A test of the linearity of the response curve for the Sony ICX285AL Exview HAD CCD chip of the Starlight Xpress SXV-H9 monochrome camera was performed in two lighting conditions ("dark", "bright") based on exposures of an evenly illuminated flat field screen. The Westerlund Telescope stopped down to 30 cm off-axis aperture provided the photons.

The response of the chip is remarkably linear though the camera has vertical anti blooming as default. The following data and plots gives ADU values and deviations from linearity following subtraction of bias current. The CCD saturates at 16 bit intensity (65500 ADU).

Results

• The shortest usable exposure time is 0.13 seconds, below which the "toe" intensity level of the response curve is independent of exposure time.
• The chip is linear to better than 1% for the range 230-61000 ADU. The "noise" level is about 0.5%.
• At low illumination intensities (<250 ADU), a "toe" in the response curve is evident. The deviation from linearity is +0.2% for 250 ADU and increases to +12.8% for 115 ADU.
  • To correct for nonlinearity at pixel intensities <250 ADU, multiply recorded (following basic image reduction) pixel ADUs with the following correction factors (F):

    ADU	F
    115	0.872
    128	0.919
    142	0.952
    158	0.970
    175	0.977
    192	0.983
    210	0.986
    227	0.989

• At high illumination intensities (>500 ADU), the response curve is linear throughout basically the full range. The deviation from linearity is <1% up to 61000 ADU.

Conclusions

The SXV-H9 camera may be used for 1%-level precision photometry without linearity corrections in the pixel intensity range 230-61000 ADU above bias level. The minimum useful exposure time is 0.13 seconds. Correction factors need to be applied for low intensity level photometry at ADU's <230.