Mismatched aperture values?

[Neal Krawetz]

I'm not sure where else to post this...

I occasionally see camera-original photos that define both EXIF FNumber and EXIF ApertureValue, but they give them different values. They can differ by as much as 0.6.

For example:
https://fotoforensics.com/analysis.php?id=78a67404f301a4ad785b3d332375345cb5f119fc.2461781&show=meta
This is a camera original from a Fuji FinePix F70EXR.
F Number = 3.3
Aperture Value = 3.2

https://fotoforensics.com/analysis.php?id=c6e481c76985cc0d3bcc42833f7fec26e8240293.1471216&show=meta
This is a Samsung Galaxy Nexus.
F Number = 2.8
Aperture Value = 2.6

When both values are present, it's almost 50/50 whether they are the same or differ.
There is also no consistency which one is bigger. (At minimum: 'bigger' seems to vary by camera.)

Does anyone know why these values differ?

I have some theories, but I don't know if any are correct:
  - The aperture on a digital camera is an approximation (no real iris). This is the range scanned when auto-determining the brightness.
  - The aperture was is an approximation, and the camera just guessed values.
  - I've ruled out HDR for these pictures, but maybe the camera records different apertures "just in case" someone wanted to enable HDR mode.
  - In the digital camera, it uses F Number for one purpose and Aperture Value for some other purpose. Both are related, but undocumented.

Any insight would be greatly appreciated.

[Phil Harvey]

FAQ 4 deals with this question.

- Phil

[Neal Krawetz]

Thanks Phil!