Removing Amp Glow ZWO 178 colour cooled

[Ricker]

I am imaging with a ZWO 178mc-cooled and am having trouble removing amp glow.  I am using darks in the same conditions as lights and stacking and imaging with SharpCap.  I am left with a image looking like below.  No matter what I do I seem to be always left with this residual amp glow signature,  Any ideas out there?

 

Rich

 

[Craney]

Take flats ??

https://astrobackyard.com/how-to-take-flat-frames/

[Ricker]

Thanks for the tip but Amp glow is part of the dark signal.  The flat is a measure of sensitivity variation across the sensor.  So darks should be enough to remove it,.

[geordie85]

I have to calibration my images with darks, flats and darkflats to fully remove the amp glow on my qhy183c. 

[wimvb]

What camera settings do you use? If you have the possibility to adjust camera offset/bias, take a look at that. With high gain, you need a larger offset to avoid clipping. What you see could be clipping in the red and blue channels, but not green.

Just a wild idea.

[Ricker]

Great suggestion. I shall give it a go tonight. Ampglow is a frustration. 

[maxchess]

I have the uncooled version of this  camera and the same problem. 

 I tried Gain=200 and got what looked like a star shaped light leak in one corner just like yours. I also took flats & Bias frames, but they did not help much.  I assumed it was a problem with my scope, but found its was AMPGLOW, see this ZWO article.  

I tried taking 90 second Darks at various levels of Gain and found it diminished until it was virtually invisible at Gain=10 .  To see the effect take a Dark into your processing program and stretch it.

ZWOs solution is to control via Darks and disable any dark optimization feature of your calibration program eg DSS.  They say that this effect only occurs in "specialist" sensors and not in those used for DLSRs.  I was staggered, sounds like a design fault to me.  

These guidelines appeared in another ZWO post:

Unity Gain ==>    Gain: 180   Offset: 25
Highest Dynamic Range ==>   Gain:  0    Offset:  25
Lowest Read Noise ==>   Gain: 270   Offset:  340
 

So my interpretation is to go for Gain: 0   Offset: 25 for DSOs on long exposure and disable dark optimization. Waiting for moonless/cloudless night to try this out. Any other experiences most welcome.  Never had this problem with my trusty 450D.

 

 

[vlaiv]

2 hours ago, maxchess said:

I have the uncooled version of this  camera and the same problem. 

 I tried Gain=200 and got what looked like a star shaped light leak in one corner just like yours. I also took flats & Bias frames, but they did not help much.  I assumed it was a problem with my scope, but found its was AMPGLOW, see this ZWO article.  

I tried taking 90 second Darks at various levels of Gain and found it diminished until it was virtually invisible at Gain=10 .  To see the effect take a Dark into your processing program and stretch it.

ZWOs solution is to control via Darks and disable any dark optimization feature of your calibration program eg DSS.  They say that this effect only occurs in "specialist" sensors and not in those used for DLSRs.  I was staggered, sounds like a design fault to me.  

These guidelines appeared in another ZWO post:

Unity Gain ==>    Gain: 180   Offset: 25
Highest Dynamic Range ==>   Gain:  0    Offset:  25
Lowest Read Noise ==>   Gain: 270   Offset:  340
 

So my interpretation is to go for Gain: 0   Offset: 25 for DSOs on long exposure and disable dark optimization. Waiting for moonless/cloudless night to try this out. Any other experiences most welcome.  Never had this problem with my trusty 450D.

 

 

Are you sure these are correct values?

For 178 model, there is no unity gain setting, since all gain values are below unity - full well of this camera is 15K and ADC is 14bit - meaning 0-16383, so at gain 0 you are still not at unity - a bit below, something like ~0.93e/ADU.

[maxchess]

To be safe I am trying some captures at Gain=10 , Offset=10 on M81, (20 x 120 secs) but with a full moon they are likely to be poor. M81 is about as far away as I can get from the Moon.  However lights so far are showing minimal Amp Glow. Hopefully with Darks, Lights  & Bias I can eliminate it.

[vlaiv]

10 minutes ago, maxchess said:

To be safe I am trying some captures at Gain=10 , Offset=10 on M81, (20 x 120 secs) but with a full moon they are likely to be poor. M81 is about as far away as I can get from the Moon.  However lights so far are showing minimal Amp Glow. Hopefully with Darks, Lights  & Bias I can eliminate it.

No need for bias. Most CMOS sensors have one issue or another with bias. You will need lights (obviously ), matching darks, flats and darks that match flats.

[maxchess]

 

13 hours ago, vlaiv said:

Are you sure these are correct values?

For 178 model, there is no unity gain setting, since all gain values are below unity - full well of this camera is 15K and ADC is 14bit - meaning 0-16383, so at gain 0 you are still not at unity - a bit below, something like ~0.93e/ADU.

This is the article that gives the example settings for long exposures.  It is a different sensor, but the principal seems to be that for Highest Dynamic Range on long exposures you set gain to 0, offset 10.  In the example they have unitary gain as (G: 139 O: 21). So I assume that G=0 maximizes well depth and offset 10 lifts the lower values.

In my simplistic way of thinking I assume that with G=0, O=10  I can usefully increase exposures until the highlights reach saturation, at which point I am getting the best dynamic range, ie making optimum use of the ADC to record the widest range of values in the most detail.

I notice from Astrobin that many ZWO camera users use large numbers of relatively short exposures 60-120 secs , rather than DLSR users who go for fewer much longer exposures.

Still struggling with this camera, will process last nights Moon affected images later. 

 

 

[vlaiv]

2 hours ago, maxchess said:

 

This is the article that gives the example settings for long exposures.  It is a different sensor, but the principal seems to be that for Highest Dynamic Range on long exposures you set gain to 0, offset 10.  In the example they have unitary gain as (G: 139 O: 21). So I assume that G=0 maximizes well depth and offset 10 lifts the lower values.

In my simplistic way of thinking I assume that with G=0, O=10  I can usefully increase exposures until the highlights reach saturation, at which point I am getting the best dynamic range, ie making optimum use of the ADC to record the widest range of values in the most detail.

I notice from Astrobin that many ZWO camera users use large numbers of relatively short exposures 60-120 secs , rather than DLSR users who go for fewer much longer exposures.

Still struggling with this camera, will process last nights Moon affected images later. 

 

 

With 178, use gain 0 if you want highest dynamic range, and that is ok (no quantization clipping will happen because even at gain 0 you are below unity for e/ADU vallue), but I would not use offset setting of other camera. There is really a simple way to test which offset value is good for your particular gain setting.

Just take your camera, set all other values (temperature, gain) and take few bias frames (or short darks, 1s or whatever). Examine histogram / pixel values of those bias frames - none of them should include lowest value possible. 178 is 14bit camera, so I would say you should not have pixels with value of 8 or less in your images. If you find such pixels - increase offset, repeat. Stop when no low values are found. Simple as that.

 

[maxchess]

OK, sounds good. Moon washed out last nights images so waiting for Moon to disappear before experimenting further.