I don’t know why or what but Rhino is amazingly bad at building this surface in a clean manner.
I’m talking about the blend of course.
It’s hard to tell from that image what issue you are having.
1 Like
When you post something like that with zero context to enable anyone to possibly know what you’re talking about let alone help, I’m just gonna say “sounds like a skill issue.” 
4 Likes
A bit more context.
What I meant is that Rhino needs a LOT of hand-holding to generate decent blends and fillets (a well known limitation btw). Left side are the result of blendSrf in Rhino vs properly hand constructed blends.
I mean we should at least have an (one click) option to align the blend spans to the incoming surfaces.
Also in the top left corner you can see that the blend surface end is not even planar, since you cannot split the planar face directly, requiering more steps.
Okay but in the situations where you want “B” there…I would just do the whole thing as one BlendSrf–you’ve already done that 90% of the way why not the rest?–and you’ve built the larger ‘fillet’ surfaces differently to heighten the contrast and get a trimmed edge on your “wrong” way example so it seems a bit contrived.
It looks like you’re trying to make an effect like Apple’s (formerly) signature puffy surfaces. You know how they did those, in Alias? Brute-force point-pushing until it’s perfect, they don’t use anything so pedestrian as “surfacing tools.” And I mean it’s true, at this point if I want a sphere I just point-edit a plane.
I’m not sure about the way Apple did it. But at this point it is not a debate on how Apple achieved their iconic surface. It is a debate about how lacking and primite the fillet and blend tools are in Rhino.
From the fact that they generate unnecessarily complex surfaces, to the fact that they do not offer any automatic options for span (profile) alignment. Also their very bad matchSrf command, if you even want to touch that and over-complicate your life further. I’m not even talking about how primitive these are for final surface generation, fabrication. Just for quick concepting.
There are numerous mentions about this, even on this forum, and it just feels so unnecessary after all this time. So again, I’m not trying to say that Rhino’s tools are incapable of allowing to achieve great results, I’m just pointing out how for iteration, they demand a lot unnecessary extra steps from you.
Rhino’s “Blend surface” tool still lacks a proper U or V isocurve alignment option. It also lacks alignment to the side edges A and/or B individually. The same goes for the “Match surface”, too.
To overcome those shortcomings, you can build a couple of blend curves between the primary surfaces and then use “Edge surface” to get a clean sueface blend. You may need to match its sides to the primary surfaces to achieve G2 continuity.
But the problem is your example here is not proving anything. It’s fake You built it 2 different ways, and your “right” way is set up such that there’s no point in even using blendSrf tot do it, it’s showing the options you want in blendsrf would only make the perfect results you want in situations where…why is that even a separate surface? There are a dozen other ways to solve that.
BlendSrf is awesome, for what it’s for. MatchSrf is fine, I’ve tried to brute-force make cleaner matches than it can with hours of point-pushing, and the truth is again unless I’m in a situation where there’s little point in the joint even existing, where you’re not lining up perfectly matched surfaces that could as easily be one surface, it’s very hard to actually do a lot better.
My point about Apple is that the more you know about this stuff, the fewer “tools” you use.
But my point was different, in that in Rhino to achieve ‘decent’ results you have to go through a lot of unnecessary steps. In this specific case where you blend two profiles against a planar top face:
- take extra steps to clean the input surfaces
- take extra time to manually adjust the span / profile / isocurve alignment
- take extra time to clean the resulting blend if it adds unnecessary spans
- take time to align the end of the blend with the planar face for trimming / joining
I haven’t used Alias in a very long time, but I remember you could input what degree and how many spans you wanted the result to have, you had options to proper align the spans and ends individually etc… everything in one tool. With Rhino you just have to go through a lot of unnecessary steps, and I want to emphasize this: unnecessary.
And just to illustrate the meaning of unnecessary I want to point out again the destructing workflow of adding fillets to solids in Rhino. The fact that you still have to save and hide your original solid, and recreate all fillets if you want to change just one radii is completely unnecessary.
That’s exactly my point. Everyone and myself included can probably live without them, but I just can’t not ask myself why there aren’t more parametric tools implement by this point in time in Rhino. Specifically for fillets and blends which just eat up so much time to adjust.
So you expect Rhino to take trimmed surfaces and replace them with untrimmed surfaces with a totally different topology? That’s what your fake example shows. It’s utter nonsense that you’re deflecting now with unrelated complaints about there being no history on fillets (which actually there is.) I’m done here.
This has always been about the blendSrf generated and the blendSrf or fillet tool in Rhino. I was pointing out that in Alias you have control over the structure of the blend irrespective of it being generated off a trim edge or surface edge. Also you have control on the ends alignment as well as many other options.
Good luck!
https://discourse.mcneel.com/uploads/short-url/kqGdxelmPZIa2UA4lyvVyR4CmsX.3dm
Misleading preview:
Resulting surface:
The simplest possible degree 1 surface (bottom left):
Overly-complicated output surface:
Future “Blend surface” in Rhino 27 (maybe?):
Examples taken from this old topic from the Rhino 6 days:
3 Likes
It’s insane how old that thread is. I agree with everything said there. Too bad Mcneel won’t even take a look inside there or aknowledge these shortcomings. It’s interesting to note also that a few tools and plugins oriented towards surfacing, or bringing more parametric options to Rhino, have since disappeared or have stopped being developed. Or at least that’s how it feels.
2 Likes
1 Like
@John_Kunst do you mind sharing your .3dm file with me? Or maybe you want to play with a trial of the cyberstrak plugin yourself (yes, I know, everyone wants this in basic Rhino ;-( )?
@Peter_Salzman FYI
In that specific thread, lots of YT’s have been logged, so I think the acknowledging part is not the problem.
That’s true for TSplines and VSR, which were both bought by Autodesk, and later killed by Autodesk. There is still XNurbs and currently Cyberstrak in development.
1 Like
Hi, Michael! I will post a simple scene. But for the trial can I try it just by downloading or do I need a special license?
@John_Kunst Just downloading and installing it will work.