I assumed that the spec sheet 84 degree FOV is the diagonal FOV, giving a HFOV of 73.7 degrees. However, when I shot stills and videos of my jeep from various heights and distances and counted pixels, I consistently ended up with more pixels than I calculated. But when I used 84 as the HFOV, the numbers were almost dead on. I'm not sure if this came up in this forum before, but can anyone confirm that 84 degrees really is the diagonal FOV and I'm calculating something else incorrectly?
Is 84 FOV or HFOV?
- Thread starter kjwhitcomb
- Start date
Meta4
Premium Pilot
- Joined
- Jan 8, 2014
- Messages
- 15,355
- Reaction score
- 9,298
It's the diagonal angle.
Lens viewing angles are conventionally expressed as the diagonal angle.
DJI specs show the P4 pro lens having a FOV equivalent to a 35mm format lens of 24mm focal length.
Checking 35mm lens specs 24mm lenses show a diagonal view angle of 84°.
Thanks for the quick response.
My question is, has anyone actually measured/confirmed this? For example, I have a Jeep that is 189.8 inches long. Took a shot from 87 meters up. If the HFOV is only 73.3 degrees, then I would expect the Jeep to be about 237 pixels. Blowing it up and counting I count only 208. If I use 84 degrees, then the calculation is 207 pixels. I repeated the measurements at longer ranges with the gimbal pointed at the Jeep in the center and got similar results (taking screen shots of the controller to record distance and altitude), though the errors are less because the total number of pixels is less.
Do you know if anyone has measured and confirmed the FOV with actual ground truth?
Thanks
My question is, has anyone actually measured/confirmed this? For example, I have a Jeep that is 189.8 inches long. Took a shot from 87 meters up. If the HFOV is only 73.3 degrees, then I would expect the Jeep to be about 237 pixels. Blowing it up and counting I count only 208. If I use 84 degrees, then the calculation is 207 pixels. I repeated the measurements at longer ranges with the gimbal pointed at the Jeep in the center and got similar results (taking screen shots of the controller to record distance and altitude), though the errors are less because the total number of pixels is less.
Do you know if anyone has measured and confirmed the FOV with actual ground truth?
Thanks
Meta4
Premium Pilot
- Joined
- Jan 8, 2014
- Messages
- 15,355
- Reaction score
- 9,298
I haven't heard of anyone doing that.
I also haven't heard of anyone having any real reason to.
The reason would be to be able to estimate the size of objects from imagery taken from known altitudes. If I can accurately determine the IFOV, record the altitude at which an image is captured, and count the pixels in an object, then I'll know the approximate size of the object. This all starts with getting a good value for the IFOV.
