Senor Analysis

[Dave_S]

When carrying out a Sensor Analysis within Sharpcap.

Should I set the camera gain, and black level setting to 100 and 0 respectively, then adjust the intensity of the light source to give the recommended Histogram level for an exposure of 0.5 to 2 seconds.

Or, should I set the gain and black level setting to that which I noramly use with my Atlair 294c Pro Tec. In my case 900 and 30 respectively.

Thanks in advance

Dave

[admin]

Hi Dave,

Assuming that you are using the direct Altair camera support in SharpCap (rather than the ASCOM driver) then no need to worry - SharpCap will adjust both the gain and the black level automatically during the analysis process as it needs to understand how changes to those values affect the brightness of the image.

At the start of the analysis, SharpCap will change to minimum gain for the step that you use to get the light level right - the exposure length that you get during that stage will (roughly speaking) be the longest exposure length needed during the analysis, since when higher gains are set late, shorter exposures can be used. Normally you want that exposure length to be somewhere in the 0.5s to 1.0s range.

cheers,

Robin

[Dave_S]

Hi Robin, intelligent software

I have another question, if I may

The analysis of the the 294c gives a e/ADU of 1.68 at gain 100, and an e/ADU of 0.95 at gain 177.

If may maths is still working (I have been retired for 19 years ), by interpolation I calculate that unity gain (1.0) is at gain setting 172.

However, the read noise v gain graph shows a distinct drop off in read noise at around a gain setting of 900, which is what I have been using, and the read noise is pretty flat from there on up.

The question is, do I go for a unity gain setting of 1.0 e/ADU or continue using a gain setting of 900.

Looking at the graph, it would appear that I could push the gain quite high without increasing the read noise, but I guess there will be a point at which "the law of diminishing returns" kick in.

I imaged for 12 years (2002-2014) with a variety of SX CCD cameras, both OSC and mono.

Now having returned to astro imaging 8 years on, I arrive at a time when CCD is well on its way out , out and enter into the world of CMOS, with all its quirks, e.g. green tinted images, gain and black level settings etc to contend with.

The green tint I have dealt with, and your sensor analysis application is an excellent 'tool' for seeing what is going, and hopefully getting the best from this camera.

[admin]

Hi Dave,

my opinion is that unity gain is a bit of an 'urban myth' in the astroimaging community - it frequently gets recommended, but I can't see any justification in the maths or science of imaging for preferring it to other gain values.

If you like understanding the reasons behind choosing one gain over another (and how to choose the right exposure length), you can read all about it here : viewtopic.php?f=35&t=456

If you prefer a presentation, you can see a recording of a talk I gave on this subject here : https://www.youtube.com/watch?v=3RH93UvP358 and here : https://www.youtube.com/watch?v=ub1HjvlCJ5Y

cheers,

Robin

[Dave_S]

I think you are right about the 'urban myths' Robin, and there are plenty of them, probably started by so called "self appointed experts".

Having spent a 40 year career in Telecommunications, with the last 20 years in R&D, my inquisitive mind does like to understand what I am dealing with, so I will visit those links that you suggest.

Although long since retired, I like to think that "the old dog can learn new tricks".

Dave

[AngusMcDuff]

Can I leave the camera connected to the telescope or should it be removed for the sensor analysis?

I use the ZWO drivers. Is there any camera setting that I need to make for the analysis? I set the temperature to -10C because I use that temperature for imagining.

[admin]

Hi,

on the telescope is fine if you can get the right illumination level while still on the scope. It may be easier to tweak the illumination with the camera removed from the telescope (either will work fine - if the light levels are right).

I would avoid using flip, overclock or 'high speed mode' during analysis, and set the colour balance to neutral (50/50). Finally, don't use darks or flats. The temperature doesn't really affect the results since all the testing is done with short exposures. Temperature affects the thermal noise for longer exposures, but that isn't part of the tests.

cheers,

Robin