HDI Linearity Test: Jan 10, 2019

Michael Richmond
Jan 11, 2019

On Jan 10/11, 2019, RIT students Brent Neves and Teresa Symons acquired a set of dome flat images in order to check the linearity of HDI. The procedure was:

• all images were taken after sunset
• B-band filter
• low-intensity lamp, set at 97 percent
• exposure sequence of
  • 100, 90, 80, ..., 10 seconds
  • 3 bias frames
  • 10, 20, 30, ..., 100 seconds

I converted the images from the native HDI FITS format to a simple FITS image format; in the process, all pixel values were divided by 2.

Brent Neves processed the images, creating a median bias frame and subtracting it from each dome flat. He then measured the mean value of pixels within a box of 200-by-200 pixels near the center of the frame.

You can find an Excel spreadsheet with his measurements and graphs here:

• link to Excel spreadsheet with linearity data

Brent Neves made a series of graphs showing the mean counts in the box as a function of exposure time. You can find all the graphs in the spreadsheet; I'll show just two here.

The first shows

• measurements in orange (going up in exposure time, shifted slightly to right for clarity) and blue (going down in exposure time)
• in black, a linear fit to all values from "short" exposures (10-50 second), extended to 100 seconds

It is clear that the measurements at longer exposure times have values which fall below the linear extrapolation of the fit to short exposures.

Remember, y-axis shows counts divided by 2

The second graph shows the same datasets, but without the offset for the blue measurements.

Remember, y-axis shows counts divided by 2

These tests suggest a small non-linearity in the response of HDI when levels increase above 30,000 counts (= 15,000 counts/2). That result is consistent with linearity tests performed in 2015:

• Technical Note 8: HDI linearity and gain tests