Getting a smoother 3D parametric plot

[george seifert]

I'm new to using 3D graphs so sorry if this is a dumb question. I'd like the look of a 3D surface plot, but the flexibility of a 3D parametric plot. The surface plot look like it fills in (interpolates?) the data between points. The parameteric plot looks choppy. My X and Y data represents movements of a positioner table so it may not always be exactly the same for each sweep of the table. That's why it'd be nice to use the parametric plot. I've searched through the 3D pallette but can't find much for changing the look of the graph - at least nothing that helps in this situation. Am I asking for something that can't be done or am I missing something?

I've enclosed a really simple VI I'm using to show the difference between plots. Download File:post-2786-1208446881.vi

Thanks,

George

[Tomi Maila]

I use the LV 8.20 introduced 3D primitives to draw 3D plots directly from polygon meshes. Using this method you can specify normal vectors for each polygon corner and the rendering will look smooth. Maybe not the easiest way but fits to our needs for 3D rendering of a brain.

[george seifert]

QUOTE (Tomi Maila @ Apr 17 2008, 10:54 AM)

I use the LV 8.20 introduced 3D primitives to draw 3D plots directly from polygon meshes. Using this method you can specify normal vectors for each polygon corner and the rendering will look smooth. Maybe not the easiest way but fits to our needs for 3D rendering of a brain.

Do you have to use the 3D picture control (not the 3D graph) for that? The 3D graph seemed attractive because it was so easy to setup.

[LAVA 1.0 Content]

QUOTE (george seifert @ Apr 17 2008, 02:07 PM)

Do you have to use the 3D picture control (not the 3D graph) for that? The 3D graph seemed attractive because it was so easy to setup.

Hi George,

The problem with 3-D Graph questions is it generally takes a lot of work to answer them.

Being lazy I wrote up a "worse case" 3-D graph example and posted http://forums.ni.com/ni/board/message?board.id=170&thread.id=143663&jump=true' target="_blank">in this thread on the "dark side".

Please take a look at that thread and then let us know what is still missing. It include comments, screen-shots, as well as all of the code I used.

Take care,

Ben

[george seifert]

QUOTE (neB @ Apr 17 2008, 01:33 PM)

Hi George,

Being lazy I wrote up a "worse case" 3-D graph example and posted http://forums.ni.com/ni/board/message?board.id=170&thread.id=143663&jump=true' target="_blank">in this thread on the "dark side".

Please take a look at that thread and then let us know what is still missing. It include comments, screen-shots, as well as all of the code I used.

Take care,

Ben

Funny I was just looking at that example just before I saw your post. Unfortunately I don't see how it'll help me tell you what's missing. If you download the VI I posted and run it once with the type set to "Parametric" and then again with the Type set to "Surface" hopefully you'll see what I mean. The surface plot looks more like what I was expecting. The problem with the surface plot though is that you only get to supply one set of XY data - if I understand it correctly that is.

Maybe if I explain what my test does it'll help. I have a Keyence displacement sensor mounted above an XY slide. The device I'm measuring (basically a hockey puck shaped device) will be moved beneath the sensor. I'll increment in the X direction to scan across the device. Then move a small increment in the Y direction and then scan across it again in the X direction. And so on until I map the whole device. I just want to display a 3D picture of the Z displacements.

George

[LAVA 1.0 Content]

QUOTE (george seifert @ Apr 17 2008, 03:19 PM)

Funny I was just looking at that example just before I saw your post. Unfortunately I don't see how it'll help me tell you what's missing. If you download the VI I posted and run it once with the type set to "Parametric" and then again with the Type set to "Surface" hopefully you'll see what I mean. The surface plot looks more like what I was expecting. The problem with the surface plot though is that you only get to supply one set of XY data - if I understand it correctly that is.

Maybe if I explain what my test does it'll help. I have a Keyence displacement sensor mounted above an XY slide. The device I'm measuring (basically a hockey puck shaped device) will be moved beneath the sensor. I'll increment in the X direction to scan across the device. Then move a small increment in the Y direction and then scan across it again in the X direction. And so on until I map the whole device. I just want to display a 3D picture of the Z displacements.

George

The parmetric plot connects elements that are adjoining each other with your data it looks like it connecting oposite ends.

Reverse every other row of X data and see how that plots for you.

Ben