"Roger Long" schreef in bericht
...
But, and Larry too, he said he was using a reference compass. Since he is
not a sailor, I doubt that it is a gyro compass.

Probably metal on his person. The reference compass may not be a
reference
unless it is a pretty good one. It wouldn't surprise me either for the
little compass in the binoculars to be off by that amount either.

I wonder how far away the reference compass is while he is making these
checks.

It sure is a lot easier to get a good magnetic reference, with proper
variation information, since GPS.

--
Roger Long

As Robert says I used a reference compass.

A RECTA DP10 with a prism optical system where you set the dial to N=0°.
Then you bring the compass to eye level and you take the sighting on the
object along the red line of the marking and read off the bearing (against
magnetic north).
The magnetic declination for my region (Denderleeuw - Belgium) is 0° 27'
West. Less then 0.5° (according to www.ngdc.noaa.gov/seg/).

I have an observation point in the open fields from where I can see 5
church towers with known coordinates. The coordinates of my observation
point are also known (measured with a GARMIN GPSMAP 76CSx)

You can now calculate the bearings from the observation point to the
different towers.
The results were then double checked on a topo map of my region.
Then I measured the different angles with the RECTA DP10 and the STEINER
COMMANDER (see table)

TOWER CALCULATED RECTA STEINER
----------------------------------------------------------
1 357.5° 358° 360.5°
2 89.6° 89.5° 96.5°
3 131.6° 131.5° 137.5°
4 187.2° 187° 187°
5 270.5° 271° 268°
----------------------------------------------------------

Notice that the difference between the calculated values and the RECTA
readings are within 0.5°.
The STEINER readings are to my astonishment is much further off (except for
tower 4).
To the E and SE the differences are positive and to the W negative.

Major disturbances of the magnetic field would affect both compasses I
suppose.
The readings of the RECTA are close enough to rule these out.

These measurements are done over and over and by different persons with
almost the same results (within 0.5°).