DistoX2 Calibration issue, Solution (had to do more than just leave the house*)

[Spike]

I thought distance didn't come into the calibration ?

Distance doesn't, however if your four orientation shots vary across a 1cm target 10m away, the max angle error between them at the disto (~0.05o) is much smaller than if the 1cm target is 1m away (~0.5o) thus helping the calibration.

If your jig fixes this in place though, so you're not targeting the laser at a particular point each time, I don't think it would cause a problem - at least not in the way I'm envisaging your jig (and the one I indent to build at some point)

HTH, I've rewritten it three times to try and be clearer - not sure it's worked...

 

[maxf]

But i'm more likely to wobbling the laser around quite a bit over a larger distance and not actually targeting the same spot each time...

I had made a another thread showing my jig, the photos didn't show up but there is a link to the album

http://ukcaving.com/board/index.php?topic=19310.0

 

[Spike]

I get what you're saying, but at least with a long shot you can tell when you've missed and therefore got a dodgy angle, and can redo the measurement. If you're targeting a closer spot of the same size, you could get the same error by hitting both sides of the same target without knowing about it.

I've got a plan for my own jig - I might try and put some of it together this weekend and if I do I'll add details to your other thread.

 

[cavermark]

Isn't it better to get practiced at calibrating by hand (no jiggety ;-) ) for the times when you need to recalibrate in a cave (or on an expedition etc.)?

 

[wookey]

Brendan, if you think that PocketTopo gave meaningfully different calibration results in this case to TopoDroid, then do tell Marco about this thread, as I'm sure he'd be interested. Topodroid nominally implements exactly the same calibration algorithm as PocketTopo so it really shouldn't matter.

 

[Hall2501]

I'm not sure - haven't properly gotten to the bottom of this yet. I managed to get the disto calibrated on expo using topo droid with you. I'll try calibrating a different disto next week & if I get the same problem. It's just strange how much the calibration readings differed so much between topodroid and pocket topo :S maybe it was something else...

Don't  worry my calibration crusade isn't over yet

 

[Hall2501]

Brendan, if you think that PocketTopo gave meaningfully different calibration results in this case to TopoDroid, then do tell Marco about this thread, as I'm sure he'd be interested. Topodroid nominally implements exactly the same calibration algorithm as PocketTopo so it really shouldn't matter.

Yeah, I'll try doing a calibration with another disto with topodroid and see if it works. It's just strange how much the compass readings differed in the first topodroid calibration. Might have been caused by something else completely.

We downloaded each set of 4 after every "roll" so each set of 4 came from the readings with the disto pointed in the same direction.

But with results like this:

Comp. Clino

65.14, -49.11
30.30, -44.28
138.19, -45.46
97.85, -49.20

instead of this:

Comp. Clino

166.2  -11.4
166.7  -11.1
167.1  -11.3
166.3  -11.3 

I'm not sure what's happend, can't just of been me not holding the disto steadily enough :S

 

[Bob Mehew]

Was your Disto completely uncalibrated before you started?  The more I ponder on 'calibration' the more I wonder what is going on.  I am far from understanding the detail of the maths underlying calibration (can any one help?) but very simply put if I have it correct, the Disto reads a magnetic sensor for compass info, a gravity sensor for clino info and a laser for distance in taking a 'reading'.  The compass and clino sensors returns 3 values each, so you get 2 * 3 values plus one distance value from the laser, 7 in all.  The 6 compass and clino sensors values are transformed into the 2 readings of compass & clino degree by some extremely clever mathematical equations which contain a range of constants.  These constants are derived from the 'calibration' process.  If your Disto was not calibrated or badly calibrated, then the values you found could be 'reasonable'. 

On the other hand I may be talking a load of crap :-[

 

[wookey]

Bob, a couple of papers have been published on the calibration process - it's quite an interesting problem due to having a lot of variables and a shortage of constraints:
http://utter.chaos.org.uk/survex/cp/CP37/CP37.pdf, page 5

And Beat's earlier article from the CREG journal: http://paperless.bheeb.ch/download/Calibration.pdf

 

[wookey]

Hmm, no edit function? Remove the comma on the Compass Points link to get a working URL to Phil's article.

 

[Bob Mehew]

Thanks Wookey, the Compass Points article by Phil is more readable and gives a few clues as to what is going on.  Beet takes that and moves on into such complexity that I am left almost speechless.  What still remains outwith my understanding is how one can take a set of 14 sets of 4 readings where the 4 readings are linked by requiring them to be precisely orientated along a given line around which one 'rolls the disto' but the 14 sets have no need for linking in a precise manner, only that they roughly point in directions of the 8 corners and 6 centers of faces of a cube.  The java coding link was not very clear to me, so I am left wandering aimlessly.  Caver.adam's jig seems a good starting point though I would like to know why.