Calibration Problem: `No correlation' and `binarized-looking' results

[marekc]

Hi Vincent, Juan, and Ioannis,

Thank you all very much for helping me with this! I'm really grateful that you've taken the time to help me with what is basically an issue with my camera and my acquisition software.

When I first got the camera, I was acquiring with the evaluation version of Maxim that came with the camera. Around the time that software expired, Orion Telescopes posted an ASCOM driver, and I switched to acquiring with the ASCOM driver and Nebulosity. (Basically, I was being cheap; this avoided having to buy Maxim.)

Interestingly, both Maxim and Nebulosity seem have acquired bias frames with values around 3000. This might fit with Ioannis's idea that the drivers might be adding an offset to the bias.

In Maxim, I was able to select options like `dark', `bias', and `light'. In Nebulosity, I don't have those options, so my bias frame is just a light frame with a 0.0 sec exposure time. (I have a cap on the nosepiece of my camera, covered with black gaffer's tape, and I put the camera outside, at night, to help keep it cool when I shoot biases and darks. The camera also has an internal shutter - but maybe that's only automatically closed when `dark' or `bias' is selected, as in Maxim. At any rate, I'm confident that I've kept light off of my chip when shooting darks and biases, regardless of the software.)


Here are some single raw frames acquired in Maxim:

http://dl.dropbox.com/u/11055667/Maxim_1x1_neg10_bias.fit

http://dl.dropbox.com/u/11055667/Maxim_1x1_neg10_dark.fit

http://dl.dropbox.com/u/11055667/Maxim_1x1_neg10_flat.fit

http://dl.dropbox.com/u/11055667/Maxim_1x1_neg10_light.fit


Here are some single raw frames acqured with Nebulosity and the camera's ASCOM driver:

http://dl.dropbox.com/u/11055667/Nebulosity_1x1_neg20_bias.fit

http://dl.dropbox.com/u/11055667/Nebulosity_1x1_neg20_dark.fit

http://dl.dropbox.com/u/11055667/Nebulosity_1x1_neg20_flat.fit

http://dl.dropbox.com/u/11055667/Nebulosity_1x1_neg20_light.fit


Thanks again so much for helping me try to figure this out!

- Marek Cichanski

[vicent_peris]

hmmm... something happens. Bias level is correct in the frames you uploaded. In fact, if you subtract the bias to the dark in PixelMath, the resulting image has values >0.

Have you tried to calibrate with the new preprocess pipeline script??


Vicent.

[marekc]

Hi Vincent,

Thanks for looking at the data. I've tried to duplicate the steps you've taken. Here's what I've gotten:

1) I subtracted a bias from a dark, and I got a lot of 0s. Am I correct in assuming this means the bias had higher values than the dark in these pixels? I assume that if I do [Image 1] - [Image 2] and get a value that's less than 0, PixelMath will give a 0 value for that pixel.

Did you perhaps mean "the resulting image has values <0", rather than "...>0" ?

2) To double-check that I'm doing the PixelMath correctly, I tried subtracting a dark from a bias, and I got a lot more pixels with the value of 0. That makes sense. Most of the pixels in the dark frame have higher values than their counterparts in the bias frame.

Still... it appears that the dark frame has a significant number of pixels with values *below* those of the bias frame.

That seems strange. I would expect the following to happen:

a) each pixel would have a small number of electrons put into it, to make the bias value.
b) As the dark frame integration proceeds, the dark current would put more electrons into each pixel.
c) As a result, the ADU value of a pixel in the dark frame would be larger than its value in a bias frame.

But, with my bias and dark frames, it's as though two strange things have happened:

i) Essentially no dark current has accumulated, even in a 60-second exposure, or a 360-second exposure.
ii) A lot of electrons were injected in order to make the bias level, more than some cameras use.

It's almost as though the dark frame isn't getting any extra electrons, even though it should be. It's as though the software has said "if there's no light falling on the chip, just stick in a value of about 3000".

I know that Sony has made some very low dark-current sensors, but this is the Kodak 8300 sensor, and I don't think it's supposed to be like that. I'm using -10C and -20C in these frames, which isn't nearly as cold as the -100C that the pros use. So, I'd expect to be accumulating dark current on top of the bias. It's strange that I don't seem to be seeing that.

I'll try to use the new preprocessing script, too. Do you think we're seeing evidence of some strange behavior in the acquisition software?

Thanks

- Marek Cichanski

[Alejandro Tombolini]

Hi Marek, I try both sets of images using individual frames as Masters and they seems work well.
Calibration using images from Maxim do not give any warning.
Calibration using images from Nebulosity give the warning!

Attached calibrated images and process.

http://www.4shared.com/file/q2mb5MWM/Maxim_1x1_neg10_light_c.html
http://www.4shared.com/file/_2J9_VF4/Nebulosity_1x1_neg20_light_c.html

Processing Console

open "file:///F:/Mis Imágenes/Maxim_1x1_neg10_light_c_5.fit"
Reading 1 file(s):
F:/Mis Imágenes/Maxim_1x1_neg10_light_c_5.fit
Reading FITS: 32-bit floating point, 1 channel(s), 3326x2504 pixels: 100%

ImageCalibration: Global context

Loading master calibration frames:
Loading calibration frame image:
F:/Mis Imágenes/Maxim_1x1_neg10_bias.fit
Reading FITS: 16-bit integers, 1 channel(s), 3326x2504 pixels: 100%
Loading calibration frame image:
F:/Mis Imágenes/Maxim_1x1_neg10_dark.fit
Reading FITS: 16-bit integers, 1 channel(s), 3326x2504 pixels: 100%
Loading calibration frame image:
F:/Mis Imágenes/Maxim_1x1_neg10_flat.fit
Reading FITS: 16-bit integers, 1 channel(s), 3326x2504 pixels: 100%

Applying bias correction: master dark frame ...
Computing master flat scaling factors ...
s0 = 0.457375

Calibration of 1 target frames:
* Using 1 worker threads

Calibrating target frame 1 of 1
Loading target frame:
F:/Mis Imágenes/Maxim_1x1_neg10_light.fit
Reading FITS: 16-bit integers, 1 channel(s), 3326x2504 pixels: 100%

Writing output file: F:/Mis Imágenes/Maxim_1x1_neg10_light_c.fit
Dark scaling factors:
k0 = 0.021
Writing FITS image: 32-bit floating point, 1 channel(s), 3326x2504 pixels: 100%

===== ImageCalibration: 1 succeeded, 0 failed, 0 aborted, 0 skipped =====
7.938 s


ImageCalibration: Global context

Loading master calibration frames:
Loading calibration frame image:
F:/Mis Imágenes/Nebulosity_1x1_neg20_bias.fit
Reading FITS: 16-bit integers, 1 channel(s), 3326x2504 pixels: 100%
Loading calibration frame image:
F:/Mis Imágenes/Nebulosity_1x1_neg20_dark.fit
Reading FITS: 16-bit integers, 1 channel(s), 3326x2504 pixels: 100%
Loading calibration frame image:
F:/Mis Imágenes/Nebulosity_1x1_neg20_flat.fit
Reading FITS: 16-bit integers, 1 channel(s), 3326x2504 pixels: 100%

Applying bias correction: master dark frame ...
Computing master flat scaling factors ...
s0 = 0.412975

Calibration of 1 target frames:
* Using 1 worker threads

Calibrating target frame 1 of 1
Loading target frame:
F:/Mis Imágenes/Nebulosity_1x1_neg20_light.fit
Reading FITS: 16-bit integers, 1 channel(s), 3326x2504 pixels: 100%

Writing output file: F:/Mis Imágenes/Nebulosity_1x1_neg20_light_c.fit
Dark scaling factors:
k0 = 0.000
** Warning: No correlation between the master dark and target frames.
Writing FITS image: 32-bit floating point, 1 channel(s), 3326x2504 pixels: 100%

===== ImageCalibration: 1 succeeded, 0 failed, 0 aborted, 0 skipped =====
8.922 s