FSXA Coordinate Converter Problems

[JRobinson]

Hi Arno,

I made a groundpoly in gmax beginning with just 20 non-tiling 1024 .bmps, each one assigned and mapped to it's own separate plane object. I exported that as a model and placed it manually with Instant Scenery 2. I decompiled the placement .bgl and used the coordinates in the GP wizard to turn it into an FS2002 style GP. I saved the placement settings as an .mgp. Since then I added 50 more adjoining planes/.bmps and processed the exported model through the GP wizard using the same initial .mgp. That seems to have worked fine.

Later I made up another remotely located GP for another area on the airport. I also placed it with IS2 and then used the coords in the GP wizard as before. I also saved those settings as .mgp.

I wanted to merge the latter remotely located GP into the first project in gmax so that I could export it as a single groundpoly. I used the coordinate converter to calculate the offset. First I used the coordinates from the .mgp files to calculate the offset. That placed the remote polygons something like 75 meters from their intended location. Next I tried using the coordinates from the original XML placement files in the coordinate converter which resulted in a similar but slightly different position.



The coordinates are as follows:

GP1 (master GP, using this as reference point)

From the .mgp
48.730921559
-116.292871088
HDG: 348.3

From the XML:
48.7309215590358
-116.29287108779
HDG: 348.299560546875


GP2 (remotely located GP, using this as secondary point)

From the .mgp
48.7266374006867
-116.297295838594
HDG: 309.0

From the XML:
48.7265787273645
-116.297353059053
HDG: 308.995971679688


Using the .mgp coords I came up with an offset of:
X= -325.536
Y= -476.410


Using the .xml coords I came up with an offset of:
X= -329.746
Y= -482.934


I merged the polys into the first project and typed the geocentric offsets into the Move Transform type-in. Seems like it should have worked, or at least it should have aligned the secondary polys to within a fraction of a meter or so. It was close, it moved everything basically the right direction, but it seems the Y coordinate is mostly the problem. I did have different headings for each GP, I subtracted one from the other and rotated the second GP by that amount on the Z axis in gmax which also looked about right. Maybe the heading discrepancy is causing the problem somehow?

Any clue what I'm doing wrong?

much appreciated,
Jim

 

[arno]

Hi Jim,

I think the heading might indeed be the problem. Also don't take the geocentric coordinates in the coordinate converter, but the flat earth ones. That can also give a small difference.

It might work better if you already rotate your polygons in gmax, so that you can place them with heading 0. Else calculating the offset of the second group will be hard, since it will be rotated with the heading of the first group.

So I would try:

Rotate your first group by the heading.
Move your second group with the offset you calculated from the origin.
Rotate your second group by its heading around the old origin of it.

 

[JRobinson]

Very much appreciated Arno, I just gave it a try (using the geocentric offset so far since I had those written down) and it seems to have done the trick, the second poly came out very close. Closer than I expected actually, I could shuffle some shadow layers around on the textures and make it work. I'll experiment with the flat-earth coords in a bit and see if that dials things in a little closer.

Amazing tools, thanks a million!

Jim