If you extrapolate 45 deg to the vertical it is not half way up the y-axis. (you can even see this in the lefthand side diagram 23:00 ). So, I think the tilts are labelled incorrectly. but its not clear to me. Perhaps the curvatures of the 5pt set-up compensate for the linear projection of the radius?
You can notice that on left diagram curved line represents the line between Z and Y points. If I got right, you mean that vertical line from eyes of observer to Y point is a "picture plan", but it's not. This vertical line shows direction of eye/camera. My concept of estimating vertical camera rotation works, you can test it in this app for example: perspectivetools.com/6-point-curvilinear-perspective
Thanks!
If you extrapolate 45 deg to the vertical it is not half way up the y-axis. (you can even see this in the lefthand side diagram 23:00 ). So, I think the tilts are labelled incorrectly. but its not clear to me. Perhaps the curvatures of the 5pt set-up compensate for the linear projection of the radius?
You can notice that on left diagram curved line represents the line between Z and Y points. If I got right, you mean that vertical line from eyes of observer to Y point is a "picture plan", but it's not. This vertical line shows direction of eye/camera. My concept of estimating vertical camera rotation works, you can test it in this app for example: perspectivetools.com/6-point-curvilinear-perspective