Dark Pixels in my Atik 314L+ OSC

[gustavo_sanchez]

Hello to all,

Maybe you can help me this issue. I am currently getting some strange dark pixels in my exposures, like an "bug swarm" pattern. They only appear on the first few exposures I take, and disappear later. Last night, I was taking 15-minute exposures @ 0 degrees C and this artifact was present on the first three frames but disappeared by the fourth one.

Any idea what could this be? The attached screenshot shows a 200% zoom region of one of my uncalibrated lights with those black pixels.

I should note that they are difficult to remove during processing, and since my guiding is fairly good, they remain after stacking.

Please help!

[johnrt]

What software are you using to calibrate and stack your frames? Are you using darks & bias? Any particular reason you only cool the chip down to 0?

[gustavo_sanchez]

This issue appears in the lights. I use Artemis Capture to take the images. I cool down only to zero because I live in the tropics and the ambient temperature at night was about 25C.

Sent from my Nokia Lumia 920

[ollypenrice]

I'm sure I've seen that before myself but I'm sorry to say that I can't remember what was causing it. I wonder if it was humidity? Could you dry the camera very thoroughly before using it? Physiotherapy gel packs from a freezer can also bring the camera's temperature down as the cooler starts to work. Just apply them to the outside. I bet this has something to do with your tropical conditions.

That kind of thing can often be fixed in Photoshop by using the colour select tool on a good range of the affected pixels and then applying the Median Filter.

Dithering between subs would also be a big help.

Olly

[Zakalwe]

It looks like ice forming on the sensor to me. The reason why it can shift is that the sensor will be warmer than the cold finger that's attached to the Peltier. The ice will eventually sublimate off the sensor onto the cold finger.

[gustavo_sanchez]

I will try stabilizing the camera much earlier to see if I can eliminate this. Regarding ice, I really doubt it; the temperatures are too hot over here. During the night, the most the camera can cool is almost zero, but not below. How can it be that the ice is present at the beginning, but not later? The camera is stored inside my house before using.

Sent from my Nokia Lumia 920

[Oddsocks]

Might be trapped electrons in the pixel wells.

Try flooding the sensor with white light by shining a torch into the OTA and capturing a few frames before you begin your imaging run.

Trapped electrons can occur for many reasons including high storage temperature, rapid temperature changes, strong static electricity fields and whether the camera is stored with a cap to keep the sensor dark until use or if it is exposed to daylight for long periods. It is best to keep the sensor covered and dark at all times until you actually image with it.

It may take a few full flood cycles to completely empty the pixel wells.

Some CCD chips have a built in Infra-Red pre-flash to illuminate and flood the pixel wells automatically before the exposure is taken but I don't know if the Atik 314 has this function.

This would explain why the dark pixels gradually disappear after you have taken several frames.

Hopefully this may solve the problem for you.

William.

[stang]

Condensation maybe? Have you tried keeping the camera in an airtight container with some dessicant when not in use?

[ollypenrice]

Might be trapped electrons in the pixel wells.

Try flooding the sensor with white light by shining a torch into the OTA and capturing a few frames before you begin your imaging run.

Trapped electrons can occur for many reasons including high storage temperature, rapid temperature changes, strong static electricity fields and whether the camera is stored with a cap to keep the sensor dark until use or if it is exposed to daylight for long periods. It is best to keep the sensor covered and dark at all times until you actually image with it.

It may take a few full flood cycles to completely empty the pixel wells.

Some CCD chips have a built in Infra-Red pre-flash to illuminate and flood the pixel wells automatically before the exposure is taken but I don't know if the Atik 314 has this function.

This would explain why the dark pixels gradually disappear after you have taken several frames.

Hopefully this may solve the problem for you.

William.

Excellent information and all new to me. Thanks.

Olly

[Oddsocks]

Did a little more research on the web last night after my first post regarding trapped electrons.

I found most of this in published university papers but very little on amateur astronomy.

This effect varies from one type of CCD to another, full frame CCDs with a mechanical or electronic shutter can show whole or partial blocked dark columns, interline CCDs can show individual blocked pixels.

Also the severity can vary between back illuminated and front illuminated CCDs.

Depending on where the trapped electrons are in the chip substrate then different techniques are needed to free them.

Combining the two techniques I found from different web sources into one procedure should clear the trapped electrons effectively.

1: Carry out two or three 30 second exposures while flooding the CCD by illuminating the OTA with a white light source for a OSC camera (green is better for a monochrome CCD).

The exposure needs to be long enough and the light bright enough to completely fill the pixel wells and overflow into the anti-blooming gate.

2: Wait for a minute or two for the CCD temperature to return to set point after the last read.

3: Carry out at least four dark-bias exposures with the OTA capped and zero exposure length (minimum possible time) with thirty seconds between, this will flush the transfer gates.

4: Wait for a minute or two for the CCD temperature to return to set point after last read.

5: Delete all frames captured so far.

6: Begin image run, try to leave at least a minute between each CCD read function and beginning the next light/dark/bias frame etc.

Hope this is the answer.

William.

[gustavo_sanchez]

Thanks a lot for the insights. Two nights ago, I started cooling the camera much earlier, and after it stabilized I took a few test exposures (about a minute in duration). I found no dark pixels.

Sent from my Nokia Lumia 920