CMOS Camera pre-processing

[Rodger King]

Just purchased my second dedicated camera for imaging.  My first was the ATIK 314L+ CCD camera and I have had some good results with that.  I never bothered with darks and never really got any good flats so most of my images were as they came off the camera. 

My newest camera is the ATIK Horizon II CMOS camera.  This is different in many ways and I want to get the best start using it by doing proper darks and flats at the beginning.  However I am struggling to get to grips with both theory and practice.

LIGHTS

My reading says that the optimum sub-exposure for my home site with a Bortle rating of 5 and camera which is rated to have a read noise of ~ 3.5 electrons at gain 1 is about 40 seconds.  But as I up the gain from a setting of 1 the read noise should decrease,  at gain setting 2 it shoulld be ~ 2 electrons and at gain 5 ~ 1.2 electrons.  This then changes the optimum sub-exposure length to ~ 20 seconds at gain setting 2 and just 10 seconds at gain setting 5.   

Is this correct ???

DARKS

I am not sure what gain setting I will settle on so I am trying to create darks for gain setting 1,2 and 5 setting.  However, the theory of darks says I must do dark exposures for the same time as I do my lights.  So I do my darks at gain setting 1 for 40 seconds, for 20 seconds at gain setting 2 and just 10 seconds for gain setting 5.    However when I tests the noise of these exposures, the ones at gain 1 and 40 seconds are less noisy than the ones at gain 2 and 20 seconds which are less noisy than the ones at gain 5 and 10 seconds.  Which is the opposite to what I was expecting ??? 

Is this right ??? 

According to my results gain1 wins out for both lowest noise and highest well depth.  Why would I choose any other gain ?

Also how should I create my master dark image Do I median the individual exposures or average them ?

FLATS

Trying to create flats inside my observatory using a teashirt over the telescope and pointing at a light source (my Ipad on low intensity with an even illumination. 

Any better ideas ?

Also how should I create my master flat image Do I median the individual exposures or average them ?

[George Sallit]

Rodger,

Basically you are correct.. CMOS cameras allow you to change the gain of the amplifier used in the camera. As you increase the gain you will increase the signal you get in the image but you will also increase the noise. So setting the gain is a balance between getting a good signal and minimising noise. For the Panasonic chip used in your camera I have found a Gain of 2 works well for LRGB images. Because narrow band filters darken the background you can use a higher Gain for narrow band filters.

Having assumed that then you need to optimise the exposure length. The aim here is to expose the image long enough that the image signal overwhelms the camera noise. For a gain of 2, with your sky conditions, you have found that an exposure of 20 secs is optimum. The camera adds 2 electrons of noise from the readout of the image so the exposure needs to be long enough to overwhelm that noise. However, I suspect the sky noise from the light pollution in your skies will dominate. Although 20 secs would be optimum, exposing for 40 secs would do no harm so I would use that. Such short exposures may allow you to do unguided exposures.

For darks, a set of  40 sec dark frames at a gain of 2 would be ideal. Unlike CCD cameras CMOS need darks of the same exposures as the lights but at 40 secs that should not be a chore. You observations on the dark frames are confusing because you have changed the gain and that increases the noise.

The reason Gain 1 is not used so much is that it can suffer from digitisation noise and to avoid hat people use a gain of 2. This gives a bonus in that the readout noise is reduced.

For dark frame combination I prefer median as it will reduce cosmic rays but averaging will work.

The teashirt over the i-pad works but may not give a totally uniform illumination across the whole screen. You could try sky flats where you put the teeshirt over the telescope objective and ‘look’ at an illuminated white screen or the sky. Or you could buy a flat frame accessory. Try the ipad first. 

Hopefully that helps but the important things to do is try out these settings and then change them to see what impact they have. If you want to calculate the exact, ideal settings for your personal conditions then Sharpcap can measure them for you.The software costs £10 a year and it is a great help in using these cameras although there are many other packages out there and a few are free.     

[Rodger King]

Thanks George.   I actually have a good and stable imaging system with a pier mounted EQ6-R with autoguiding inside an observatory and I have built,  so long exposures isnt a problem.  With my ATIK 314L+ CCD camera I was doing 5 minute exposures without issue.  Whats the problem doing much longer exposures than 40 seconds with the CMOS camera ?

[George Sallit]

No problems but you may find that all the stars will be white and show no colour as they will be saturated. I use 300-sec exposures and just live with all the stars being white, unless doing star clusters.

[Tim Haymes]

When using the ATIK, what does gain 1,2  and 5 represent in terms of the whole scale ? – in other words, what is the maximum gain setting in the software ?    I have Sharpcap, so i would like to investigate sensor characteristics  more fully with my QHY174.
Also, does temperature effect CMOS characteristsic ?  Im thinking it should.

[Rodger King]

The gain goes upto 30 maximum.  But at that setting the well depth is only 512 so not worth going that far.

[Rodger King]

Could not wait to get my darks and flats ready so went out and took my first image.  Just 15×40 second subs using an Ha filter.  Really pleased with result.  Cant wait to do more with this new camera.

[Tim Haymes]

That is a very nice image Ryking.   Reading around the subject i found this presentation by Robin Glover enlightening.
https://www.youtube.com/watch?v=3RH93UvP358

[Author]

Posts