Dave,

Looks like you're on to something. Couple of points:

- I started with a "raw" .CAI file and tested two conversion cycles;
one using Conv-cam and one using he utility that comes from Cambridge.
- The files produced are different over and above just the LRecord you
describe. The CAI utility produces a record that's about 1kb larger
and it's clear that the L-Record format is different.
- When you run the CAI2IGC conversion the output is different as well.

- Just modifying the LRecord (comment record) probably isn't the
answer, even though it shouldn't corrupt the hashed GRecord validation
step.

Looking back at the instructions from the OLC, it seems pretty clear
that it's critical to follow the conversion sequence exactly as
described; ie. start with the raw CAI file only - run the Conv-cam
utility - then use the .IGC file output from that step as the input to
the CAI2IGC step. Where I think people are getting tripped up is that
they get both the .CAI file and the .IGC file when downloading using
GlideNavigtor/PocketNav. They use the .IGC file from this step as the
input to the CAI2IGC conversion.

There are now dozens of files out there on the OLC that are red
flagged, and I haven't yet heard back from the OLC guys after emailing
them; I'm sure they're reasonably busy right now!