Mike wrote:
Is this also the source of the faux AH found in L8000?

Not quite:

http://git.xcsoar.org/cgit/mirror/LK...rawTRI.cpp#n18

The rendering code was rewritten to make it look more fancy (LK8000
uses only the calculated bank angle, but not the pitch angle). The
formula has never been changed in LK8000, it's still the old XCSoar 5
formula, which means it's still pretty unreliable, lacking real AHRS
input:

http://git.xcsoar.org/cgit/mirror/LK...tions.cpp#n513

The XCSoar 6 formula has been improved to consider acceleration sensor
input (if available):

http://git.xcsoar.org/cgit/master/xc...puter.cpp#n324

Not so much better, though. I'd say neither one is usable.

Max