I placed the camera level, facing 90° towards a window. On the window, I placed a measuring band. The distance from the camera's mount to the measuring band is 870mm. This means that the distance to the sensor is 890mm, since the register distance of Micro Four Thirds is 20mm.

Here is the setup:

And this is the resulting image, after in-camera distortion correction:

It's easier to measure the width when looking at 100% crops of the endpoints:

We can see that the total horizontal distance is 1039mm.

Doing the basic trigonometrical calculation, we get the horizontal angle of view to be:

*2*arctan((1039mm/2)/890mm)° = 60.5°*.

Now, we want to find the diagonal angle of view. The diagonal width is

*sqrt(1039*. Hence, the diagonal angle of view is

^{2}+(1039*3/4)^{2}) = 1298.75*2*arctan((1298.75mm/2)/890mm)° = 72.2°*.

Panasonic's specifications state that the diagonal angle of view is 75°. However, we know that the angle is specified at infinity focus. And in my case, the focus is just below 1m. It is also a fact that the angle of view can change with the focus, especially with internal or rear focusing mechanisms. So this is probably the explanation for the difference.

There's an easier way:

ReplyDeleteDFOV = 2*(arctan(d/2f))

where

d = sqrt((w^2)+(h^2))

w = sensor width (17.3mm)

h = sensor height (13.0mm)

f = focal length (14mm)

(DFOV: "diagonal field of view")

It changes at closer distances because f increases as you focus closer.

That assumes that the specified focal lengths and sensor widths/heights are correct and everything is working properly as specified. The type of measurement he shows here is something that, for certain applications in the real world, you would want to do at least once for various focus in order to verify that the camera specifications are correct, and that there is nothing wrong with your camera. You would then use your simple theoretical calculations to make sure you get agreement with the camera specs.

ReplyDelete