I was under the impression that line calling is done in the following way.
Freeze time at the moment the base of the ball first touches the ground.
Take a "bird's eye view" of the ball, ie imagine looking at it from directly above. (In essence, project the 3D object to the 2D court surface.)
If any of the ball is intersecting the line, then the call is in.
In particular, _the base of the ball need not touch the line, as the 'back' of the ball may be 'overhanging' the line.
This seems natural enough: "is any of the ball intersecting the line?". In particular, the trajectory of the ball doesn't matter: if it's hit flat, almost horizontal, then the cross-section at the point of impact is exactly the same as if it were falling horizontally; one need only see the snapshot, not the whole trajectory.
Now, in the Fritz/Isner match yesterday, it seems that this is not the case.
The cross-section is clearly intersecting the line -- the hawk-eye image on the left is at different time to the video on the right.
Compare this, though, with a Federer/Nadal match of which most of us are aware (it was in Summer 2008...): https://www.youtube.com/watch?v=Y3Ufsrx5J-4. Again, at impact, the ball clearly 'overhangs' the line. However, at no point is the ball actually in contact with the line. This ball was called in by hawk-eye.
So it seems that there is inconsistency. Can anyone share any light on this?