Introduction and request for help

[pfile]

well, i did not want to download everything, so i simply downloaded one of each frame and put all of them into ImageCalibration. i selected calibration of the dark and flat frame, but no dark scaling.

i selected no dark scaling because the dark is 180s long and so is the light.

the result looks great.

if i turn on dark scaling (optimization) then i get the 'no correlation' message. this is probably happening because the dark signal is actually very low in the dark; it is comparable to the bias signal. at -35C it appears not much dark current builds up after 180s.

anyway, while PI can sometimes produce better calibration results when scaling a dark of equal length to the light, here i would not worry about it and simply not attempt to scale the dark.

as for the bad column problem, can you point out which one(s) of the lights has this problem? i'm sure it's just a matter of configuring CosmeticCorrection properly for these frames.

thanks

rob

[bianry]

Thanks rob.
I am not a computer with PI right now but I will try it tomorrow.
And I will take a look at the light frames again to identify where the dark column is best seen.

Many thanks for your effort. Will report back.

regards

Mats

[Harry page]

Hi

I downloaded some bias and darks and find the bias has more noise than the dark ( tiny amount more )

Scaling works by subtracting the bias from the dark leaving only the dark current to be scaled by simple maths

But in your case you end up with negative numbers and this causes the problem 

In your case I would not use darks at all just use the bias for everything

Regards

Harry

[sreilly]

If you aren't scaling anything, how do you just use a bias and not the dark? I could understand maybe no bias and the dark.

[Harry page]

Hi

I should have said do not use scaling at all and just use a bias for the dark as well

I have not used a dark in years 

If using the BPP script just put a bias in the bias box and leave the dark box empty and the script will use the bias for everything 8)

Harry

[sreilly]

Not to split hairs but wouldn't that be dependant on the characteristics of your particular camera? Some have much lower noise than others. Take the Sony chips vs the Kodak chips. I haven't tried that with my two SBIG cameras, both Kodak chips, but maybe I need to see what happens.

[Harry page]

Hi

I can only speak from using my cameras , the sx h35 is a Kodak chip

If you are using say a dslr darks will still be required

Harry

[bianry]

Ok.
Thanks to all who has helped with this.

Turning off Optimization (scaling) in the calibration (both BPP script and manual) fixed the problem with 'Warning: No correlation...' issue. One down.

Now I need the fine tune the calibration parameters (integration of darks and flats).

What I would need is some help with interpreting the results from the Integration console.

My own visual subjective appreciation of the result tells me that the integrated picture I get by using iTelescopes calibrated frames LOOK better than when using my own calibrated frames.

BUT if I understand the output of the integration script there is more noise in the image made by iTelescopes frames.
Ie. the picture LOOKS better to me but it is actually worse.
I am the first to admit that I have a very limited experience in visually inspecting an astronomical image and it might be that my lifelong experience in handling "normal" pictures is fooling me.
OR it might be that I do not understand the readout from the console.

Image integrated with frames calibrated by BPP

Code: [Select]

Output from Noise evaluation script
?K = 7.157e-005, N = 4276627 (67.98%), J = 4

Output from integration script
Integration of 17 images:
Pixel combination ......... average
Output normalization ...... additive + scaling
Weighting mode ............ noise evaluation
Scale estimator ........... iterative k-sigma / BWMV
Range rejection ........... range_low=0.000000 range_high=0.980000
Pixel rejection ........... Winsorized sigma clipping
Rejection normalization ... scale + zero offset
Rejection clippings ....... low=yes high=yes
Rejection parameters ...... sigma_low=4.000 sigma_high=2.000

MRS noise evaluation: done
Computing noise scaling factors: done

Gaussian noise estimates  : 7.1574e-005
Scale estimates           : 1.1722e-004
Location estimates        : 1.2052e-003
SNR estimates             : 5.5517e+003
Reference noise reduction : 1.5111
Median noise reduction    : 1.6573
Image integrated with frames calibrated by iTelescopes system

Code: [Select]

Output from Noise evaluation script
?K = 8.347e-005, N = 4712765 (74.91%), J = 4

Integration of 17 images:
Pixel combination ......... average
Output normalization ...... additive + scaling
Weighting mode ............ noise evaluation
Scale estimator ........... iterative k-sigma / BWMV
Range rejection ........... range_low=0.000000 range_high=0.980000
Pixel rejection ........... Winsorized sigma clipping
Rejection normalization ... scale + zero offset
Rejection clippings ....... low=yes high=yes
Rejection parameters ...... sigma_low=4.000 sigma_high=2.000

MRS noise evaluation: done
Computing noise scaling factors: done

Gaussian noise estimates  : 8.3471e-005
Scale estimates           : 1.0707e-004
Location estimates        : 1.5804e-003
SNR estimates             : 3.4925e+003
Reference noise reduction : 1.2206
Median noise reduction    : 1.2180
The one that LOOKS worse to me (calibrated by BPP) has an SNR of 5.5517e+003 and the one that looks better has an SNR of 3.4925e+003

Should I not strive for a higher SNR??

Also: How should I read the output from the noise evaluation
?K = 7.157e-005, N = 4276627 (67.98%), J = 4
Have tried to search the forum for an explanation what the output means but so far no luck.

There is a lot of questions here but I am really struggling to LEARN what I am doing. Have used PS for +15 years and the only thing I learned there is workflows. Do this to achieve that. I never understood why something worked.

PI is very different and I appreciate that a lot.

Attaching Dropbox links to both images.
https://www.dropbox.com/s/bgildmbfcacs4yz/own_calib.fit
https://www.dropbox.com/s/micnpyeaxt8q7v7/iTelescope_calib.fit

Regards

Mats

[bianry]

well, i did not want to download everything, so i simply downloaded one of each frame and put all of them into ImageCalibration. i selected calibration of the dark and flat frame, but no dark scaling.

i selected no dark scaling because the dark is 180s long and so is the light.

the result looks great.

if i turn on dark scaling (optimization) then i get the 'no correlation' message. this is probably happening because the dark signal is actually very low in the dark; it is comparable to the bias signal. at -35C it appears not much dark current builds up after 180s.

anyway, while PI can sometimes produce better calibration results when scaling a dark of equal length to the light, here i would not worry about it and simply not attempt to scale the dark.

as for the bad column problem, can you point out which one(s) of the lights has this problem? i'm sure it's just a matter of configuring CosmeticCorrection properly for these frames.

thanks

rob

Rob

The bad column(s) is best visible in light frames 4,10-17
Xpos 2060 -> 2075 about

regards

Mats

[pfile]

ok i'll look when i get home later

rob