Surface Isocurves from Patch


#1

Hi, everybody!
I noticed such problem… I use Patch, I cut a surface of a curve and I apply BlendSrf - the defect is shown here! Isocurves joint at a surface (Patch) badly influences work of BlendSrf.
Untitled.3dm (384.4 KB)


(Pascal Golay) #2

Hello - you can do several things to make this better -

  1. Use the controls to add and align cross section shapes.
  2. RemoveMultiKnot on the result.
  3. makeUniform on the result
  4. MatchSrf again on both edges.

But, I’d use a Loft and MatchSrf here too - Loft the edges, with Rebuild, maybe 32 points, then MatchSrf on both edges. If it does not join up, use Refine in MatchSrf… LoftTransition_PG.3dm (172.6 KB)

-Pascal


#3

Thanks!
However similar strongly complicates work with the cut surfaces.:frowning_face:


#4

And why it in general occurs? How these harmful points at defect are formed?

@sochin, @Helvetosaur, @Toshiaki_Takano, @davidcockey


#5

@Modeler3D Wow flattered that you think I would have any idea and include me in this company…I am a crude sub-d type of modeller that fixes these type of flaws with a file and emery in post production… if I start looking at zebras I know I am in the wrong place for what I do :slight_smile:

You are in good hands with the other guys.


#6

I have assumed that you can know something. :slightly_smiling_face:
A smooth edge, and the smoothing surface with bends - as so?


(David Cockey) #7

When using BlendSrf you can adjust the seam location, and then use Add Shapes to control the isocurve directions.


#8

Thanks! :grinning:
It is a little manipulations with BlendSrf, plus RemoveMultiKnot - result quite pleasant.


(Tom) #9

it may also improves the shape if you revolve the top part instead of pull it in like you did. Best blending usually has iso alignment.


#10

I don’t understand something… It is possible in more detail?


(Tom) #11


#12

Do you suggest to adapt the top surface (to make at it more smooth edge) to the lower surface? Smoothing will be simpler… And the bend of curves at smoothing - poorly influences, well the direct defect turns out and what?


#13

I’d just Add Shapes to make the S curved isocurves go away…
I’m not that experience enough to be too picky yet…

But, point of making a good Nurbs surfaces seems to be making sure
U and V are as rectangular as possible to have a good shape and
something that fits with the other tangent surfaces.

You can play around even more with control handles while in blendsrf…

@pascal’s loft seems to be what we’re after I think…

To make the isocurves more straight, you can Alt+Mouse Click and Snap to the edge points so that, you change the angle to be quite straight.
After few added shapes, I think you can allign fairly well.

SomethingBetter.3dm (360.0 KB)


(Pascal Golay) #14

It’s intersting that you can see that the ugly bits on the blend surface tend to occur on “corners” of the top surface UV. This is why @TomTom suggests making the top surface a revolved one.

-Pascal


#15

The problem something isn’t solved… :frowning_face:
Even if I do the top surface round (considerable simplification of smoothing) - defects all the same are present.
2018-02-09_173844
2018-02-09_174031

However if I do a new surface, strongly I distort it - BlendSrf works perfectly well. How so?
2018-02-09_175209
1.3dm (731.0 KB)

The surface of Patch and new surface - they differently interact with smoothing.

@pascal, @davidcockey, @TomTom
Somebody can tell why degree of a surface of Patch has such strong influence? It solution!


2.3dm (5.0 MB)


(Pascal Golay) #16

Hello - I’d say the revolved center surface is the configuration you want, but if you just wildly make random surfaces and want the transition to be smooth, you’ll have to work at it.

Higher degree surfaces are internally smoother, presumably that has some influence on the blend.

-Pascal


#17

At the usual cut surface I don’t notice it (serious influence on smoothing), only the surface from Patch has such trick.