Here’s how the 3-d and 2-d views are different: in the 2-d view, you don’t have much freedom to move your head. Given this, if we don’t honor the horizon level, the 2-d view might somewhat unusable. Furthermore, honoring the horizon level in the 2-d view lets us provide a “2-d like” behavior in the 2-d view which is what users expect.

(The point of having the 3-d cockpit in 2-d view is _not_ to make the 2-d view into a 3-d one – that’s what the 3-d view is for. It is to allow authors who have invested a lot in 3-d and very little in 2-d to reuse their 3-d art assets.)

In the 3-d view you have 5 DOF..under that situation, having the “natural” center of the field of view be not the middle of the screen means unequal real-estate in all viewpoints.

In the end, I don’t believe that the horizon reference is a matter of “acf spec” in the 3-d view – there has never been a setting for it, and X-Plane’s legacy behavior (using the 2-d setting, sometimes producing absurd results) doesn’t strike me as a good idea.

Your changing views that both of them draw 3d panel (whatever view type there is) and look through the front glass.

Since 9.0 both of them drew the same view, now they don’t. For me it’s less then expected.

“the logic place only applies to 3-d cockpit view type, not whether the 3-d cockpit is drawn.”

So 3-d cockpit view type changed, 2d cockpit view type (even if 3 cockpit is being drawn here) didn’t. Hence to incoherence and shift when you switch between those view types (if I get it correctly).

Just open a plane with 3d pit only (no such plane in default X-Plane ones afaik) and repeatedly hit ctrl-o to switch 2d<->3d (in both cases 3d should be drawn but in 3d mode it will be shifted up). You should see what I mean.

If you observe different in beta 7, please report a bug.

The change is only for the “real” 3-d view, where we have 5 DOF and can thus adjust the relationship between the panel and horizon by moving our heads (which induces parallax).

This was not a case in previous betas or 9.2x

Are you saying that we should change the default pilot head position (because it is now necessary to change the head position to get the same ratio of sky to panel as before)?

If so, the necessary change is not a change in head _height_ (translation) but head angle (roation) – that is, you need to look down, not scroll down.

I would argue that the requirement to look down is _more_ correct…920 basically lets you see the entire glass display with peripheral vision, which is IMO unrealistic. In real life you’d have to look down (and if it isn’t at an angle, it will be sloped away from you) or move your head lower (and losing the ability to see the runway).

In other words, the geometry of the cockpit and view position should be based on real world values, not optimal viewing. It is up to the user to change the view dynamically to see the content desired (more sky, or more panel), and in virtually any real setup I expect the user not to be able to see everything at once.

cheers
ben

I don’t agree with your assertion.

– If the pilot has to move his head up (I think this is what you mean) to see over the dash board, that just means that a normal-height pilot’s head position (represented by the pilot viewpoint in Plane-Maker, which is really a starting viewpoint before the user moves his head) is simply not that high up relative to the dash board.

– Moving the horizon’s mapping to the screen has _absolutely no effect_ on the relative positioning of the dash board and the pilot’s head. If you set the pilot’s head position to have trouble seeing the runway without “sitting up a bit”, you will get the same results with the old horizon and new horizon.

cheers
Ben