2 NoiseEvaluation

[RenaudVL]

Hello,

I wanted to compare the noise from my Central DS camera a CDS600D (basically a cooled T3i) to a regular Canon T3i

Both dark frames at ISO 800 for 300sec. The CDS600D was at -28 deg and the T3i at +15 deg.

I have subtracted a BIAS from each Dark frame, I took only one of each.

This is the result from NoiseEvaluation, is there a significant difference between the two, considering a 43 deg difference...

Thanks

T3I_DARK_DARK_300s_800iso_+15c_20160130_16h29m01s794ms_c
Calculating noise standard deviation...

* Channel #0
?R = 6.915e-04, N = 1551086 (8.61%), J = 4

* Channel #1
?G = 5.002e-04, N = 4112375 (22.81%), J = 4

* Channel #2
?B = 5.898e-04, N = 2479233 (13.75%), J = 4


run --execute-mode=auto "/Applications/PixInsight/src/scripts/NoiseEvaluation.js"

Processing script file: /Applications/PixInsight/src/scripts/NoiseEvaluation.js

CDSDARK_DARK_300s_800iso_-28c_20160130_16h45m04s137ms_c
Calculating noise standard deviation...

* Channel #0
?R = 4.313e-04, N = 1402956 (7.78%), J = 4

* Channel #1
?G = 3.105e-04, N = 3542694 (19.65%), J = 4

* Channel #2
?B = 3.665e-04, N = 2110012 (11.71%), J = 4

[RenaudVL]

The thing is that I don't know how to interpret these number...

[oldwexi]

The lower the value, the less noise.

Test it.
Create an Image with PixelMath
for example set all Pixels to 0.2

run the noise control all numbers are Zero because no noise at all
if you have only one Pixel value in he whole Picture.

Or take an Image.

Run the noise control

run any noise reduction tool.

run the noise control again.

You will see the numbers of the untreated Image are higher than
those of the denoised Image.

how are these numbers calculated.
* average of all Pixels is calculated
* The average value is subtracted from each Pixel, this gets you minus and plus differences
* These difference are brought to an absolute value (remove sign)
* These values are multiplicated each with itself
* These values are summed and divided by  (number of values -1)
   Here you geht the so called variance
* From the variance you take the square root
  thats the value which is shown as Sigma K, the Std. Dev.

Can provide you an Excel table which shows this.
Can provide you a PI project which does the same with PixelMath on a test image.

Gerald