Help in noise evaluation script

[gvanhau]

Hello
Visually I am seeing that my calibrated images are more noisy than my original ones
In order to verify this I run the noise evaluation script 
But I dont know how to interpret the result.
So here is the result of running the script on two couple of images.

Is N the resulting Noise in the image?

Regards
Geert


NGC104_010r100
Calculating noise standard deviation...

* Channel #0
?K = 7.880e-004, N = 6640302 (77.85%), J = 4


run --execute-mode=auto "C:/Program Files/PixInsight/src/scripts/NoiseEvaluation.js"

Processing script file: C:/Program Files/PixInsight/src/scripts/NoiseEvaluation.js

NGC104_010r100_c
Calculating noise standard deviation...

* Channel #0
?K = 8.152e-004, N = 6816046 (79.91%), J = 4


Reading image:
G:/Mis Imagenes/2012-10-15 NGC104/NGC104_3-007R.fit
Reading FITS: 16-bit integers, 1 channel(s), 3362x2537 pixels: 100%

Reading image:
G:/Mis Imagenes/WorkDirectory/NGC104_3-007R_c.fit
Reading FITS: 32-bit floating point, 1 channel(s), 3362x2537 pixels: 100%

run --execute-mode=auto "C:/Program Files/PixInsight/src/scripts/NoiseEvaluation.js"

Processing script file: C:/Program Files/PixInsight/src/scripts/NoiseEvaluation.js

NGC104_3_007R
Calculating noise standard deviation...

* Channel #0
?K = 1.016e-003, N = 5587014 (65.50%), J = 4


run --execute-mode=auto "C:/Program Files/PixInsight/src/scripts/NoiseEvaluation.js"

Processing script file: C:/Program Files/PixInsight/src/scripts/NoiseEvaluation.js

NGC104_3_007R_c
Calculating noise standard deviation...

* Channel #0
?K = 1.054e-003, N = 5850371 (68.59%), J = 4

[mschuster]

Geert, the first number is the noise. N and the percent are the number of pixels and the fraction of the total in the image used to estimate the noise. J is number of layers. The script uses wavelet decomposition to find those pixel free of deep-sky structure, ie background pixels. Noise is then measured for these pixels.

I like to convert the noise number to electrons by multiplying by gain and camera resolution. For a 16-bit camera with 0.5 e-/ADU multiply by 0.5 * 65535. Then it is easy to compare the result with camera read noise for example.

Mike

[gvanhau]

Thanks Mike.