Parameters for the ASI2600MC Pro ZWO ASI2600MC Pro · Daniel Arenas · ... · 58 · 5040 · 19

Hi everyone,

Dark flats are most of the time useless. 
I suggest to read the paper that Siril staff wrote on this 

(Sorry this is the french page, I guess the english translation is somewhere in their website)

https://siril.org/fr/2021/12/stop-%C3%A0-la-mode-des-dark-flats/


In a nutshell, they only recommend darkflats for very long flats (over 10s) or for cameras with very strong ampglow.

So most of the time, a  masterbias is suffiscient enough and you save a bit of processing time if you had produced one with your dark librairy.


Edit: Here is the english translation :
https://siril.org/2021/12/enough-with-dark-flats/

I would be very careful about not using dark flats--note in thatarticle they do say be careful, you should verify yourself that it is okay.

But lets look at data. Here is my data for the ASI2600MC Pro, I cool to 0C to limit dark noise. This plot shows single exposures in the dark, the shortest exposure is the bias, the others would be called darks. All but the one marked continuous are one-off single exposures. What I am plotting is the average signal value over the bias/dark, so this is the bias signal plus any thermal contribution. All were set to 50 bias, which as you can see gives about 500 adu--this camera gives 10X adu for whatever bias is set.  You can see the shortest exposure, the bias is a bit odd, sometime higher value than somewhat longer times. So if you use the bias as a dark flat it will be somewhat wrong unless it is the same exposure as the bias. Your flats will not correct to give you a nice flat field with this camera using a bias as your flat dark. Further, anything longer than 0.1 s increases the average value of the background, so anything over 0.1 s with this camera must use a dark with the same exposure--and as you can see it gets more sensitive with longer exposures. For this reason with this camera I would not recommend more than 1 second exposure, it is just too sensitive. And it is worse than that, taking images continuosly, as you would to average your dark/bias or your lights increases the average value. I believe this is due to sensor heating, despite cooling to 0 degrees, the instantaneous temperature of the sensor is higher than if you took a single exposure. So take your dark flats at the same exposure and with the same wait time between frames as your flats--if not your flats will be poor. This is my experience, match the dark flats and flats for good flats, it doesn't work well otherwise with this camera.

This is a very low noise camera, I would suspect other modern CMOS cameras will be similar, and indeed, others have reported the bias signal is often odd compared to any other exposure. Don't use a bias for a dark flat or a light unless you have proven it is okay.

Note, this is just looking at one aspect, the average level, there may be many other reasons not to use a bias.
Rick

Stuart Taylor:

Daniel Arenas:

Stuart Taylor:
Set to offset 0 the camera reports a minimum pixel value of 0
So I then shot 20 bias frames at gradually increasing offset settings and integrated them using a minimum value method and measured their lowest pixel value.

What I found is that an offset of 5 produces a min pixel value of 0.0006 (which is >0) so it's actually not necessary to go to any higher offset.

Oh! Excellent @Stuart Taylor !

That's the test I wanted to do next days. But maybe I can use your value to see if, for me, it's OK too. I say that because I've read that it doesn't matter if the model and make of a camera is the same, it depends on the sensor and there aren't two sensors identical. 

PD: I'm not any kind of expert, I just reproduce what I've read in forums, or I've seen in videos.

I was surprised that only an offset of 5 makes a measurable difference to the min pixel value. So I have been using far too high a value for quite a long time. For me it only needs to be 5

Well, I did my test.

I shot some darks, 300s, -10ºC, gain 100 (these are my usual parameters) and try to change the offset from the camera from 5 to 5.

Offset 0:




Offset 5:



Min is still in 0. Let's try with offset 10:



Let's enlarge this graph with the magnifying glass icon (same offset):



Offset 15:





The minimum is shown in 1 but regarding the graph it seems that it's still clipped. So, let's ensure and use offset 20.

Offset 20:






In that case and with a minimum of 53 I think it's the correct one. As a curiosity, I thought that the minimum in bias would be 53 also when setting offset in value of 20, but not. Here you are the graph of a bias with the most short exposure time 3,2E-05 s.





But min 26 is Ok so, offset 20 is the best value for me. Of course, I could try different offset values from 15 to 20 and maybe 17 o 18 would be enough not to clipping, but I preferred to have round figures and a little bit more margin of security from the minimum of 0 not to clip the histogram in Bias frames.

Another question, please, when I set the temp in -10ºC (for example) I noticed that the camera is not able to maintain it invariant. I mean, sometimes the temp is -10, others -9,80 or -9,60 then raise a bit again in -9,80... Is it normal or maybe is because it's summer here in the north hemisphere and the ambient temperature was about 27ºC, so it was too high for the camera to achieve the goal of -10ºC and maintain it (meaning that in winter will be easier). I noticed that the camera was working on the 100% so that was the clue for me to know that the ambient temperature was too high.

Kind regards and clear skies.

Daniel.

Hello perhaps you could have a look on this topic
It's in french but with all the translation tools it will be easy.
Some user of the forum have made a hard and complete job (i think) about the CMOS camera.

On of the great french user is writing about his cam, and we are lucky he use a ASI2600MC
Some other user explain other things in a very good quality

I'm a noob and i really appreciate the job they made.
They have write it in a way that a noob like me can understand.
So i think that it could be usefull for a lot off 2600MC users

Have a good day

Where I live (Eastern Canada) I can usually get away with -10C year-round for my lights and darks. Because I bring my optical train inside after imaging and shoot my flats with a light panel, I shoot flats and flat darks at 0C. I usually use gain 100 offset 50 unless I need a high dynamic range or if I’m just shooting RGB stars for a NB image in which case I use gain 0 offset 30.

hi

 the easy way to know you offset 

open asi air pro plus   
shot a one image and save it 
then open it in pixlnsight  then click  FILE , FITS header 
and there is the  default offset value 
 for ZWO    50 is the  default value for 2600, 2400, 6200 cameras 


CS, BRIAN