VBO POI's

[gplracerx]

When I try to import a .vbo file into Circuit Tools, there's always a problem with the trigger POI's. Out of curiosity, I plotted the POI gates and a run from an autocross this weekend. Here's the result:


The orientation of the gates is about 90 degrees off and the location of the center point isn't all that good either. I think the problem may be related to the RaceLogic convention that west longitude is positive instead of negative. If, when you plot, you use the second longitude location first and the first second while keeping the latitude order the same, the orientation is a lot better. That is for a point with coordinates x1,y1 and x2,y2, you plot them as x2,y1 and x1,y2:



Also, the POI's in HLT are at one end of the lines rather than in the middle.

[Harry]

This is probably like it should be. According to Racelogic, triggers are defined like:

The first word is the split point type (Start, Finish, Split).
The four coordinates are:
1. Longitude
2. Latitude
3. Longitude measured 10 samples in the past
4. Latitude measured 10 samples in the past
All coordinates are in minutes. North is positive, South is negative. West is positive, East is negative.
This defines a "straight" line along the circuit, all split points are saved using this format.
If you load such a split point definition into any software (e.g. Circuit Tools) it will generate a gate perpendicular to the line and positioned at the end of it. Gate width can usually be adjusted in software (I think 25m is the default).

- Harry

[gplracerx]

That explains the orientation of the start and finish and two of the splits, more or less. But if the other end of the line is ten samples in the past, shouldn't one end of the line always touch the course? I don't understand the orientation of the orange and purple lines, split 2 and 4 in order on the course.

[Harry]

LapTimer does not use the actual line driven as the end point, but the position you have set. This means the actual line does not necessarily touch the "trigger line end point". You will pass it left or right and this is why a corridor is necessary. Same for the 10 samples back. This is not a sample from an actual recording, but a back-extrapolation using either the trigger direction or the driving direction at the point the line has been passed (need to check in code). Looking into your charts, it looks like you had set a driving direction?

- Harry

[gplracerx]

Never mind. The real problem is that with directional POI's, which I'm using, HLT defaults to wide corridor, which, with the default setting. is way too wide for an autocross on a relatively small parking lot. That leads me to move the actual location of the POI off the course so it won't trigger at the wrong time. Then when the .vbo file is created, the locations are wrong. I've set the width of all corridors to 5m to solve that problem. I don't do track days, so I won't have to change it back. I may need to tweak that up a little, but it will probably still be less than 10m.

[Harry]

If someone gets it managed, it is you!