mrsusan
2011-10-18, 12:11 PM
Hi,
I have received extremely large txt files (XYZ) to be used to make a surface in C3D. Bringing them in isn't a problem, but the contours display very jagged.
Ideally, I'd like both a smooth, accurate surface with a smooth profile, but it's mainly the profile I'm interested in (although seeming as it reads the surface, I'll have to mess around with that first).
I have tried sampling every, say 10 points which smooths it to an extent but I find that there are frequent large spikes in the profile, as well as it still being quite jagged. Using surface simplification and smoothing only seems to yield very minimal results. Sampling 20+ gives me even bigger data spikes.
I am using 2010, I understand that there is a profile line of best fit feature in 2011, but does anyone know if there is a way to smooth out the surface properly without data spikes or lots of data loss? Or do I tell the surveyor his data is bad?
Attached a picture as it would probably explain my predicament better than I can!
Thanks in advance
I have received extremely large txt files (XYZ) to be used to make a surface in C3D. Bringing them in isn't a problem, but the contours display very jagged.
Ideally, I'd like both a smooth, accurate surface with a smooth profile, but it's mainly the profile I'm interested in (although seeming as it reads the surface, I'll have to mess around with that first).
I have tried sampling every, say 10 points which smooths it to an extent but I find that there are frequent large spikes in the profile, as well as it still being quite jagged. Using surface simplification and smoothing only seems to yield very minimal results. Sampling 20+ gives me even bigger data spikes.
I am using 2010, I understand that there is a profile line of best fit feature in 2011, but does anyone know if there is a way to smooth out the surface properly without data spikes or lots of data loss? Or do I tell the surveyor his data is bad?
Attached a picture as it would probably explain my predicament better than I can!
Thanks in advance