Twisted curve and sweep SOP

   5057   4   1
User Avatar
Member
49 posts
Joined: Feb. 2007
Offline
I have this problem with a twisted curve when I use a sweep operator. I see sweep sop it have Twist control, but I can't use it handly to fix this issue.

The curve is generated by a equation from wrangler, and must be a close curve. I've tried closing curve from code and with ends SOP.

I need have to control closing point of curve profile at start and at ends from sweep operator, else for curve rotation.

Attachments:
Captura de pantalla 2017-01-01 a les 20.44.58.png (1.4 MB)

User Avatar
Member
670 posts
Joined: Sept. 2013
Offline
It´s nicer to post a hip file.
https://procegen.konstantinmagnus.de/ [procegen.konstantinmagnus.de]
User Avatar
Member
49 posts
Joined: Feb. 2007
Offline
Uploaded hip file.

I've seen the problem appears when the equation curve is heavy twisted. With a simple circle for example, works as expected when arc is closed in the same Circle SOP or keep arc opened, then using Ends SOP to close it.

I think in a way to control the ends up vector, but maybe there is a easy solution to keep a correct curve continuity respect its points orientation.

Attachments:
twisted_curve_wrangle.hipnc (78.2 KB)

User Avatar
Member
2624 posts
Joined: Aug. 2006
Offline
I have tried everything and I get issues with the winding order, even after rebuilding the curve. Maybe revise how you create the twist and do that as a post created line effect.

Rob

Attachments:
rse_fixed_twisted_curve_wrangle.hipnc (117.8 KB)

Gone fishing
User Avatar
Member
49 posts
Joined: Feb. 2007
Offline
Thank you circusmonkey!

Using your method with Polyframe and copyOp instead SweepSOP tangent + normal attributes, it seems works ok. Using a open profile and close with SkinOP is valid for my purporse too.

After I've seen that it have a bit of twist in the ends of curve. However it's suitable in most situations of twist and divisions.

It would be possible add a value to an attribute?

I want to try adding a small increment to current normal to attempt compensate the current deviation.
Edited by Unaided - Jan. 3, 2017 10:44:03

Attachments:
fixed_twisted_curve_wrangle.hipnc (84.0 KB)

  • Quick Links