Dark Current Enable Tool

[Iliah Borg]

Here are some Nikon D810 ~300 sec shots taken at different ISO settings
http://s3.amazonaws.com/IliahBorg/Nikon ... C_1686.NEF
http://s3.amazonaws.com/IliahBorg/Nikon ... C_1685.NEF
http://s3.amazonaws.com/IliahBorg/Nikon ... C_1684.NEF
http://s3.amazonaws.com/IliahBorg/Nikon ... C_1683.NEF

[astronomer]

Here are some Nikon D810 ~300 sec shots taken at different ISO settings

Thanks,

From the histogram, the bias level always centered at 600ADU. I think it runs into old Canon issues now. In an actual sensor output using my dark current tool, you would expect black point gradually deviates from the ADC offset. This deviation is called dark current and the average deviation varies from pixel to pixel, which is called Dark Signal Non-Uniformity.

In astronomy, we stack many dark frames to get the average dark signal value from each pixel and subtract it from light frame to do the calibration. If this is not done correctly, the next step of flat fielding will be problematic.

Now dark frame usually will not be perfect, thus scaling factor can be applied. But now with everything moved to 600ADU, you could not estimate the dark current.

[astronomer]

Lenses that I use when doing these experiments do not transfer aperture values to the camera. This switches off aperture-related scaling effects. I never saw shutter speed affecting gaps, but maybe I'm missing something. Gaps, as I saw, depend on particular camera sensor calibration (what Nikon are calling WB pre-conditioning), and yes, it is an operation over integers. I saw digital scaling depending on ISO setting on, say, Canons; but not on Nikons within the calibrated ISO range.

What astronomer was meaning is the gaps in the data seen in above picture in the red and blue channel. The gaps are because the R/B channels are scaled in integer space, and the scaling is a function of the aperture/shutter/ISO setting of the camera, not related to your actual awesome glass.

Both are right, I had a detailed post months ago before the raw cooking is hacked. Generally there's a fixed scaling to Red and Blue channel which is calibrated for each sensor during color adjustment I guess. And a dynamic digital scaling applied to all channel depending on the aperture value being set.

@Iliah,
Manual lens will disable the dynamic scaling only. But it usually won't solve the problem since new lens has much better optical design, such as less coma and astigmatism, both good for astrophotography.

Anyway, I'm not digging into EXPEED 4 now since all those are based on ARM, and the TFT table holding the calibrated data is completely different.

[Iliah Borg]

> From the histogram, the bias level always centered at 600ADU.

Right. Let's see some "mode 3" shots tomorrow. It switches off long exposure noise reduction / median filter.

> we stack many dark frames to get the average dark signal value from each pixel and subtract it from light frame to do the calibration

Sure.

[Iliah Borg]

> Manual lens will disable the dynamic scaling only

And leave WB preconditioning / R/B calibration scaling as it is.

> new lens has much better optical design

I use EL lenses to evaluate in-camera raw processing, and rarely for imaging purposes - unless I need to shoot in UV or IR.

[astronomer]

Right. Let's see some "mode 3" shots tomorrow. It switches off long exposure noise reduction / median filter.

The mode 3 will not be effective on any camera. It just applied a different type of algorithm to image. And it will still leave artifacts with adjacent hot pixels pairs.

[Andrew]

Hi. I am very excited by this post- I use D800 for scientific research. Was the link taken down for this program? I can't seem to find it.

[astronomer]

Hi. I am very excited by this post- I use D800 for scientific research. Was the link taken down for this program? I can't seem to find it.

Now the link is restored to the updated files.

[Andrew]

Thank you very much for this tool.

Has anyone tested this with using remote control software to control camera over USB? I have been trying with the following procedure but It seems to put the camera in an error state and then it needs to be power cycled to communicate again-

1. run dark current enable tool, leave program open
2. run camera control software
3. camera control software gives error saying the device is busy

Thanks,
Andrew

[astronomer]

Thank you very much for this tool.

Has anyone tested this with using remote control software to control camera over USB? I have been trying with the following procedure but It seems to put the camera in an error state and then it needs to be power cycled to communicate again-

1. run dark current enable tool, leave program open
2. run camera control software
3. camera control software gives error saying the device is busy

Thanks,
Andrew

Hi Andrew, the 2 programs are competing for MTP access to a single device.
Can you try exit my tool, unplug and replug USB, then run your control software.