Shetland Attack Pony-True Pulse-Suuntos-D3 shot comparisons

[cavescom]

Last weekend Aaron Bird and I got together to shoot 12 survey shots
using...
1) Shetland Attack Pony(digital compass/inclo -
http://shetlandattackpony.co.uk
2) True Pulse 360 - http://www.lasertech.com/360/
3) Traditional Suuntos sighting instruments
4) New Disto D3(discount code is cavers) -
http://www.laserstreet.com/leica-disto-d3.htm

Aaron and I both shot the different shots independently and wrote the
results down. Here is a link to those results
http://www.cavediggers.com/SurveySidebySideSummary.pdf
NOTE: there is also a link to the .xls file that contains the raw
data at the top of this pdf file.

OTHER NOTES: Aaron and I had originally tried this a week earlier
but quickly realized the True Pulse and SAP(Shetland Attack Pony)
needed calibration. The Shetland was shipped from overseas so most
likely got knocked out of calibration on the journey to the U.S.
The SAP(Shetland Attack Pony) however wasn't so easy to calibrate
i.e. the shot process is very simple but the software had
installation issues. I never did get the software to complete the
calibration completely, but instead had to send the raw.txt file to
Phil at SAP. Phil then generated a calibration file and sent that
file for me to load. Hopefully the software issue will be resolved
soon because the SAP is an absolutely fantastic device thus far to
use and worth every penny of the $600. In fact my goal is to have
another unit by year end for backsights.
**Phil just emailed and said a new version is up and running so I will test
this weekend and post if it is successful in running the entire calibration routine.

SAP(Shetland Attack Pony) usage -- Simply hold the device at any
angle...point the laser to the next station....hold the button down
for 1.5 second(this is adjustable via software setting) and read the
results. An extra feature is the recording/saving of a leg # on the
device. This leg# is written down in the survey book with that shot
so that later any blunders in reading the device or in the recording
of the measurement can be discovered!
The SAP is without a doubt the single greatest survey device I have
used and will eliminate many common survey errors.

Suuntos -- Both Aaron and I discovered how easy it was to read these
devices wrong.....maybe just once or twice in a series of shots but
it was enough for us to recognize the benefits of a device like the
SAP or True Pulse.

Disto D3 -- An easy to use...ultra small device that delivers what it
is designed to do....i.e. distance and inclo!

True Pulse 360 -- Easy device to use but relegated for use above
ground on longer shots such as a surface survey. Aaron will publish
more on this device.

My advice....get an SAP you won't be disppointed then add an extra
column in the survey book for Leg# and go survey.....the data crew
will become so much more efficient that they will have more than
enough time to draw cross sections at each station. This in turn
frees up the sketcher so that the labor is more balanced and the
survey moves forward at an even pace for all involved.

Additional ICDS article: http://incavedigitalsurvey.com/board/Ya ... 1205624025

Mark Passerby, InCaveDigitalSurvey.com

[Ralph E. Powers]

Well that may be good n all but finding a harness big enough to get that pony down the rope... I dunno.

[George Dasher]

You can always hang the pony from a cow's tail!

[barcelonacvr]

Mark: Once again,thank you for the massive amounts of data and effort you share freely on a constant basis!

I am going to buy one of these units.I assist on some surveys that need extreme accuracy (conserving lands through ANSI etc)and some scientists were wondering if the unit can be brought down to a half a degree accuracy.Do you think this is something that would be possible?For my own surveys one degree is fine as I doubt human eyes (or at least mine) will do any better with Suunto's etc.

[John Lovaas]

Mark-

I was wondering- is there any chance that you do a few more shots to get a loop out of that survey test? I entered some of the data, and it would be interesting to see how the different readers and instruments close the loop.

It would also be interesting to run the SAC(and heck, your Suuntos as well) on a "known" survey loop that was set up with "Big Boy" tools. I was reading the calibration instructions for the SAC, and I see you need a compass and clino anyway to set up the thing. It also mentioned the calibration is invalidated when the battery is shaken loose- does that mean the device has to be recalibrated when you change the batteries?

I entered the Suunto and SAC results as individual surveys all starting at your "Station 1" and I generated a plan and profile view. Your Suunto and SAC survey lines are pretty darned close!

jl

[Jeff Bartlett]

even without having the equipment to set up a real compass course, loop closure error is definitely going to be the most useful way to compare the performance of a typical Suunto compass/clino with the SAP.

[John Lovaas]

even without having the equipment to set up a real compass course, loop closure error is definitely going to be the most useful way to compare the performance of a typical Suunto compass/clino with the SAP.

Hi Jeff- I agree completely(off topic- I'm glad the Wunderalls fit, BTW!)

I guess my thing about using a "Big Boy" survey course(produced with a total station or other high precision/accuracy instruments) is that it lets you know just how close(or far off) your instruments(and readers) are to truth. I can only speak to data from two survey courses, but I know one course had a +/- 3 degree range in what folks with Suuntos were getting on compass and clino. The readings at either end of the range were instrument-based; anyone could read the instrument and get the same crappy numbers ;-)

The kicker would be to set up a compass course that minimizes human input or interaction- a giant wooden(no, it should be a dimensionally stable composite) jig that the instrument was clamped upon, for example.

[Bob Thrun]

I entered the Suunto and SAC results as individual surveys all starting at your "Station 1" and I generated a plan and profile view. Your Suunto and SAC survey lines are pretty darned close!

The plan view plot shows strong evidence of bias error. mp2 reads consistently high and ab2 reads consistently low.

[Bob Thrun]

even without having the equipment to set up a real compass course, loop closure error is definitely going to be the most useful way to compare the performance of a typical Suunto compass/clino with the SAP.

A single loop closure indicative of accuracy, but not definitive. It is like throwing darts at a bullseye. Accuracy helps, but someone less accurate might get lucky and put a single dart closer to the center. Also, accuracy, consistency, and precision are different things. A compass could be 180 degrees off and still get good loop closures.

George Dasher, who runs the compass course at Old Timers' Renunion, is better qualified to speak about the element of lick than I am.

[John Lovaas]

The plan view plot shows strong evidence of bias error. mp2 reads consistently high and ab2 reads consistently low.

Hi Bob-

I had not noted that before.

The survey lines are around 144' in length, and the maximum horizontal+vertical divergence of the 5 surveys is less than 3 feet, I think. I'd be interested to hear which of the instruments(#1 or #2) were used to calibrate the SAP. I see I mislabelled the Pony as SAC- Shetland Attack Compass; DOH! My apologies to the designer!

[George Dasher]

I don't run the OTR Compass Course any more. Dusty Gulden took over the job in 1999 so I could work on the 2000 Convention. I didn't want the job back once the Convention was over.

First of all, your compass can be off a bunch, but as long as it is off the same amount for each shot, it can give you consistent readings and a good loop closure. The results will just plot skewed compared to everyone else's.

Second, once you get within a certain accurate loop closure, the contest becomes pretty much a turkey shoot. Things like getting over the station and being just a little off on an azimuth can make or break your loop closure, so luck becomes more important than accuracy. Keep in mind though, that it is careful and accurate surveying that gets you to the point where luck takes over. I never was able to get that close (which is about 0.2 feet). It will be interesting to see if the SAP, because of its ability to get directly over the station and not having to be held horizontal, will do better at getting smaller loop closures on the OTR Compass Course.

I have never been able to find anything published on compass courses' results. I loaned out my own Suuntos for the OTR Compass Course, determined an average “correct shot” for each shot (using all the non-blundered data), and then made data comparisons between each of my two compasses.

I found that people shots generally stayed within two degrees of the average “correct shot” for each shot, but that the individual Suunto compasses varied between three or four degrees negative and three or four degrees positive of the “correct shot.” The same person using the same compass could shoot a consistent two degrees positive one year, and two degrees negative the next year--while using the same compass. This meant (in my opinion) that the compass course was measuring the quality of the person, not the quality of the compass.* I wrote an article on this for The Compass and Tape on this, and my conclusion was that compass courses (and the correction factors they generate) do not work to “calibrate” compasses.


* This fact is consistent with MP2 being a consistently "high" and AB2 being a consistently "low" in the example described above. You're measuring the person, not the instrument.

[John Lovaas]

The same person using the same compass could shoot a consistent two degrees positive one year, and two degrees negative the next year--while using the same compass. This meant (in my opinion) that the compass course was measuring the quality of the person, not the quality of the compass.*

George- did you ever have multiple year records of the same person and instrument? If I saw a 4 degree shift(from +2 to -2) in a person's readings within in a year, I'd think about wear and tear on the instrument- rather than wear and tear on the reader ;-) If the reader's shots shifted 4 degrees the other way the following year, I'd look at the reader, who is attending OTR. God only knows the shape they're in ;-)

I wrote an article on this for The Compass and Tape, and my conclusion was that compass courses (and the correction factors they generate) do not work to “calibrate” compasses.

Yup. Compass courses test the instrument and the reader- and it is hard to separate the two variables without manufacturing a device that can align an intrument on multiple predetermined bearings or inclinations, with minimal human input and interaction.

[Bob Thrun]

Yup. Compass courses test the instrument and the reader- and it is hard to separate the two variables without manufacturing a device that can align an intrument on multiple predetermined bearings or inclinations, with minimal human input and interaction.

The effects of instruments and readers can be separated by analyzing a set of data where everyone reads every compass. Jos Burger presented such a set with twelve readers and six compasses in Compass Points 30 (http://www.chaos.org.uk/survex/cp/index.htm). The readings were all between the same pair of points and each reader took only one reading with each compass. Jos Burger concluded that instrument errors were larger than reader errors.

I noted that one compass was about 3 grades higher than the others. I found the average reading for each compass. Then I tabluated how much each reader differed from the compass averages for his set of shots. If the one odd compass was removed, reader variations were about the same as instrument variations. A letter with my comments was published in CP 31. I did not present my detailed analysis because it was not my data.

At the last two NSS Conventions I handed out copies of my unpublished paper about the DC Grotto compass course. It did not involve loops. It had radial shots between a center post and ten periphal posts. There were 31 compasses and nine readers. Over 1600 compass readings were taken. Because not everyone read the same compasses, I was unable to separate reader and compass variations.

[George Dasher]

John:

Yes, I had the same user(s) with the same compass(s) shoot the course year after year. I particuarly remember my brother-in-law, who shot one year a consistent 2 degrees from the average, and the next year a consistent -2 degrees from the average--with the same compass.

I usually sketch with people who bring their own instruments, so my two sets of instruments lead a relatively charmed and aboveground life.

That's all.

george

[John Lovaas]

Yes, I had the same user(s) with the same compass(s) shoot the course year after year. I particuarly remember my brother-in-law, who shot one year a consistent 2 degrees from the average, and the next year a consistent -2 degrees from the average--with the same compass.

Hi George-

Did your brother in law(or other folks) run the course a third(or more) year? That's what I meant by multiple years. LIke I mentioned before, a 4 degree shift in a reader's results in one year would make me think that somebody drop-kicked that instrument at some point during the preceding year.

So if year 1 was +2, year 2 was -2, and then year 3, 4, 5, X was -2 as well- I'd vote for a dinged instrument- unless the reader was a beginner instrument reader at year 1. If the error was more random over time, I'd vote for a dinged reader ;-)

jl