David's Astronomy Pages Notes - Session 599.2 (2017-01-16)

Notes (S599.2)

Notes Main

Home Page

Notes (S599.3)

	2017-01-16
	ZWO ASI178MC - Read Noise Tests
	Powershell - Useful Things

 ZWO ASI178MC - Read Noise Test (2017-01-16)

Tests carried out on  ZWO ASI178MC camera using SharpCap for Image Capture using the following procedure:
(Note:  tests carried out indoors at an Ambient Temp of 18 degC)

 -----------------------------------------------------------------------------------------------
 - Take Two Flat Frames (F1 and F2)
 - Take Two Bias Frames (B1 and B2)
 - Load frames into AIP4WIN
 - Measure Mean(F1+F2)  
    - Use AIP4WIN's 'Multi-Image/Image Math' faclity to add F1 and F2 frames together,
    - Then use AIP4WIN's 'Measure/Statistics/Image' facility to obtain the mean of the summed frame
 - Measure StDev (F1-F2)
 - Measure Mean (B1+B2)
 - Measure StDev (B1-B2)
 - Calculate Gain (in e/ADU)
      where Gain = (Mean(F1+F2) - Mean (B1+B2) ) / ((StDev (F1-F2))^2 + (StDev(B1-B2))^2)
 - Calculate Read-Out Noise RMS
      where Read-Out Noise = (Gain * StDev(B1-B2)) / Sqrt(2)
-------------------------------------------------------------------------------------------------

Initially this produced a Gain of 0.18 e/ADU and a ReadOut Noise of 2.34 e RMS

The calculated Read Out Noise was close to the expected value of 2.2 e RMS (from ASI178's published charts)
but the calculated gain was quite a bit smaller than the expected value close to 0.8 e/ADU.

However the frames are ones that have been stretched to 16 bit from original 14 bits in the camera
When the input measures are each divided by 4 (to take 16 bit data back to 14 bit data) and the numbers are recalculated Gain value becomes 0.71 e/ADU (i.e. much closer to the expected value). The calculated read-out noise (2.34 e RMS) remains the same.

As a separate check the original F1, F2, B1 and B2 frames were put through operations to divide each pixel value by 4 (to return them to 14 bit range), before the Mean(F1+F2), StDev(F1-F2), Mean(B1+B2), StDev(B1-B2) were re-measured. Running the calculations again gave a gain of 0.71e/ADU and a Read Out Noise of 2.29 e RMS. This is approximately the same result as above.

Thus the test has produced the following numbers :- 
  Gain  :  0.71 e/ADU  (14 bit)     (at amplification gain of 0 )
  ReadOut Noise :  2.3 e RMS    (at amplification gain of 0 )

The Gain value of <1 means that we're unable to select a amplifer gain on the ASI178 camera that corresponds to Unity Gain (i.e. the point where we have a Gain of 1.0e/ADU).  An Amplification Gain of 0 is the closest that we can approach Unity Gain.

Note:   A Single Frame from ASI178MC camera

 (Binning 1x1) is  3096 x 2080 pixels (6,439,680 total pixels).

Test 1
Bias Frames (actually 0.032s)
- at 0 gain:  Mean ADU 0 (Min 0, Max 0)
- at 400 gain:  Mean ADU 0 (Min 0, Max 0)

Colour Balance :
Best White Light Balance from R64, B98

AIP4WIN Cropped Red Channel

Histogram showed Red to be cropped in a rather unusual way in that the histogram value that its cropped at is lower than green channel. Changing the Colour Balance shifted the position of the R Channel's curve, but didn't change its cropped nature. It would seem that the Red Channel values are being cropped at the highest 14 bit value before being stretched to 16 bit and then adjusted according to Colour Balance Settings.   To prevent this the exposure time must be reduced slightly.

Back to Top

Powershell - Useful Things

Problems: SharpCap produces FITS files with the ".fits" extension. However AIP4WIN doesn't automatically recognise this file extension as an image file (have to use Show All Files option in order to to open/select such a file). Hence there is a need to be able to quickly change a collection of ".fits" file to  ".fit" files

In just the folder:
Dir *.fits | rename-item -newname { [io.path]::ChangeExtension($_.name, "fit") }

In the folder and in all subfolders below
Get-ChildItem -Recurse -Include *.fits | rename-item -newname { [io.path]::ChangeExtension($_.name, "fit") }

Back to Top