Possible unwanted behaviour of scale factor?

[Olly Betts]

On Wed, Apr 01, 2015 at 10:55:39PM +0100, Andrew Atkinson wrote:
> Just been processing data for a sump using a distox2 in a waterproof box
> (for lack of a better word)
> 
> So the box adds a bit on the back which means you have to adjust the
> zero, and the water mean that you have to scale the readings by about 0.752
> 
> So
> 
> *calibrate tape -0.05 0.752
> 
> However according to the manual and the quick leg I have checked
> 
> (magnetic bearing) = ((reading)-(compass zero err)) * (compass scale factor)
> 
> Which I take it is the same for the tape.
> 
> Now following survex principles for trying not to amend data before
> processing, and as I have measured the zero offset in meters the above
> does not work, which leads me to wonder if this is the correct behaviour
> for tape.

I think it's generally the right definition of the zero error - you'll
typically set up a reading which should be zero, and see what the
instrument says for that.  In the "end has fallen off the tape" case,
which is probably the most common tape zero error situation, you just
read the marking for where it now ends.  If you have a bonus extra
before zero (as some tapes CUCC bought a while ago did), then again you
can just read off the scale for that section (or if the extra section
isn't marked, double it over at zero and read it off against the marked
part of the tape).

If the tape had stretched, you'd still want to record the markings from
the tape, and have them corrected by the stretch (if the tape has
stretched, it's time to get a new tape as it probably hasn't stretched
evenly along the whole length, but if you're in the field somewhere
remote you may have to do the best with the equipment you have).

That's not very helpful in your situation though - measuring the zero
error with the device would be much more complicated than just using a
ruler, and likely to introduce as much error as it avoids (I guess the
zero reading is also affected by the beam travelling through the case
and any air between the lens and the inside of the case as well as the
additional distance the box adds behind the device).

How do you determine the 0.752 scale factor?

> Anyway something to be thought about
> 
> Possible ideas are if not units
> 
> real = ( reading - offset) * scale
> 
> If units eg *calibrate tape -0.05m 0.752
> 
> real = reading * scale - offset

Yeah, optional units here is probably the answer.  Internally we can
just adjust the offset by the scale factor so it only requires a change
to the *calibrate command handling.

Cheers,
    Olly