The voltage and power calibrations in Fig 3 and Fig 4 in the last comment were incorrect. The following figures show the correct calibration.

* Fig 1: time series data of 3 PDs showing the DC output. (No change)

* Fig 2: row spectrum in counts. (No change)

* Fig 3: Updated calibration of PDs in voltage.

- Assumption for calibration: the PD output range is ±20V and the ADC is 2^16 bit.

- IR PD setting: no PD gain, no whitening filter gain, 0 filter stage, 0.255 filter bank gain.

- IR S-pol setting: 30 dB PD gain, no whitening filter gain, 3 filter stages, no filter bank gain.

- IR P-pol setting: same as s-pol PD.

(1) IR PD DC output: 4.2 counts / (2^16 counts) * 40 V / 0.255 = 1.0e-2 V

(2) IR S-pol PD DC output: 24 counts / (2^16 counts) * 40 V = 1.5e-2 V

(3) IR P-pol PD DC output: 390 counts / (2^16 counts) * 40 V = 2.4e-1 V

* Fig 4: Updated calibration of PDs in watts.

- PDA100A2 data sheet: Responsivity 0.39 A/W, gain at 0 dB: 1.51e3 V/A, gain at 30 dB: 4.75e4 V/A

(1) IR PD DC power: 1.0e-2 V / (1.51e3 V/A) / (0.39 A/W) = 1.7e-5 W

(2) IR S-pol PD DC power: 1.5e-2 V / (4.75e4 V/A) / (0.39 A/W) = 8.0e-7 W

(3) IR P-pol PD DC power: 2.4e-1 V / (4.75e4 V/A) / (0.39 A/W) = 1.3e-5 W