interpreting bit depth and bit alignment

[berrypi]

I am trying to use a Point Grey Grasshopper CCD camera (GRAS-14S3M-C) which has a 14-bit ADC.

In NI Max, I configure the settings as below.

The gain is set to 1 (0dB) and the gamma setting is ignored, so that the camera output is linear.

The question is what do the settings in the box "16 bit pixel representation" do to how the Camera data is interpreted?

Given that the ADC on the camera is 14-bit, I assume that I should set the "Actual Bit Depth" to 14-bit.

If bit alignment is LSB, I would expect the captured image to pad 2 MSBs with 0s, and take the camera 14-bit output and interpret it as varying in value from 0 to 2^14-1. However, I see that the pixel counts continue to vary from 0 through 2^16-1 = 65535.

In the end, what I want to do is convert the counts registered on the CCD to number of incident photons. Assuming a quantum efficiency of say 30%. But how do the settings in the box "16 bit pixel representation" influence this calculation?

It would be very helpful if you could reference a note/ white paper which sheds light on this.

[Hatef.fouladi]

image in gray scale just will be available to show in 8bit or 16 bit so the data came from your hardware(ccd) scaled into one of this two kind of showing image(16 bit or 8 bit)
now if you want to calculate QE  or not,  you know the Qe and just want to know number of photon  your accurate will be 14 bit it means that if you know what is your maximum working voltage and number of electron transmit for this voltage inside your ccd that make your ccd in Saturation mode then you can calculate amount of incident photon with this formula
Nph= number of electron *Qe  and your accuracy to counting number of photon will be( max num  of electron/(2^16))*Qe