Understanding in-game car coordinates

[jars121]

Hi all,

I'm trying to make sense of some AC telemetry, specifically the carCoordinates[3] packet, which provides the coordinates of the car in X, Y and Z axes. The Y axis isn't of interest for the time being, as I'm only trying to understand these coordinates in 2-dimensions.

I've just done a lap of Barcelona, and have made a scatter plot in Excel. I've then overlaid the AC Barcelona map (the map.png file in the C:\Program Files (x86)\Steam\steamapps\common\assettocorsa\content\tracks\ks_barcelona\layout_gp directory) as shown in the image below.

The telemetry data is shown in blue, the AC map.png is in white. Note that I've actually had to mirror the AC map.png in Excel; it appears that the telemetry data is inverted horizontally?

Can anyone make sense of the above? It looks like there's some scaling, offsetting, etc. required to convert the in-game carCoordinate values into meaningful real-world data. The map.ini file in the /data directory has SCALE_FACTOR, X_OFFSET, Z_OFFSET, etc. values, which I've played around with, but I think they're used for the in-game map overlay, rather than converting the coordinates themselves. I may be wrong about that though!

Any ideas?

 

[Stereo]

The map.ini converts into image coordinates from X and Z, so the offsets say where one of the corners of the image is (I believe top left?) It does also have a scale for the ingame display, though. Unlike cartesian graphs, image coordinates are X (left to right) and Y (top to bottom) so they do tend to get flipped

I'm not sure why it's rotated a bit. Maybe the lighting.ini SUN_HEADING_ANGLE? Your X and Y coordinates on the graph are 2x differently scaled though so no way it would fit.

 

[jars121]

The map.ini converts into image coordinates from X and Z, so the offsets say where one of the corners of the image is (I believe top left?) It does also have a scale for the ingame display, though. Unlike cartesian graphs, image coordinates are X (left to right) and Y (top to bottom) so they do tend to get flipped

I'm not sure why it's rotated a bit. Maybe the lighting.ini SUN_HEADING_ANGLE? Your X and Y coordinates on the graph are 2x differently scaled though so no way it would fit.

Thanks stereo, your input is much appreciated.

That's what has me confused, there seems to be some non-linear scaling going on, at the very least the scaling differs between the X and Z axes for some reason. I've seen some AC telemetry software which plots the car's position on a map of the track, so there must be some way to correlate the carCoordinate values with the track map.

 

[jars121]

Just wrapping this thread up, I've solved this issue. It looks like the in-game coordinates were totally fine, my Excel scatter plot on the other hand was not! The auto-scaled axes were causing the observed scaling/offset issues; when I manually set the axes in Excel the carCoordinate data lines up very nicely with the in-game-producted map.png image (the only exception being the horizontal mirroring as mentioned above).

 

[deadly_pants]

Hi, I'm trying to make the same thing, but my track plot is a little twisted. Does anyone know how to fix this? It supposed to be the vallelunga circuit with multiple laps

.

 

[Neilski]

Hi, I'm trying to make the same thing, but my track plot is a little twisted. Does anyone know how to fix this? It supposed to be the vallelunga circuit with multiple lapsView attachment 575356View attachment 575357.

Given the tiny min-max range on the horizontal axis I'm gonna guess you used the wrong pair of axes. If AC did what I'm gonna choose to call the "normal/expected" thing, Z would be up (well, vertical) and X/Y would be the horizontal plane. AC didn't do the normal/expected thing though

 

[deadly_pants]

Given the tiny min-max range on the horizontal axis I'm gonna guess you used the wrong pair of axes. If AC did what I'm gonna choose to call the "normal/expected" thing, Z would be up (well, vertical) and X/Y would be the horizontal plane. AC didn't do the normal/expected thing though

Thanks, I found the right coordinates to plot. Previously I used the CarCoordinates first 2 item but the solution was to use the TyreContactPoint.

 

[Neilski]

Thanks, I found the right coordinates to plot. Previously I used the CarCoordinates first 2 item but the solution was to use the TyreContactPoint.

Glad it's working for you, but the car coordinates should be fine too (provided you don't use the Y axis).