Do flats need to be taken at same gain as lights?

@Freestar8n has given an explanation in a post in another thread:
https://www.astrobin.com/forum/post/62215/

Therefore, I would take them at the same settings. Just to be sure to have the same sensor conditions 

Björn

Alfonso Caschili:
As long as the flats are properly calibrated it should not matter.

This is where I am coming from.  In fact, by lowering the gain I am more likely to be in the linear operating range of the sensor.  Also, by doing so I am less likely to have issues with short exposure times and "strobing" if I can call it that of a cheap LED panel.

As long as you calibrate your flats consistently the gain you use isn't an issue. In fact it wouldn't matter whether the sensor is in linear or non-linear response mode.

Björn Arnold:
@Freestar8n has given an explanation in a post in another thread:
https://www.astrobin.com/forum/post/62215/

Therefore, I would take them at the same settings. Just to be sure to have the same sensor conditions 

Björn

The underlying question is a good one.  Just to review, there are three important things that flats correct:

1) Vignetting, which includes both light partially blocked by internal components in the optical train as well as dust motes.
2) Cos^4 radiometric light fall off, which is due to how irradiance decreases with field angle in any optical system.
3) Fixed pattern noise, which is due to PRNU or "photo response non-uniform.  This is simply small variations in responsivity between pixels and it's a property of the sensor itself.

When we calibrate an image, the basic idea is to first subtract the dark signal and then divide that result by the master flat.  Of course nothing is ever as simple as it seems so we have to consider how darks and bias offsets figure into the operation.  When we take the flat data that will make up the master flat file, the data will contain both dark current and the bias offset.  Here's the problem.  In order to correctly divide the flat data it must be first normalized to have values between the values of [0,1] and that means that we must FIRST remove the dark current+bias offset before doing the normalization.  Here it's important to recognize that the normalization process is exactly the same as adjusting the gain on the sensor, which means that the sensor gain doesn't matter when you take flats!  What does matter is that you take any flat-dark and/or any bias data at the same camera gain and temperature.  That's important so that you are correctly subtracting the correct amount of dark signal and/or bias offset before scaling the flat data.  Obviously if you are using a camera with negligible bias offset and exposures short enough that dark current is also negligible, you may not need to worry very much about removing that stuff from the Flat Master.

As a matter of good standard operating practice, it's always a good idea to use all the same settings on your camera when you take calibration data; but, in the case of flats, you can get often get away with using a different gain settings if you have too.   Again, that depends a little on your camera, the exposure time that you use, and whether you have or need flat-darks and bias.

John

Andy Wray:

John Hayes:
What does matter is that you take any flat-dark and/or any bias data at the same camera gain and temperature.

Just for clarity for any newbies on here ... you are saying that the flat dark or bias needs to be at the same gain and offset as the flat itself; NOT the lights.  That, at least, is what I am doing.

Yes.  That is exactly what I am saying.  Thanks for making it clear.

John

Apologies for rephrasing a question I asked in another thread but didn't get answered there:
How susceptible are flats to slight focus changes? If you refocus during a session and you have slight focus shifts for the same filter, do you have to take flats for each setting of the focuser or can you neglect minor focus shifts by a few focus steps?

Currently I take flats for each session and it seems ok, but I just recently acquired a focuser and don't have much experience. The focus shifts within one session so far seem to be within 30 steps of a ZWO EAF. Focus inbetween sessions varies much more, I assume due to atmospheric conditions.

Clear skies
Wolfgang

Wolfgang - these small changes are irrelevant. I’ve very often taken flats indoors and lights outdoors with significant temperature and hence focus position differences and they have always worked without issue. What is more important with my ZWO filter wheel is to make sure it is calibrated before every session and set to  unidirectional motion. Repeatable positioning of filters is much more important than slight changes in focus position.

What is more important with my ZWO filter wheel is to make sure it is calibrated before every session and set to  unidirectional motion.

I would agree on the unidirectional setting, particularly if you are using small filters compared to your sensor (1.25" in my case with a 4/3rds sensor).