Manfred Ruhmer is flying twice as far.

[philderbeast]

In task 9, HG Worlds 2019, Manfred Ruhmer is flying 199.875 kms in flare-timing but only 106.856 in FS. David Guiducci and Christian Tiefenbacher distances flown are much the same.

[kuaka]

I remember this in airscore. Manfred was just short of one of the turnpoints and went on to fly more of the task as if he had taken the turnpoint. His instrument obviously thought he had. Most likely the wrong tolerance in his instrument and flare timing. Maybe FAI sphere v ellipsoid issue

[philderbeast]

Thanks @kuaka. These are the extraction parameters I used:

ft-extract-input "$COMP" --give-fraction=0.0001 --give-distance=5 --haversines

[kuaka]

The competition was scored using wgs84. See local regulations here where it is specified https://airtribune.com/22nd-fai-world-hg-championship/info FAI sphere was obsoleted by 2019

…

On Sun, 22 Nov 2020, 23:59 Phil de Joux, ***@***.***> wrote: Thanks @kuaka <https://github.com/kuaka>. These are the extraction parameters I used: ft-extract-input "$COMP" --give-fraction=0.0001 --give-distance=5 --haversines — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#468 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AETWEH5WF3XB6DFFEOYFCCDSRDVJXANCNFSM4TXGIXNA> .

[philderbeast]

I had intentionally gone with haversines despite the comp rules. I still haven't quite pinned down why more recent comps take longer to score.

[kuaka]

Ok. If you use the wrong model you will get errors like this from time to time. Especially on large cylinders. Pilots have been caught out by making the same error with their instruments. Manfred possibly had the wrong setting here too. Even without turnpoint taken/not taken errors you cannot compare results with official as all distances will be slightly different.

…

On Mon, 23 Nov 2020, 08:11 Phil de Joux, ***@***.***> wrote: I had intentionally gone with haversines despite the comp rules. I still haven't quite pinned down why more recent comps take longer to score. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#468 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AETWEH64OEIWG4R7SPYB5BLSRFO4ZANCNFSM4TXGIXNA> .

[philderbeast]

The turnpoint tolerance for this comp is written to the *.fsdb at //FsScoreFormula@turnpoint_radius_tolerance and it is larger than I thought:

-- $ ft-extract-input "$COMP" --give-fraction=0.0001
++ $ ft-extract-input "$COMP" --give-fraction=0.001

<FsCompetition name="22nd FAI World Hang Gliding Class 1 Championship">
  <FsScoreFormula id="GAP2018" turnpoint_radius_tolerance="0.001" />

[philderbeast]

With the following, I get 105.488 km vs 106.856 km for reach for 117 vs 118 reach points, flaretiming vs FS. However for effort I then see 13.1 km vs 0.0 km and 13.6 vs 52.6 effort points.

ft-extract-input "$COMP" --give-fraction=0.001 --give-distance=5 --vincenty

Looks like the map projection is showing him making B91.

[kuaka]

airScore had the same issue with maps. The library to plot a circle of radius x was not allowing for the projection so there had to be a correction. The great circle distance between the waypoint centre and the optimised turnpoint (which, if using vincenty would be the turnpoint radius) was used to adjust the radius of the plotted circle.

FAI-CIVL/FAI-Airscore@eb20156

[philderbeast]

I logged the problem of drawing projections of turnpoint cylinders on the map as #486.