I agree, and that’s the sense I took them in. There’s also what I’d call “measured FOV” where reviewers have measured what FOV they’re actually getting looking through the lens.
I suspect that “advertised FOV” is generally based on the panels alone from an assumed ideal position of eyeballs. It ignores the limits of the lenses and any other limitations. I don’t see any other way that a higher “diagonal” number can be claimed. This is bringing a measuring standard from TVs into VR. But it really makes no sense in VR because lenses are not rectangle shaped, and so the widest dimension of what can actually be seen is really the horizontal, not the diagonal as in TVs.
Still, this is what all companies seem to use as the basis for their advertised FOVs as far as I can tell. Whether or not any particular company agrees with measuring this way (which is clearly BS, especially the diagonal number), they have to fall in line, or they’ll be advising a relatively smaller FOV than their competitors when the device really has the same or better FOV.
The edges and especially the corners of the panels can not be seen because they’re beyond the edges of the lenses which will be roughly oval shaped. The area that can be seen is a cutout from the rectangular panels. What percentage of the panel area is able to actually be viewed varies between VR headsets. But they tend to all lose roughly around the same percentage. I think this matches more or less what you’re calling the “reported” number. Since that’s derived from the geometry of what’s actually being rendered, and it necessarily needs to match what pixels are potentially actually viewable to avoid wasting computing resources. In effect, it’s compelled to be honest.
And yet “measured” FOV can still be a different number than this due to various other factors. Particularly that the user’s IPD, how the VR headset sits on their face, and the positions of their eyeballs varies. There can be other factors that come into play, too. For instance, I think on the widest IPD setting on the Quest 2, the outer edges of the lenses end up going past the limits of the panel. So the edge of the panel, rather than the edge of the lens becomes the limiting factor in that specific case (not totally sure about this). Anyway, these kinds of things are why FOV measurements between different reviewers vary.
The upshot is that “advertised FOV” is equal to or larger than “reported FOV” which is equal to or larger than “measured FOV”. And it’s important to make sure you’re comparing the same measurement between VR headsets. If you compared measured FOV on one VR headset vs advertised FOV on another, you’re going to get a misleading result.
That Varjo was apparently able to increase their “reported FOV” with a software/firmware update is an unusual case. You wouldn’t expect that to be possible since FOV is normally fixed by the physical lens design. But this implies that the Aero was not previously making full use of the extent of its lenses. And potentially might still not be.
I’m particularly interested in this and would like to see the measured FOV reevaluated against other VR headsets again after the update. Since the previous unusually small FOV was a major factor in my decision to not purchase the Aero. This update could potentially change my mind.