Bias master frames comparison

[johnrt]

Evening!

I was recently sent some data to process from a fellow Atik 314l+ owner, which consisted of lights, flats and bias frames. I stacked the bias to make a master for calibration and gave it a quick stretch to compare to my own bias frame with an identical camera. I was quite shocked to discover such a huge difference in the stretched images.

The image attached shows the two bias frames, mine on the left and the other camera on the right. As you can see my camera exhibits vertical banding down the left hand side of the chip, and looks darker when in reality it has been stretched further than the other bias frame.

My camera is powered via 12v mains adapter bought from Modern Astronomy.

I would be very interested in some thoughts on this!

[RikM]

I am afraid I don't have a clue John. But for comparison, here's a master bias from my SXV-H9, which is the same chip. I gave it an almighty stretch otherwise it looked essentially smooth.

[johnrt]

I might drop atik an email with the comparison and see what they think.

[JacobvonChorus]

My Atik 314L+ also has a bit of vertical banding. Its 100% repeatable and 100% removable so I have't had issues. The pattern in my images match the pattern Craig Stark had in his test of his Atik 314L+ exactly. Same intensities.

Jacob

[johnrt]

My Atik 314L+ also has a bit of vertical banding. Its 100% repeatable and 100% removable so I have't had issues. The pattern in my images match the pattern Craig Stark had in his test of his Atik 314L+ exactly. Same intensities.

Jacob

Do you have an example you can upload to compare to mine?

[ollypenrice]

Cables, power supplies, temperature...

Olly

[johnrt]

Cables, power supplies, temperature...

Olly

Powered from mains via 12v adapter bought from Bern @ MA, I've tried the camera with lots of cables both old and new, always the same since I bought the camera new from Ian Kng. Temp - set to -25C on my camera permanently, these bias were done indoors so where the camera can get down to about -10/15C.

John

[rfdesigner]

you will get chip to chip variation, even on a single wafer, I work in semiconductor development, we had a wafer recently where all the passed chips were round one edge, while the rest of the wafer was dud. With these chips you can and should expect substantial variation of performance. I know Christian Buil collated some results from a series of tess of readout noise and showed that 314 cameras can have anything from ~3e to 6e of readout noise, I think the median was 4e, as per the spec.

The trick is to image such that your frame length and signal strength render these variations void.

The key parameters you can't do much about are QE and dark current. QE appears to be pretty consistant on independant tests, and dark current is essentially leakage current wich is usually controlled for during wafer production (there are some test structures on most wafers, so basic parameters can be measured and wafers discarded if too far out or the process tweaked if wafers start being consistantly a bit slow or fast)

What would be interesting from your images is to show what the scale is of the compression.. i.e. set your black point to the median value -3x readout noise, set your white point to median value +20x readout noise on both images. Then you have like for like.

Derek

[andyo]

When I read this post I noticed my bias are nearly the same I thought this was normal as the lights are ok and don't show this after calibration,Below is a post of one of my master bias frames also taken with an Atik 314l+ the banding is on the rightside would be interested to see what atik say.

MASTER BIAS 1x1.fit