Mesh - Converting Splines into Contours - scrn shot inc.

[Anonymous]

Hi People,

I'm experiencing an irration converting Spline Lines into contours using the Mesh Tool.

Everything works but when the contour is converted to be part of the mesh, I get hundreds of nodes along the contour as oppose to the sever present when it was still in 2d spline line form.

Anyone else experiencing this problem?

I'm using Archicad 10 Student Version with the latest updates downloaded running XP with latest Service Pack.

Cheers [/img]

[Dwight]

You probably have your magic wand setting too fine.
This setting controls the precision of curve conversions.

However, if you are in the process of converting a site plan into a 3D terrain, contours are a bad way to do it. Contours are themselves interpolations of other data, so don't get caught up in their sinous beauty.

It's a trick! Efficient site modeling is an art:

It is faster in all ways, drafting and imaging, to impose a 5x5 m grid (finer or coarser, depending on the steepness and regularity of the site) and only plot the grid intersections. If there is an important outcropping, you would add nodes at that place. Smoother areas could have even fewer nodes.

[Anonymous]

Definitely a magic wand settings issue (options menu - magic wand settings). Most site information is delivered to you in a contour format so you are following a correct procedure. I often use the segment length methodology (see attached image).

[Anonymous]

Thanks guys... I followed the instructions in regards to the Magic Wand Setting and indeed it was set far to fine. Now I'm finally getting normal contours

Dwight - Yeah I must confess that I love the little contour lines on the site plans and often follow them blindly without question

But with your way of modeling a site plan by the grid setting, I'm still not exactly sure of how'd you'd do it. Are you suggesting that we draw a grided Mesh model and only adjust those nodes' height accordingly?

Cheers

[Dwight]

You DO leave the contour lines on the drawing.....

Clarificating my original suggestion:

You wouldn't do this if you had real site data - not interpolated contours - and needed total accuracy.

I suggest that sites need not have supreme accuracy or slavishly follow contour lines because that method puts in too many nodes. You'll also notice that by the time you reduce magic wand tracing accuracy with contours enough to get acceptable numbers of nodes that the thing is short-cutting all the extremely curved contours: shows low.

My method doesn't require a grid mesh, but that doesn't hurt to start. My method involves varying mesh node density to adequately describe land slope: fewer nodes where the slope is: away from building footprint and land is of shallower slope, more nodes as you get close to the building footprint or to describe important outcroppings.

Using only a strict grid yields a robotic site mesh while an adaptive node array as i propose artistically describes the site.

[Anonymous]

I use contour lines.

Dwight, I see your reasoning about point density and artistic site models, etc., but I find the contours much easier and quicker to manipulate and everyone understands them in an instant.

I find it is very rare to have enough verified points to form a realistic looking site and moving the contours until it "looks about right" is much faster (for me at least) than manipulating single nodes.

[Dwight]

My technique is best used on irregular sites where there's way too many irrelevant jiggies and the node count skyrockets without improving site description.

[Anonymous]

I agree that Dwight's meshing method has some valuable features in some circumstances: its fast and variable compared to the general method of splining every contour line at worst case wand or spline setting. The grid can be flexible, whatever you decide fits the site & design requirements.

However, on current residential projects the DWG from the Surveyors is the "bedrock" that everything relates to. Those contour lines are what the plan checker uses to assess if you are meeting their height restrictions, and are what any variance panel will consider as gods truth if you need to submit your documents to them.

Dwight's right, the contours on the Survey drawing are interpolations. Our usual surveyors use an autocad add-on, MicroSurvey, to do the interpolation. On our typically steepish, rocky View! sites I'm not going to second guess its contours, even though I might have over a hundred of their points shown on the drawing as well. The surveyors contours are 'Legal'.

We generally draw over the survey contours with the spline tool and take considerable care in doing that to assure ourselves that our site model is as compliant as we can make it with the survey drawing.

I set my iMac up on a Development Officers desk a while ago and gave him a demo of Archicad and how we made the site and building models work together (plug for iLugger bag). At his choice of three locations I used the section tool, then the line and dimension tools to quickly varify to 1/8" the clearance of the roof ridge. It felt a bit cruel, actually, he was very impressed, but how does Dwight put it..."the green bile of envy rose..."
Poor guy has to carefully check compliance of every set of drawings submitted for approval by hand. So 20th century!

My conclusion is that careful site modeling allows you to sleep well if your drawings need to comply with strict height or view requirements.