12-12-2022, 06:56 PM
Hi,
thanks for your answer. Yes, I tried it all
I narrowed the problem down and understand it a bit better now.
I think it's the hard change in normals on the spline when doing a swirl.
With a generated mesh on the main spline this isn't a problem, but as soon the generator shape gets offseted, either with "TRS Shape" or with "Path Relative Transform", then there is a sudden change in direction on the offseted mesh on start and end of a swirl, because it relies on the normals of the spline.
Completely makes sense to me, so this thread may morph into a feature request beg, rather than an actual problem with curvy
I think it would be an awesome feature:
A new parameter that fades in/out the swirl at the start and end of the swirl segment/anchor group.
A distance percentage value for instance.
With this, we could do swirls over an anchor group and say "fade in from the first CP over the first 10% of distance" and "fade out to the orientation of the last CP over the last 20% of distance".
Additionally this would decrease the tuning demands for swirl turn parameter and increase the flexibility overall
Let me know what you think about the idea, or if you see a possible solution to achieve this with the already existent system.
regards
Michael
thanks for your answer. Yes, I tried it all
I narrowed the problem down and understand it a bit better now.
I think it's the hard change in normals on the spline when doing a swirl.
With a generated mesh on the main spline this isn't a problem, but as soon the generator shape gets offseted, either with "TRS Shape" or with "Path Relative Transform", then there is a sudden change in direction on the offseted mesh on start and end of a swirl, because it relies on the normals of the spline.
Completely makes sense to me, so this thread may morph into a feature request beg, rather than an actual problem with curvy
I think it would be an awesome feature:
A new parameter that fades in/out the swirl at the start and end of the swirl segment/anchor group.
A distance percentage value for instance.
With this, we could do swirls over an anchor group and say "fade in from the first CP over the first 10% of distance" and "fade out to the orientation of the last CP over the last 20% of distance".
Additionally this would decrease the tuning demands for swirl turn parameter and increase the flexibility overall
Let me know what you think about the idea, or if you see a possible solution to achieve this with the already existent system.
regards
Michael
