Our Registrar General's directions no longer include the requirement for rounding dimensions on our plans that are lodged for registration to 10's of mm or 10's or 20's of arc seconds. Our dimensions are required to be correct to 3 decimal places, and bearings should be shown as measured. Intuitively there seems to be no reason why rounding my plans would give the next user an advantage of any kind, so I don't do it. When retracing my own fieldwork I'm finding that agreement is closer to the mm level rather than the cm level, so I figure showing 3 decimal places in my distances is a reasonable thing to do. For azimuths, the nearest second seems over-adequate for azimuths, but closer in truth than 10" or 20" favoured by others in my area. I can't think of a reason why I'd take the additional step of adding or subtracting totally random amounts to the calculated dimensions between my monuments. Hopefully people using my plans will appreciate that the parcels and connections close nicely, without adjustment in their CAD. That's how I see it anyway.
The people I've asked about rounding their measurements seem to see it as some kind of indirect statement about the achievable accuracy by a cadastral surveyor. I'd think this could be better conveyed by a direct statement of accuracy on the plan face, if the recording authority permitted it. Just tell people directly what to expect. That is if the surveyor actually know what to expect from their work. Also, seemingly, nobody wants to be known in their area as "mm Mick". I've heard this used as a term of derision a few times.
This topic would rank pretty low (near last?) on the importance scale as far as cadastral surveying and plan-making goes, but it's just something that has piqued my interest. Perhaps because it seems to me like it's mostly an acquired, ideological position, rather than a practice that results in a demonstrable benefit. As I said, I just turned the rounding off. You have my measurements and calculations, unaltered, and best I could make them. So, why do you do it, or not? What is your ideology on this issue of little importance? Have you declared a jihad on the mm?
Just out of curiosity, what country are we talking about?
Hello Scott, I'm in Australia
There are good reasons for rounding and there are reasons to leave the precision alone by what is kicked out by the black box. Sometimes you might go one digit beyond what you think your precision is so to not pollute the data with error due to rounding. You might decide to round to what you know your precision is. I highly doubt that if you send three surveyors to measure that same line, you will get the exact same number. I think it is and/or should be the stamping professional surveyor's decision as to his way to display it. It should be a governmental checker's duty to make sure the plat has the minimum standards on it (in my opinion). If I disagreed with someone else's precision, I would make suggestions, or bring it up in discussion. Not make some declaratory blanket statement that this is how it is. (I don't know what a Registrar General is other than what I infer from the title, I am just talking in general).
I would suggest that if you want to carry out your report beyond what your precision technically is, you have a statement on the plat like: "It is estimated that the precision of the data shown is +/-XX with 95% certainty. Values are shown to a higher precision to prevent additional imprecision due to rounding error".....something like that.
Of course in Australia, the opposite may be true. 😉
Australia has some excellent documentation on base lines and calibration. Seems counter intuitive to not show data to mms.
Historically our government has slowly been tightening up its dimensional measures.
Positions provided in New England datum was tenth of a meter, NAD27 was to hundredth of a foot, NAD83 to a mm.
I agree with you, Conrad.
Every professional Surveyor should be aware of what's achievable. When I see bearing to the second and feet to the hundredth through a half mile of steep forest in the 1960s I am aware that plus or minus a foot is to be expected but I still appreciate that they provide their answer that way.
Many years ago (before everyone had a data collector) I watched a local surveyor reading distances to 0.005 ft from his instrument and recording them to the nearest 0.02 in a field book. Similarly for angles, whatever the increment was. I scratched my head and asked why and was told because the board required plats to be shown to a precision of 0.02 ft. I thought that strange.
If I was rounding, I certainly would not do so until I had performed adjustment on the data. As said above, rounding is adding random error to every measurement, and in this case more random error than I would have expected to exist in the instrument reading. And I would interpret the board requirement as "0.02 or better" unless I had a ruling to the contrary.
If your precision is actually to to the nearest mm then it should be reported to the nearest mm, but why would you do a boundary survey to that precision. It is very hard to measure a property line to the mm. I have never met a land owner who cares about a mm.
If you don't round till after all adjustments and calculations are made you are not introducing any error at all.
I have never accepted the numbers that Autocad based software spits out.
The rules of rounding I learned do not apply in their software and most of the time when there is one length on one side of a line and multiple lengths displayed for the breakdown of the other side of the line, the values rarely add up to the same value.
I am always having to edit each segment of cad produced labeling against the rules of rounding as my mentor taught and it is a process that includes inversing between each segment and noting each value to 3 or more decimal places and correcting them as I go.
As I was taught, for example:
0.025 = 0.02
0.035 = 0.04
Autocad results:
0.025 = 0.02
0.035 = 0.03
What state are you in Conrad? Is this part of eplan?
Seb, post: 394388, member: 7509 wrote: What state are you in Conrad? Is this part of eplan?
Hello Seb, I'm in NSW, but none of this has to do with ePlan TIFF or the turd that is LandXML
Here in Tassie, we are about to have a discussion about what is and what isn't required on survey notes. This is being driven by the eplan idea and I'm worried that a lot of the info currently shown will be deemed surplus in today's coordinated era.
We currently have a plan of survey that has distances rounded to two places and rounded bearings (different roundings dependent on length). We also have survey notes that show everything to the mm.
Seb, post: 394405, member: 7509 wrote: We currently have a plan of survey that has distances rounded to two places and rounded bearings (different roundings dependent on length). We also have survey notes that show everything to the mm.
As for me, I'd rather have a list of the coordinates of the various marks and corners shown on a plan with their uncertainties with respect to some permanent and reproducible datum. In modern practice, it's routinely possible to have coordinates kilometers apart with relative uncertainties at the 5mm level, so why isn't it more direct to say as much instead of trying to make the bearings and distances along the chain of connections between them more clumsily express the idea?
Isn't the real point of the bearings and distances to allow the computation of the actual relative coordinates of the various points positioned by the survey as well as providing an independent check upon any coordinates stated on the plan?
Conrad, post: 394301, member: 6642 wrote: The people I've asked about rounding their measurements seem to see it as some kind of indirect statement about the achievable accuracy by a cadastral surveyor. I'd think this could be better conveyed by a direct statement of accuracy on the plan face, if the recording authority permitted it. Just tell people directly what to expect.
Absolutely. More consistent with modern practice, too.
A Harris, post: 394387, member: 81 wrote:
As I was taught, for example:
0.025 = 0.02
0.035 = 0.04
Autocad results:
0.025 = 0.02
0.035 = 0.03
It seems you were taught bankers' rounding: towards even from the middle
A Harris, post: 394387, member: 81 wrote: I have never accepted the numbers that Autocad based software spits out.
The rules of rounding I learned do not apply in their software and most of the time when there is one length on one side of a line and multiple lengths displayed for the breakdown of the other side of the line, the values rarely add up to the same value.
I am always having to edit each segment of cad produced labeling against the rules of rounding as my mentor taught and it is a process that includes inversing between each segment and noting each value to 3 or more decimal places and correcting them as I go.
As I was taught, for example:
0.025 = 0.02
0.035 = 0.04
Autocad results:
0.025 = 0.02
0.035 = 0.03
There's a way to change how AutoCAD rounds but I don't remember the setting off the top of my head.
Stephen Ward, post: 394436, member: 1206 wrote: There's a way to change how AutoCAD rounds but I don't remember the setting off the top of my head.
The way I control autocad's rounding is to click the little red X at the top right hand corner, then load surveying software!
Sorry, cheap shot, I know...
Conrad, post: 394403, member: 6642 wrote: Hello Seb, I'm in NSW, but none of this has to do with ePlan TIFF or the turd that is LandXML
What are your issues with LandXML? I rarely deal with it, but my sense is that it is a good format / standard.
.025' = .02' or .03'.... I think we are splitting hairs here.
Rich., post: 394483, member: 10450 wrote: .025' = .02' or .03'.
The purpose of rounding to the even number, instead of always up or always down, it to avoid the accumulation of a slight bias. If you need to add up a lot of numbers that have been rounded the same direction the total will be further from the unrounded total than if you round even.
