Info

<a 0 : rms noise in ADU before any charge shift

Eq. (5) shows a linear dependence between the variance of the signal in line (column) i and the line number i. To estimate this equation, we will use the best fit line of the set of points (i, < (i)).

To measure the variance of each line, two flats (same level) and two biases are taken. The bias images are subtracted from the flats and these two new images are divided, one by the other. The result is multiplied by the mean of one of the images. The final result will give an image (R) with fixed pattern noise (Photon Response Non Uniformity, PRNU) flat fielded out. The total noise (in pixel units, ADU) in this image R will be composed of photon noise and read out noise essentially. To measure the photon noise, the read out noise3 is subtracted from the total noise.

To create the plot variance of lines (columns) versus lines (columns), we measure the variance for each line (column) of R. This value is divided by 3 to obtain the variance of one line and the point (i, < (i)) is plotted (see Fig. 1). During the calculation of the variance, bad pixels are eliminating by using the sigma clipping method.

The best fit line will give us:

JU : y-axis intercept (in ADU2) V : slope of the streak

The constant term JU will give us the variance of the signal before being affected by the charge transfer inefficiency and the slope, V, divided by 2 J, the charge transfer inefficiency. The charge transfer efficiency, a, is then:

This parameter is measured by subtracting two bias images. The variance of the pixels in the difference image is equal to two times the read out noise squared. The variance from this image can be directly subtracted from the variance measured in the image R to have only two times the photon noise squared in ADU.

Figure 1. Image A and B show the plot ((, cC (i)) at different CTEs with the best fit line cutting through. In A, the CTE is 0.999970 and in B, 0.999996. These plots have been made with simulated data. The dimensions of the images are 1K*1K. The number of electrons in the images is 1000 plus a Poisson noise. The CTE is applied after. Similar pattern are observed with real data. In the x-axis, the line count increases away from the readout register.

Figure 1. Image A and B show the plot ((, cC (i)) at different CTEs with the best fit line cutting through. In A, the CTE is 0.999970 and in B, 0.999996. These plots have been made with simulated data. The dimensions of the images are 1K*1K. The number of electrons in the images is 1000 plus a Poisson noise. The CTE is applied after. Similar pattern are observed with real data. In the x-axis, the line count increases away from the readout register.

Was this article helpful?

0 0
Telescopes Mastery

Telescopes Mastery

Through this ebook, you are going to learn what you will need to know all about the telescopes that can provide a fun and rewarding hobby for you and your family!

Get My Free Ebook


Post a comment