Autogen view distance (2) - Lake Pontchartrain

[edgarone 0]

Hello,in relation to the Pontchartrain-bridge question in another thread I came accross a strange observation: The distance to the first point in a VTP-bridge object seems to determine the display of this object.Background: This bridge across Lake Pontchartrain is quite long. In the FS world it crosses actually from one cell into another one. Its coordinates are even larger then the "safety zone" of 0-4079 or 12141-16320. So it is not possible to reproduced it with one single VTP line from the northern shore to the south. But you can build it with 2 bridge lines, starting at the shore and meeting each other at the cellborder in the middle of the lake.(It has the drawback, that both bridgeforms may differ, because we have no true influence on the choosen model - see the screenshot.)Now I made the following observation: If I approach the (southern) bridge from the southern end (actually the start of the VTP line), everything is displayed well. When I approach the (southern) bridge from the north (maybe starting a the meeting point at the cellborder) - nothing is displayed at all. When I slew south, the whole bridge springs into life at a certain point, when I am close enough to the southern end.It does not help to add more points to the bridge line. It just looks like the concatenation of several bridges.Is there a way to circumvent this behaviour?Cheers, EdgarAttached are 3 files: a map and view at both bridges at the cellborder. The Zip-file contains the bridge bgl + source and an exclude bgl for the default bridge.

[LuisFelizTirado 33]

Fascinating results, Edgar. Your images came out as text files, however. Please put them up again, as it would be nice to see them.Could this be something to do with the data stream? I seem to remember that the order in which the points are defined has great influence, and this might explain what you are seeing. Have you compared your bridge with the default? That should be an interesting exercise.This problem has some very interesting implications, since that bridge, being so long, may have visibility problems anyway. Just how do we define V1 and V2 nowadays?Best regards.Luis

[Guest luissa]

>It does not help to add more points to the bridge line. It>just looks like the concatenation of several bridges.Hi Edgar,About adding more points - is the visibility the same? As per your observations it looks as if the "reference point" in VTP bridges is the first point. So dividing the bridge in several 2-point segments could be a solution. I am just guessing.Kind regards, Luis

[rhumbaflappy 113]

Hi Edgar.Each component of bridges are actually library objects. You should be able to place bridge segments by XML code.Another possibility would be to make a gmax object. You could even use the library segments if you were willing to experiment with the measurements and use the attach tool. An added bonus there is that it could then have a platform(s) made, and the surface would be landable.Dick

[arno 1]

>This problem has some very interesting implications, since>that bridge, being so long, may have visibility problems>anyway. Just how do we define V1 and V2 nowadays?For mesh elements the v1 and v2 are done automatically. In the XML code to place objects the v1 is also done automatically by the scenery engine and the v2 (bounding box) is defined in the MDL object itself.

[arno 1]

Yes, I have done some testing with placing these library objects.About the visibility, when the bridge comes into view does it appear all at once or does each segment pop-up one by one?EDIT: Can someone give me an airport close to the lake? My US topography is not good enough to find it :).

[edgarone 0]

The bridge appears all at once.The same is true for the default bridge in the north. Here is the northern end the weak end. You can't even see its northern end in the FS, because it simple disappears (at once), when you slew north.The two bridges of the sample are similar to the default bridge, i.e. the common z-shaped form (2 major points and 2 hooks).As I said, adding more points in the VTP line does not help with the visibilty issue - the whole bridge is displayed or not. However the segments looks like connected single bridges (the ramps are mangled).Richards idea might work, because all bridge segments are defined and placed by themself. However it seems quite a challenge to form a 24-miles long bridge :)Cheers, Edgar

[edgarone 0]

Hi Luis,I misread your answer. Your suggestion might help too: using a string of smaller bridges. But there stays the issue of the ramps - there is already one issue at the cellborder, where both bridges meet each other. And then the problem of misaligned bridge shapes...Best regards, Edgar

[arno 1]

>Richards idea might work, because all bridge segments are>defined and placed by themself. However it seems quite a>challenge to form a 24-miles long bridge :)When I tested it, I made a API macro for FreeSC that did the calculations for you. So you just entered the length and the macro went through some loops to put all the segments together.Something like that should also be possible when using XML code.

[LuisFelizTirado 33]

Absolutely, Arno. What I mean is that I liked having control over these variables and while I can understand that the FS Team prefers to replace this with a LOD model scheme, it will still take me a while to get used to it. And most of my models are very simple low-polygon count shapes anyway; they do not really merit LOD models just so that I can determine when they are displayed.Best regards.Luis

[arno 1]

Hi Luis,I fully agree with you here. I also like to have full control over such things myself.But as a lot of people forget to set the v1 and especially the v2, I can understand the move away from it.I am also still learning how to make optimal use of LOD models and for simple objects I just add a distance check to the MDL, so that it works like the old v1 again :).

[Guest GCBarni]

When I created Confederation Bridge, which spans between New Brunswick and Prince Edward Island and is somewhere around 13 kilometers in length, I used individual sections just like the way the 'real' one was constructed. These are *.api objects and assembled / linked together using AFW.Now this bridge is not as long as the Pontchartrain crossing, but it is considerable and suffers the same affects being discussed in this thread. No matter which direction one approaches, the sections come into view as one gets closer; in other words, the complete bridge is not visible until the furthest point is within the sim's range. Each segment of the bridge behaves as a single object, because they indeed are, and is subject to the same scenery generation rules, so 'pieces' become visible as one crosses the strait. If you approach from 90 degrees to the crossing, say the middle of the span, these centre parts come into view first, and then as you close, adjacent parts also display towards both shores until eventually the complete bridge is visible. Note that range is also dependant upon altitude....the higher up you are, the closer you have to be to the objects before they will display, which is normal.I have the maximum allowable range set within AFW and no V=2 value.If interested, look in the library for confedbrg.zip; be sure to get the applicable one for the version of the sim. Begin a flight towards Charlottetown from Moncton, or vice versa and the Bridge should come into view.

[Guest luissa]

Hi Edgar,I understand what you are telling. Each 2-point segment would be translated to a kind of sequence:- starting part- central part- ending partI remember that the default object library has several bridges sections which can be called by its own ID. So you could chose 2 of them for the end parts and use several for the central parts. That, of course, would not be "trully VTP bridges", but it might work.Regards, Luis