Maximum pixel value too low?

[Altais]

Hello.

I've encountered something I don't understand when I open a CR2 file from my Canon EOS 1100D in PixInsight.

Under the DSLR_RAW entry in Format Explorer, I use the "Pure Raw" preset in preferences. When I measure the 14-bit pixel value in the center of a completely saturated star (using a probe size of 1 pixel), it reads 3826. Since the CR2 files are 14-bit, shouldn't it read 16384 (or slightly less)? Or am I missing something here?

This is problematic because I'm trying to write a SNR calculator for my camera, and now I don't know if I can trust the pixel values that I use for calculating an approximation of the shot noise.

Lars

[Juan Conejero]

Hi Lars,

Digital cameras store raw images as 16-bit unsigned integer data, irrespective of detection bit depths. So don't select the 14-bit integer range; always select the 16-bit range to read raw pixel sample values.

And yes, you can trust raw pixel sample values that you read in PixInsight

[Altais]

Thanks a lot!

[Cleon_Wells]

Hi Juan,  the T1i Canon spec is (14bit , Canon Original), what ever that is; did Canon drop the bit depth from 16bit  to 14bit to increase the frame rate  of the processor for video?
I remember back in 2010 I could increase the Brightness level in the DSLR_RAW format preferences and get a 16bit DN clip level when I created a .fit file
I have wonder way this DSLR_RAW Brightness adjustment no longer does this.
Cleon