Author Topic: Strange Statistics  (Read 2926 times)

Offline astroedo

  • PixInsight Addict
  • ***
  • Posts: 171
  • Io ne ho viste... cose che voi umani...
    • L'arciere celeste
Strange Statistics
« on: 2012 September 10 14:43:25 »
Hello to the forum.
My name's and i'm new to pixinsight: this evening i was playing with Menus, processes and so on.

I tried to get statistics from an image (A dark frame) and i get strange results:
The image is a single 60 seconds dark frame from an Black and white Orion CCD (16 bit pixel depth).
If i use statistics i get this results:

K
Mean         34360.0
Median      34359.0
AvgDev     17.2
StdDev      118.3
Variance    0.2
Minimum    34252.0
MinPos      14,218
Maximum   62044.0
MaxPos     198,551

If i do the same with another software (Maxim) i get

Maximum    29276.000
Minimum     1484.000
Average     1591.988
Std Dev      118.309

As you can see Average, maximum and minimum values are completely different and the only consistent data is the standard deviation.

I tried the same with a color Meade DSI II camera obtaining an even odder result:

Pixinsight

Mean          497.209
Median       456.776
AvgDev       91.703
StdDev        828.911
Variance     10.484
Minimum      0.000
MinPos        516,456
Maximum    65535.000
MaxPos      678,9
Maxim

Maximum    65535.000
Minimum     5018.200
Average     5477.336
Std Dev      765.438

Another strange thing is the Variance value: the StdDev value is the square root of variance, but in the results this simple relation is completely missing.

Can someone explain me this odds results?

Thanks

Cieli sereni
Edoardo Radice

Offline oldwexi

  • PixInsight Guru
  • ****
  • Posts: 627
    • Astronomy Pages G.W.
Re: Strange Statistics
« Reply #1 on: 2012 September 10 16:19:59 »
Eduardo!
Check the file save in Maxim.
I assume you saved the data as signed integer which
adds more or less 32k to the values.
Save as floatingpoint image in Maxim transfer it ti PI and you will get the
wanted result.

Gerald

Offline astroedo

  • PixInsight Addict
  • ***
  • Posts: 171
  • Io ne ho viste... cose che voi umani...
    • L'arciere celeste
Re: Strange Statistics
« Reply #2 on: 2012 September 11 10:56:27 »
Thank You Gerald for fast reply, unfortunately your solution does not works  :-\

I do not know how the image was saved (and wich was the software) because it is not mine.
Saving in IEEE Float format The statistics are changed but are still different.
The first image gives

Mean         254.641
Median      252.312
AvgDev     40.573
StdDev     278.978
Variance   1.188
Minimum   0.000
MinPos     14,218
Maximum  65535.000
MaxPos    198,551

Any other ideas?

Note that this is a minor issue for me i Would like only to understand this strange behavior

Bye

Cieli Sereni
Edoardo Radice