Re: FS4 coordinate system

[astewart]

The FS coordinate system is quite in tune with most of its other software
features, i.e totally weird.

The origo of the system can be found in the centre of the (then) SubLogic
office building in Champaign, IL. The x-axis runs NS, the y-axis EW.

One grid units equals 1/10 of a US mile (628 ft, I think) and remains the same
throughout the grid system.

The grid is overlaid on the official US ICAO sectionals, which use a Lambeth
conformal polyconic projection. That projection always renders meridians as
straight lines, meaning that one sectional can be joined to another along the
edge meridian, _provided_ the projection parallells remain the same - which
they do for Continental USA, as far as I can recall.



Are you therefore saying, if I had a map (chart) sectioned using the ICAO
system and units, and I would pick a coordinate from this chart and enter it in
the location section in FS4, my screen would correspond accordingly?

This serves me very little purpose. If I wanted to use an official FAA
sectional from my local area (or even one for Chicago) to practice my IFR
flying, the latitude and longitude markings on that sectional would be useless.
I could not enter that info into FS4 to place me at a desired point.

Why is it someone at SubLogic could not include an algorithm that could at
least do an apporoximate conversion from an entered latitude/longitude to the
ICAO units? Am I missing something here?

Andre

[selberga]

The FS coordinate system is quite in tune with most of its other software
features, i.e totally weird.

The origo of the system can be found in the centre of the (then) SubLogic
office building in Champaign, IL. The x-axis runs NS, the y-axis EW.

One grid units equals 1/10 of a US mile (628 ft, I think) and remains the same
throughout the grid system.

The grid is overlaid on the official US ICAO sectionals, which use a Lambeth
conformal polyconic projection. That projection always renders meridians as
straight lines, meaning that one sectional can be joined to another along the
edge meridian, _provided_ the projection parallells remain the same - which
they do for Continental USA, as far as I can recall.



Are you therefore saying, if I had a map (chart) sectioned using the ICAO
system and units, and I would pick a coordinate from this chart and enter it in


the location section in FS4, my screen would correspond accordingly?

This serves me very little purpose. If I wanted to use an official FAA
sectional from my local area (or even one for Chicago) to practice my IFR
flying, the latitude and longitude markings on that sectional would be useless.


I could not enter that info into FS4 to place me at a desired point.

Why is it someone at SubLogic could not include an algorithm that could at
least do an apporoximate conversion from an entered latitude/longitude to the
ICAO units? Am I missing something here?

Andre


To start at the bottom: no, you are not missing anything at all.

It is one of the many mysteries surrounding the internal structure of MSFS why
(a) they did not pick something valid for the whole earth and to be found on
maps, for example UTM or Georef, and (b) why they did not at least try to turn
out an algorithm valid for Continental USA - it might be too difficult to deal
with coordination between maps for larger areas.

Put very simply, what SubLogic did - probably in reality as well - was to paste
together all 1:500,000 ICAO sheets covering CONUS on a suiable sheet of
something or on a wall, then drew a line which was geographically N-S and ran
through Champaign, IL, and as step three drew a line perpendicular to the N-S
line. That system was used as the starting point for a regular 528-ft grid
covering CONUS.

It means that the parallells are all curved and towards the edges - say, at the
same longitude as Seattle - you find that the E-W line, which ran through 0,0
at Champaign, might actually be found somewhere in Bahia de California (a
friend of mine, Professor of Geography, says so, but I have not checked it
myself).

The entire system was based on paper maps: there is no geodetic base for the
system - nor is there any such thing for the coordinates in FS5.x and later,
for that matter.

If you declare a square in a .bgl scenery you state (1) lat/long lower left,
(2) lat/long upper left and (3) distance between the points. If there existed a
geodetic datum, that information would be superflous - just give us the
lat/long position of two diagonally opposite corners, and it is possible to
compute the distance with a precision of tenths of millimeters, if you have any
use for it.

It has been written, that the FS5.x and later Earth looks a bit like a wine
cask (barrique) - it is made up of 36 polygons, those around the equator
vertical, the other ones slopeing and tapering, each strip 30 deg wide, and the
barrel finally covered by a polygon top and bottom.

I do not know the exact answer to the question about the shape of the Earth;
what I know is that you cannot transform real-world data to the FS projection
planes without incurring a penalty in the shape of serious distortion at
latitudes >50 deg or thereabouts, for want of a proper geodetic datum and a
usable shape.


Lars Strand