how does a 14bits camera generate a 16 bits image with IMAQdx?

[xu27]

Hi,

I am using Guppy Pro F-031B firewire camera.   And I am programing it with IMAQdx. 
The AD converter of this camera is 14bits,  and I was expecting that the counts per pixel of this camera should not exceed  2^14.  In the attached picture,  the maximum count for each pixel is 18132, which is larger than the 2^14.  I do set a 16bits grayscale output.  How did a 14bits camera generate a 16bits output usually? 


My setup:
1. Ni IMAQdx with Microsoft driver.
2.   VideoMode = "Format 7, Mode 0, 656 x 492"
       PixelFormat = "Mono 16"
       BitsPerPixel = "Use hardware value"
       PixelSignedness = "un Signed"
       ShiftPixelBits = "false"
       AutoExposure, Brightness,Gain,Gamma=ignored

Thank you very much!

best,
wen

[MoviJOHN · Accepted solution]

IMAQdx does not shift or stretch the data, as far as I know. Change the endianess, and see if that makes a difference.

[xu27 · Accepted solution]

John,

Thank you! 

I found the problem.  My camera default setting is shift the bits. So 1111   1111   1111   11 is manupulate into 1111   1111   1111   1100.  

Best,

Wen 

[BlueCheese · Accepted solution]

Yes, many cameras left shift the data so the range is identical regardless of the number of bits. This allows image processing code to run identically on any image from 10-16 bits.