• Which the release of FS2020 we see an explosition of activity on the forun and of course we are very happy to see this. But having all questions about FS2020 in one forum becomes a bit messy. So therefore we would like to ask you all to use the following guidelines when posting your questions:

    • Tag FS2020 specific questions with the MSFS2020 tag.
    • Questions about making 3D assets can be posted in the 3D asset design forum. Either post them in the subforum of the modelling tool you use or in the general forum if they are general.
    • Questions about aircraft design can be posted in the Aircraft design forum
    • Questions about airport design can be posted in the FS2020 airport design forum. Once airport development tools have been updated for FS2020 you can post tool speciifc questions in the subforums of those tools as well of course.
    • Questions about terrain design can be posted in the FS2020 terrain design forum.
    • Questions about SimConnect can be posted in the SimConnect forum.

    Any other question that is not specific to an aspect of development or tool can be posted in the General chat forum.

    By following these guidelines we make sure that the forums remain easy to read for everybody and also that the right people can find your post to answer it.

Point placement accuracy

Darren

Darren

Messages
25
Country
unitedkingdom
Hi Arno

I've found that AGN library objects placed from very accurate *.shp point data using scenproc are being positioned in the simulator offset slightly from their expected positions. The offset angle will vary considerably and ranges in distance between ok and about 10-20m or so from the right spot. I'm pretty sure there must be some small error in the way scenproc is calculating the shp to agn position or maybe that FSX is adding the offset at runtime?
 
Hi

Yes I know Chris. I've not added an offset in the script nor do I want to in this case.
What I'm trying to say is, that the *.shp points which fit the imagery perfectly in Global mapper ( WGS84) will result in objects being placed away from that accurate location when they are generated through scenproc.
 
Hi

Sceneproc creates the points accurately. If I export the points to shape file and view in Global mapper again, then they still fit the imagery perfectly. Doesn't matter if the points are imported as "points" or if they are created with PLACEPOINTSALNGLINE script. That's not the problem.
It's only as they are converted to AGN placed objects that a mis-positioning issue is introduced, which is why it leads me to believe sceneproc is calculating something incorrectly when writing the *.AGN.
The mis-positioning / offset ( whatever you want to call it) varies a bit. I think the amount it varies may be related to it's x,y location within the tile, but I'm not sure about that exactly.
 
Hi George

Thanks...unfortunately though it can be quite a bit more than +/- 2m. At it's worst it can be more like 20m or more away from it's expected location.
 
Hi,

I don't think the accuracy is only limited to 2 meters, if you look at the way the autogen format works it is more accurate then that. Especially for library objects, for polygons you still have the rasterization with its own resolution.

Let me explain what scenproc does when creating library objects. Hopefully that brings us closer to finding the solution.

The autogen format does not store the objects as points, but it stores them as polygons. Just as you would draw them in annotator. So in scenproc I know make a polygon around the point. I use the length and width as provided in the step for that.

I think to stay accurate these need to match with the length and width in the default xml file. But I'm not sure what happens when the length and width of the actual object are different. Not sure if they are placed around the center or one corner in that case.

Also not sure what happens if you have an autogen class with different sized objects.

I'll see if I can do some more experimenting on this.
 
But I'm not sure what happens when the length and width of the actual object are different.
Click to expand...

That sounds very possible Arno. The object I've noticed the issue with mostly is with a railway gantry spaced along a railway line. The gantry object dimensions are 1m x 25m, so if your idea above were true the big difference in width / length with this particular object would certainly exaggerate the issue...as indeed it appears to have done. Could your idea still account for why the offset varies so much, from ok to about 20m or more adrift though?

So in theory then, I can test this and center the object simply by giving it an equal x,y dimension in the cfg file. I'll try this evening and let you know the outcome.
 
Last edited:
Hi Darren,

Did you also provide the size as 1x25 meter in the scenproc step that places the library objects?

I still plan to do some tests what happens when the polygon size is different from the object size. Maybe that explains the offsets.

I don't think the problem is that the length and width are very different, but if the polygon has a different size from the object I think that could be the problem.
 
Hi,

I just did a test where I placed some autogen library objects and then imported the AGN files again and converted them back to SHP. I do indeed see a constant offset between the input points and the resulting polygons of the library objects. But for me it seems there is always a similar offset, not a variable one.

I have to check if this is a bug in the importing the AGN data or if it is a bug in creating the AGN library objects. But for sure there is something going on here...
 
Hi

Did you also provide the size as 1x25 meter in the scenproc step that places the library objects?
Click to expand...

Originally yes I did, both processing step and default.xml dimensions for the object matched... but... while testing last night I changed the dimensions to 2x2m squared and then ran sceneproc to generate the AGN. Only after creating the AGN did I think to change the dimensions to 2x2m in the default xml too. I started flightsim expecting to see the object centered on the points, but instead found that they were still offset. So I then ran the step again without making any changes and found that this time the points were centered correctly.
So it would seem that sceneproc relies entirely on the dimensions it finds in the default.xml file, and is ignoring those that we give it in the processing step.

But for me it seems there is always a similar offset, not a variable one.
Click to expand...

With my railway gantry object the offset would vary as the track curved and meanders across the landscape. Were your test points all lined up in a straight line Arno?
 
Hi,

Scenproc doesn't read the length abd width from the default.xml file at all. It can even run without it. It only takes the dimensions from the step itself.

My points were placed along a curved road.
 
Hi,

I can confirm there is a bug in scenProc indeed when it comes to placing library objects.

Attached is a screenshot from Annotator. I have created 4 library objects from the same point feature, just with a different size provided in the step. I would expect them to be all centered on the point feature, but as you can see the bigger the size, the further away it gets.

So I'm going to debug the code now that makes the polygon...
 

Attachments

  • Image2014-04-23 2122.50.833.jpg
    Image2014-04-23 2122.50.833.jpg
    266.2 KB · Views: 478
One step closer, the polygons now all share the same center position. But the size of the polygons is still different, so there must be another bug in there. I'll continue searching tomorrow...
 

Attachments

  • Image2014-04-23 2230.57.538.jpg
    Image2014-04-23 2230.57.538.jpg
    320.6 KB · Views: 334
Hi Darren,

I think I have fixed it now. Can you check with the release of tomorrow to see if it also works for you?

The offset you reported in the heading of the objects is still there (I think). I have an idea what is wrong there, but I still need to debug that.
 
You must log in or register to reply here.
Back
Top