Mangled 3d file

[Bill Gee]

Hi Graham -

The back-compass shots I left out of the data were done because they messed up 
Therion.  When Therion tries to average shots that are nearly north-south, it 
often comes up with due west!  The result is a seriously mangled map.  

Take readings of forward=359, backward=181.  What is the average?  It should 
be 0, but Therion comes up with something more like 270   (359 + 181) / 2.  
The only way I could get Therion to produce a correct map was to take out one 
of the two compass readings.

The first shot where things go wrong is from A10 to A11.  All three shots (A9, 
A10, A11) have both forward and backward compass data.  None of them are close 
to 0/360, so they should average correctly.  The splay shot from A4 to L1 is 
on the wrong side of the main survey line - and it also has both forward and 
backward compass readings.

There *IS* something about the backward compass readings that is getting 
cavern confused.  I added your two *data lines before and after A10-A11.  The 
problem moved to A11-A12.  Just for grins I changed the*data line at the top 
of the file to completely ignore the backward readings.  Voila!  It gave me a 
good plot.

Regards - Bill Gee


On Tuesday, March 03, 2015 20:43:46 Graham Mullan wrote:
> View by Error will not work as there are no fixed loops. Without wishing to
> re-jig all your data myself, I suggest that where you have legs in which you
> have left out the back compass reading that, instead of just omitting (or
> rather using a dash) for the offending reading that, before the leg you
> insert a new data line:
> 
> *data normal from to tape compass clino ignoreall
> 
> To apply to that leg and then, after that leg insert, once more:
> 
> *data normal from to tape compass clino backcompass backclino ignoreall
> 
> I suspect - and this would prove - that your data is being compromised by
> these false readings and that this has knock on effects along the file
> 
> What surprises me is that Therion does not produce the same problem, as this
> program uses the Survex engine to crunch centreline data.
> 
> Graham