RTS Command

[fer310b]

Should you be able to input the parameters of a Helmerts Transformation result into the RTS Command and also come up with the same transformation?
I have tried to test this but always get different values when applying the RTS command to the original points.

[Brian Sloman]

Hi, I just ran a test on this and I do see very slight differences between the results of Helmerts' Transformation vs RTS, but the differences are within about 0.005m and I think are due to rounding, particularly with the rotation factor which is only reported to the whole second. However, if you have any concerns that you are getting incorrect results in either routine, please submit a technical support ticket here and one of our support reps can assist.

[fer310b]

Thanks for the reply!
I will run a scenario and try it again, because my results were certainly more than millimetres. That way I will have some specifics.

[Brian Sloman]

Hi, one more thought occurred to me - the test I ran was using local coordinates near 0,0 (using the sample file "Examples\sample-helmerts.dwg") but if you are using larger coordinates, any error that gets introduced due to rounding in the rotation factor will become more exaggerated the further you are from the origin. So if you're using UTM coordinates it could be many meters different: for example, an inaccuracy of 0.5" propagated over 5,000,000 meters would represent a difference of over 12m. sin(0.5")*5000000m=12.12m
Does this sound more like what you are seeing?

[fer310b]

Yes that sounds more like it. In New Brunswick are provincial coordinate system uses a false northing and so the coordinates we use are like 7500000 & 2500000. So certainly the precision of the rotation would be crucial. I believe when I print the Helmert result I use the the rotation used but as you said earlier, it only displays to the nearest second if I remember correctly.