PixInsight Forum (historical)

PixInsight => General => Topic started by: UlteriorModem on 2018 July 25 07:35:05

Title: Subframe selector script with a cmos camera?
Post by: UlteriorModem on 2018 July 25 07:35:05
When setting the system information for the subframe selector script, in the field "Camera gain x.xx electrons per data number".

Is this literally the camera gain setting used eg; 139.00 (for unity gain)? Or am I misunderstanding this?
Title: Re: Subframe selector script with a cmos camera?
Post by: pfile on 2018 July 25 09:16:53
if 139 is unity gain then the camera gain setting here is 1. the units are electrons per DN. 139.00 would differ from camera to camera and perhaps even driver to driver for the same camera.

rob
Title: Re: Subframe selector script with a cmos camera?
Post by: UlteriorModem on 2018 July 25 10:01:31
Ah okay I think I understand.

In looking at the product manufactuer's curves it appears that this number actually goes down until at a gain of 300 it would be zero.

The camera is an ASI1600mc pro.

The curves can be found on page 8 of this document.

https://astronomy-imaging-camera.com/manuals/ASI1600%20Manual%20EN.pdf

Thanks for the help!
Title: Re: Subframe selector script with a cmos camera?
Post by: bulrichl on 2018 July 25 10:13:20
Yes, you have to differentiate between the gain setting of the driver and gain in electrons/ADU. Gain [e-/ADU] is the conversion factor, defined as:

gain = number of electrons per pixel / number of counts per pixel

One has to consider that with most (all?) CMOS cameras the driver is scaling the counts, e.g. for a camera with 12-bit A/D converter: from 12 bit to 16 bit, i.e. the driver is multiplying by factor 2^4 = 16.

With ZWO's CMOS cameras, the gain setting is in units of 0.1 dB. With other brands this scale might be different. The plot in the camera manual shows the gain [e-/ADU] vs. gain setting [0.1 dB] of the driver.

When you input the camera gain in Subframe Selector, it can display Median and MeanDeviation alternatively in ADUs or in electrons. However, you will not need that in order to select the good frames as these representations differ only by a multiplicative factor, i.e. the gain.

Bernd