SubD Bridge "Straightness" not behaving as expected

Firstly congratulations to the team for making SubD available. It’s excellent. :clap:
I’ve finally had an opportunity give the workflow a red hot go (modelling some ceramic moulds) and I’ve started running into a few difficulties.
Fortunately most of those problems have been addressed with the great tutorials here: Rhino 7 SubD Tutorials [McNeel Wiki]

However this one remains:

Planar open bridges behave exactly as expected, however holes do not seem to take the normals into account. The result is always flat. If the loops are properly flipped so as not to go inside out, then command fails with Bridging result not valid.

I’m guessing that the connected edge tangents are used, but in many cases the connected face normals would make a lot more sense.
I have modelled the expected result manually.

These other three work ok, however they are very linear still, even with Straightness set to 0

It’s been half a decade since i’ve used T-splines or Maya, but i seem to recall this process being very intuitive. Even if my memory is failing me, would it be an improvement to utilize face normals for face bridge orientation? Could this be an option in the command?
or would this break something else?

bridge.3dm (362.6 KB)

By the way I do recognise that the workaround is fairly simple: extrude the faces, delete them and bridge the resultant naked edges.
bridge-workaround-small-culled

But wouldn’t it be nice if this could give the same result?
bridge-fail

1 Like

If I understand you correctly, you need such a model. This is done by the regular Scale command through Gumball.
You can also use the Multipipe command on the transition or Soft Transform

1 Like

No @inju, think you misunderstood. I modelled what i expected to see with the bridge command. It is maybe not clear in my screengrab that the actual model is 2D. the specifics of the interpolation was not my problem the direction is the problem. I realise it would have been clearer if i kept thickness of the bridge the same in both scenarios. The scaling is not the issue. If the input to bridge is 2D, the result is 2D. I was asking for the developers to consider the scenarios where this produces an invalid result as my gif demonstrates. I believe it would make sense for the the new bridge vertices to be offset normal to the originating face, if bridging edges it makes sense to use the edge tangents

Hello - what should happen when the ‘right’ diretion is the tangent though (more often than not, I would say, by a lot) ? I can see it is not in a flat thing, but at what point does it change, how would that be controlled, via some UI?

-Pascal

Hi dharman

I think it is difficult to compare with programs (max, maya, etc ) that use meshes.

Sub-D is Not mesh.

We need to find a new modeling algorithm? expression technique?

Hope this helps

-Thank’s RMA Team that’s developing better Rhino 3D tools

1 Like

Hi Pascal, yes maybe it’s a switch. Would it make sense for bridges with face inputs to use the normals, and for bridges with edge selections to behave as it currently does? I can’t picture how that would impact other cases. Do you see that breaking anything else? I don’t want to introduce UI bloat.

Any changes on this ?.. The current subD bridge still seems to act the same way the flat output