oct 10, 2023

Different sides of the meridian? Different altitude? Shooting toward a city one night and away the other?

Jens

Topic starter

Oct 11, 2023

So I shot the target in two consecutive nights, temperature was pretty much exactly the same, the target was at exactly the same spot as the night before (i just turned the camera for a different target later in the night) and used individual calibration frames for each night.

I found out just by stacking the light frames I get pretty much the same noise pattern in the first night.

My guess is, that the dither settings on the first night of 5 pixels wasn't sufficient enough. Probably that increase up to 10 pixels changed things. But that's just a guess.

Wei-Hao Wang

Oct 11, 2023 · Edited: Oct 11, 2023

Since you said each night has its own calibration frames, you may take a look at the master darks and master biases from the two nights to see if there is any difference. You can also try to stretch the uncalibrated raw lights from the two nights to see if there is really a huge difference. Then to stretch the unstacked single calibrated lights from the two nights to see if there is a huge difference. These tests can tell you exactly what or which step went wrong: the calibration frames? the calibration process? or the stacking process (which is related to whether you dithered enough)?

Helpful

Ali Alhawas

IOTD Staff

Oct 11, 2023

Hi,

For 1st guess.. temperature.
2nd guess.. calibration issue.
You never can tell all factors are the same even in 2 consecutive nights, Sometimes, Factors are changed in the same night.
When we talking about temperature, We meant the sensor temp. not feeling temp, And that can't be known without measurements.

CS!

Jens

Topic starter

Oct 11, 2023

Thing is that I tried to stack just the light frames to see what would happen and the same noise pattern persited. So the calibration frames can't be the issue. do you think that the temperature can have such a huge effect on the noise? it also doesn't really look like walking noise.

The asi air acutally gives a Sensor Temperature readout, it tells me it's about 24c° which seems to be normal for this camera in this usecase.
I'm guessing it gets that data from the overheating sensors in the camera.

For me it's just important to get behind it in order know what to do and not to do. this is basically a whole night of data that is unusable.

Wei-Hao Wang

Oct 11, 2023

The biggest difference between the two is the horizontal banding. Mismatched dark temperature and insufficient dithering usually lead to walking noise, but not horizontal banding that goes across the entire frame. If your camera mysteriously produced different strengths of banding in the two nights, you probably can see such a difference in the bias and dark frames as well. (The opposite is not necessarily true, unfortunately.) That's why I suggested you to look at the calibration frames. This does not necessarily mean the calibration process went wrong. You really need to check everything you can in order to find a clue.

Well Written Helpful Insightful Concise Supportive

Stefan Böckler

Oct 11, 2023

When I imaged with an EOS1100Da I found that excessive dithering, i.e. every second image with huge steps, cured noise in general and specifically banding (though this has only been an issue in hot summer nights).

I guess the individual subs from the first night show the banding, right? Is this also true for the indivdual subs of the second night? If yes, I think stacking got rid of it due to more dithering. If no and the individual subs of both nights differ in terms of banding, it's probably not the dithering.

Regards
Stefan

Helpful Insightful Respectful Engaging

dkamen

Oct 11, 2023

You should also check the subs one by one one by one. If the SD card (or the camera's writer) is problematic, one or two may get mangled.

Here is what my D7500 does to roughly 1% of the subs:

If one or two such subs manage to sneak into the stack, I end up with artifacts much like yours.

PS in my case the preview JPEGs (which is what the camera display and what most photo management programs use) are written correctly, it is only the raw data that shows the problem. IOW image thumbnail looks good in the file manager but when you open it in Darktable or APP it shows the artifacts.

Helpful Insightful

Kristof Vandebeek

Oct 11, 2023

A few questions:

- what is the average ADU of both individual uncalibrated subs (so for a good and a bad sub)
- what is the average ADU value of a bias or a dark frame that you use?
- how long did you expose (and was it the same for both subs)?

I see a similar pattern with my ASI533MM when I underexpose.

Patrick

Oct 11, 2023

Looks like the sensor temp on the first stack was higher, but maybe something with dark calibration went wrong too.
I would start with the following in siril:
Calibrate all your files only with master bias and master flat. Then open the frame list of the calibrated sequence.
Set the stretch preview to histogram.
While in the frame list you can just press arrow down to quickly go through your frames. Check for any significant change in contrast at dim parts of the nebula.
If you find any significant differences i think its either increased sensor heat or increased light pollution (maybe from a neighbour?)

If there is significant differences between the days, do a new stack of day 1 and day 2 with the calibrated frames without darks and do the histogram stretch. The shape of gradients will tell if its lightpollution or sensor heat.

Helpful