Steve Turner wrote:

> Using the "stick it out of the window of a moving car" method of
> calibration ain't as easy as it sounds.

  Hint: You need a separate driver!

> Second, automobile speedos are not made for great precision.

  Quite so; they don't even register below 20k/h (15mph).  I really
*must* build my "car computer".  These seem to have waned in popularity
now, except amongst rally users (maybe even there too?).

  An alternative comes to mind; a cyclist riding in a large circle on a
tether?  Still air and smooth tarmac required of course.  It turns out
that cycle computers appear to use algorithms, quite possibly table
reads, to do the period inversion and are not that accurate either.
Which stymies the other suggestion, that they can be used as a cheap
"car" computer.

> Third, the calibration method assumes zero ambient wind speed, which
> simply isn't realistic most of the time.

  I can tell you there's been a lot of it over here in the last few
days.  It is HOT!

  Cheers,
        Paul B.
(Who has in the past, "fox-hunted" single-handedly; i.e., spun the yagi
whilst driving and watching the meter!)