I don't recall whether or not I've posted about this before, but it's a problem I encounter occasionally with resections in SurvCE (I'm using version 5) and wonder if anyone else has ever run into it.
I'll shoot 2 or more points, but the resection fails. If it's just 2 points the resection is unsalvageable, if there are more than 2 (the usual case) I can repair the resection by disabling the one that's causing the problem, once I identify it by trial-and-error. In the past I've just chalked the problem up to something I must have done wrong, reshot whatever I needed to, and moved on.
Today when it happened again I decided to look a bit deeper. I had shot 3 points that had all been set earlier in the day from the same setup, and expected a slam-dunk result, but instead it failed (residuals were in the 1.5-foot range). I started disabling points one at a time and recalculating. The third shot -- point 15 -- turned out to be the "bad" one -- disabling it brought the 3D residuals below 0.01'. I stored the resected point, then set about figuring out why point 15 didn't fit. I returned to the backsight and re-zeroed, then shot D+R to point 15. When I compared the check shot to the original, the differences were below 0.01'.
I'm at a loss to explain the failure, so any insights that might allow me to avoid the problem in the future would be appreciated.
Thanks!
I have found that bad rod height settings account for a majority of failures in resections and in StarNet adjustments as well.
It's happened to me enough times to make me half expect it.
I've found that repeating the resection but shooting only on one face - then it works.
But I don't want to shoot only one face...
Also changing the order you shoot the points can make it work.
I have put it down to just an unresolved glitch in SurvCE - and I do recall this being discussed here before.
Delving into the raw data I see what happened, though I don't know why: when I shot 15 during the resection, the DC recorded the direct reading as reverse, and the reverse reading as direct, so it calculated the point 180° in the wrong direction, which blew up the resection.
Bummer that this happens as often as it does, but like cf-67, I guess I'll just have to expect it and accept that there'll be some do-overs.
@jim-frame Are you resecting 1-man? Are you doing D & R manually? If so maybe you forgot to manually flip back to D before you started the resection to 15?
I use Fieldgenius v11 with my Leica TS15 and after I set zero on the BS in direct mode then somehow the software always uses the R shots correctly and never calcs them 180 out. When I am staking something where elevations are critical I will set a point direct and then manually flip the scope and use stakeout to check the stake I just set. Usually the screen pops up N+/-0.01' E+/-0.01' EL+/-0.01' so I know the stake and elevation are good and obviously the R shots are being calc'd correctly.
Note: When I say manually I am clicking a button on the DC that causes the scope to flip 180 and lock onto the prism again.
Yes, I'm resecting 1-man (I do everything 1-man). But I let the DC and the instrument handle all the pointing, so the flop glitch is happening somewhere between the DC and the instrument (and I suspect it's SurvCE's doing). I'll have to pay more attention to see if I can figure out what the common conditions are when these failures occur. It'd be nice to know what to avoid, especially when one of the resection points is inconvenient or dangerous (i.e. in traffic) to occupy -- those are the failures that break you heart.
Ya 1-man all the time here too and it seems like only the one shot that is difficult (timed the traffic signal and scooted out and back to shoot a Mag/CityDisk in traffic) and that I really do not want to shoot again is the one that gets fucked (I bumped something during the scoot and the shot is not saved).
That has happened to me. I assumed it was my mistake. Maybe not always. My other problem with SurvCE resection is: usually I need MY location of the points I have resectioned to. It has the data, but won't generate the points and I have to shoot them all over again. Am I missing something?
I have reported things like this to Carlson Tech Support in the past and received a patch via email before I could get back to the office....
No you are not missing anything and my software, Fieldgenius does the same as SurvCE and IMHO that is for a good reason. Assuming the points you resected from are hot (not GPS) and the residuals are small then the only reason to re-shoot the points you resected from would be to check the software result BUT I would not store a new point there because there is error (residuals) in the computed resected point under the TS and storing new points at the original points used in the resection will probably induce more error. I do not change my original control used in the resection.
Now when I resect using GPS points (they are in a separate GPS layer and point range) I play with which ones to hold to get the best residuals then I shoot them all again and store with new point numbers in my robot layer. I do not use those GPS points again, they are only kept as reference points if I need to revisit what I did.