[Csantucci]

Attached you have the 2 modified runactivity exe files to replace within x.3915, that correct the problem.
The patch is operating when all these conditions are met:
- the FOV preset in the options is 45 degrees
- Cab2DStretch is set at 0
- the aspect ratio is bigger than 4:3
- a cabview is available.
The patch forces the horizontal FOV at 60 degrees and consequently recomputes the vertical FOV (which will be < 45 degrees) for the cab camera. Other cameras are unaffected.


Pls. check if the patch is OK and if it works correctly also changing cabs, changing cameras and so on.

14/8/17: Attached files deleted because patch committed in x.3917

[Borislav]

Ok, I am not sure we are on the same track here, but this patch does nothing on my computer.

If I use a cab with .cvf configured for SS on my laptop WS monitor, and leave the default value at 45, I get the aforementioned discrepancy.

I get the same discrepancy if I use a cab with .cvf configured for WS monitor in windowed SS mode, with 45 default value.

Again, the only way to make the same .cvf file look correct in BOTH cases is to change the default 45 value to its appropriate value, depending on whether I am using .cvf configured for SS on my laptop WS monitor, or .cvf configured for WS monitor in windowed SS mode.

[Csantucci]

This is what I get with an 800x600 window:


This is what I get with a 1280x720 window:


This is what I get with a 1600x900 full screen:


The three pictures are coherent referring to the relative position of the world.

[Borislav]

In all three cases the default value of Viewing Vertical Fov in display options is left at default 45?

[Csantucci]

Yes, and %cab 2d stretch is zero.

[Borislav]

I'll download today a fresh x.3915 version and try again. Perhaps it has got to do with my updated ORTS. I've had problem before with updated version not working properly.

[Borislav]

Ok!

I have managed to get it working the way I described above. HOWEVER, there is now a different "problem" - when I scroll the cab up and down, the vertical direction changes dramatically.

Picture 1 - scrolled all the way up:


Picture 2 - scrolled at around the middle - this is what the cab view should look like concerning direction and position:


Picture 3 - scrolled all the way down (notice how the nose of the cab is pointed way downward):

[Lindsayts]

Sorry for coming in to the discusion late, I am having great difficulty with my internet service, A couple of points...................

(1) One should NOT assume a particular field of view angle as this CAN be altered, I in fact I set mine much narrower as one gets a far better view of signals.

(2) The aspect ratio of the cabview as built should be kept unaltered, but should be scaled to suit so it fits the monitor.

(3) One should not assume the cabviews a particular aspect ratio, OR (up till now) has allowed any aspect ratio for the cabview, I do all mine 16 x 10 as it fits my 2560 x1600 monitor correctly and cab stretch is set to display this correctly. When 4 x 3 cabviews are used cab stretch is set to suite.

(4) Unfortunately there is NO correct way to display cabviews on todays monitors as a correct cabvview would require a monitor that was much taller then even 4 x 3 monitors and I personally find scrolling the cabview is NOT really the answer.

LIndsay

[Csantucci]

Borislav, on 13 August 2017 - 10:06 AM, said:

Ok!

Good that it works!

Borislav, on 13 August 2017 - 10:06 AM, said:

I have managed to get it working the way I described above. HOWEVER, there is now a different "problem" - when I scroll the cab up and down, the vertical direction changes dramatically.

As far as I know this is a wanted feature. Vertically scrolling the cabview has been considered as rotating the observer's head around the x axis, so a rotation has been applied to the external world.
I think this effect was there also before my patch, wasn't it?

Borislav, on 13 August 2017 - 10:06 AM, said:

Picture 3 - scrolled all the way down (notice how the nose of the cab is pointed way downward):
Open Rails 2017-08-13 08-01-06.jpg

Looking at the nose's position unfortunately pinpoints a quite unresolvable problem : the nose should be part of the external world, and therefore should move together with it. But this is impossible, because the nose is part of a 2D cab. The problem wouldn't exist in a 3D cab.

So, what do you suggest the world should do when vertically scrolling the cab?

[Borislav]

I see. Yes, you are right. This effect exists without the patch as well.

Is it possible to make the scrolling work like in a picture viewer? Meaning that you pan around the image, while the image remains stationery?