hatrick Posted April 26, 2013 Share Posted April 26, 2013 (edited) Hi, Could someone please take a look at the file attached. The goal in this example is a constant arclength of 1 impartial from the bending factor. Thank you Lucas prevent_stretch.hip Edited April 26, 2013 by hatrick Quote Link to comment Share on other sites More sharing options...
anim Posted April 26, 2013 Share Posted April 26, 2013 here is one of the ways to do it for your case prevent_stretch_fix.hip Quote Link to comment Share on other sites More sharing options...
hatrick Posted April 26, 2013 Author Share Posted April 26, 2013 it's incredible how helpful and fast this community is! ... thank you anim ! Quote Link to comment Share on other sites More sharing options...
hatrick Posted April 30, 2013 Author Share Posted April 30, 2013 (edited) any ideas for a method using the carve SOP. In my case i'd like to control the perimeter of a surface after! it is build up from curves. So the idea was to convert to nurbs and use the carve SOP for the cut/length operation. The measurement would happen with an extracted curve similar to the prevous example. Cheers Lucas Edited May 1, 2013 by hatrick Quote Link to comment Share on other sites More sharing options...
hatrick Posted May 1, 2013 Author Share Posted May 1, 2013 (edited) a further issue is that i only can define a maximum seg length in the resample no minimum ... does anyone have an explanation why the initial setup with the carve SOP does not work? Edited May 1, 2013 by hatrick Quote Link to comment Share on other sites More sharing options...
anim Posted May 1, 2013 Share Posted May 1, 2013 I think carve isn't working correctly because it's not sampled by distance (or arc length) but by U-coordinate, which is not linear so if the whole length is 10 units and you want 20% from that, setting carve U to 0.2 will not necessairly give you that, it will give you length from u = 0 to u = 0.2 which can be of any length especially if the control points on curve are more non-uniformly distributed in 3d space why do you need minimum segment length? maximum segment length is actually exact segment length that will be used unless you reach the end of the curve Quote Link to comment Share on other sites More sharing options...
edward Posted May 2, 2013 Share Posted May 2, 2013 Could someone please take a look at the file attached. The goal in this example is a constant arclength of 1 impartial from the bending factor. Sorry, didn't look at your file, but try using a Basis SOP with Parameterization set to Approximate Arc Length. Quote Link to comment Share on other sites More sharing options...
hatrick Posted May 2, 2013 Author Share Posted May 2, 2013 @anim. The bending process not only causes stretching but compression. So it seems the max seg length in the carve SOP prevents the curves from going over the wanted value but not below. So i thought something like the minimum seg length could solve that issue. @eward. Thank you! i'll give it a try. Quote Link to comment Share on other sites More sharing options...
hatrick Posted May 2, 2013 Author Share Posted May 2, 2013 I'm not getting how to use the BASIS SOP here .... thank you Quote Link to comment Share on other sites More sharing options...
anim Posted May 2, 2013 Share Posted May 2, 2013 (edited) ... The bending process not only causes stretching but compression. So it seems the max seg length in the carve SOP prevents the curves from going over the wanted value but not below. So i thought something like the minimum seg length could solve that issue.... I'm getting lost here, you are applying some deformation on the curve which makes it longer -> in that case it's easy to trim it to original length with no problem using Resample SOP if the deformation causes it to shrink how exactly do you expect houdini to compute missing part? straight line in tangent direction? continuous arc in outgoing curve curvature? in any way you need to recreate it yourself, then you can run Resample SOP to trim it to exact length and keep the same number of segments if your problem is easier however, like if you want bend without altering the length of the curve you can make one using VOP SOP or even easier download from exchange http://www.sidefx.co...0&versionid=670 or maybe http://www.sidefx.co...1&versionid=721 I haven't tested them but they sound like doing the correct type of bend, because houdini's bend is a joke Edited May 2, 2013 by anim 1 Quote Link to comment Share on other sites More sharing options...
hatrick Posted May 2, 2013 Author Share Posted May 2, 2013 (edited) probably i missed the point that building up the curve from below the wanted length would be a more complex task?! ... anyway. Please see the scene file attached that should illustrate the issue in more detail. The stretching comes from several deform operations. I asked sesi to take a look at the file. i'll share the feedback here. In the meantime i'm very curious about your thoughts. (The scene contains a great "book" asset from a talented artist, who was kind enough to share his work on the web. Unfortunately i can not remember his name) cheers lucas COMMUNICATION.rar Edited May 3, 2013 by hatrick Quote Link to comment Share on other sites More sharing options...
hatrick Posted May 3, 2013 Author Share Posted May 3, 2013 At first sesi came up with a simple and effective, but non procedural solution. Shearing the geometry befor the bend operation with a transform SOP gives quite nice control over the stretching behavour. Never the less an exact, procedural method would be great! cheers Lucas Quote Link to comment Share on other sites More sharing options...
hatrick Posted May 14, 2013 Author Share Posted May 14, 2013 Here is a procedural setup provided by SESI support! Rollup2.hip Quote Link to comment Share on other sites More sharing options...
anim Posted May 15, 2013 Share Posted May 15, 2013 Here is a procedural setup provided by SESI support! that's clearly not maintaining original length - original -> 17.6 - deformed on frame 60 with 50 segments -> 19.3 - deformed on frame 60 with 1000 segs -> 19.3351 what problem do you have with single Resample SOP solution? it's not preserving length 100% but it's much closer than any other methods I've seen - original -> 17.6 - deformed on frame 60 with 50 segments -> 17.5696 - deformed on frame 60 with 1000 segs -> 17.5988 so until Carve SOP can work in normalized arc length values as well (not just parametric uv) I think resample SOP is the way to go or maybe I just completely misunderstood what you want curve_maintain_length.hip Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.