Hello - if your 5 by 5 surface is degree 3, then you will indeed get multiple Beziers out but if you can make them degree 4 by 4 and 5 points, your troubles should be gone… is that a possibility in your process?
-Pascal
Hi,
no, we´ve got polysurfaces made from single 4x4 deg3 (“Bezier”) Surfaces, that we want to convert into one single nurbs surface.
Hi Joerg -
OK. I was wondering really about this earlier part of the process:
I thought if you could start with degree 4 by 4…
-Pascal
Let’s say I have 4 simple Bezier deg 3 Surface with each 4 control points=16x4 =64 control points, and I would love to get the same 5x5 Bspline Surface with 25 controlpoints. The idea is to reduce the number of control points, and get the"free" continuity between spans by doing this.
Assume that I have a polysurface which consists of 50 single 4x4 cp “bezier” surfaces. I want to convert to one single nurbs surface, of course without fully multiple knots…
By the way: What sense is behind a mergesrf command, that makes fully multiple knots?
Who needs that? IfI want to make a kink, I would use insertkink. If I merge surfaces I generally don´t want to make a kink-or?
By the way: Why don´t we have a removekink command?
Hi Joerg -
MergeSrf without fully multiple knots (Smooth=Yes) changes the shape a little or a lot, depending…
RemoveKnot = ‘RemoveKink.’
-Pascal
Hi Joerg -
MergeSrf without fully multiple knots (Smooth=Yes) changes the shape a little or a lot, depending…
Yes, so that doesn´t make any sense.
Yes, that works for simple situations, but if you merge more surfaces, how do you know which knots must be removed?
Mergesrf (smooth=no) does not change the shape. If you need the shape to stay exactly the same that is the option to use. In some cases (actually many cases) you can remove the extra knots and the shape will still remain unchanged.
The solution is to have a Rhino function that can identify when the extra knots are not needed to maintain the shape and remove them. The fact that Rhino does not have this capability is a serious enough flaw that it could be called a bug.
To me, it looks more like you are just trying to avoid dealing with the problem.
It is irrelevant what caused the problem.; You should be looking at how to fix it.
This is a serious problem in Rhino. It causes numerous other problems. One big problem is when you end up with fully multiple knots in edge definitions.
If Rhino could figure out how to clean up edges getting rid of the erroneous fully-multiple knots users would have a lot fewer problems with naked edges. Almost all the orphaned naked edges that Rhino so commonly produces are caused by fully multiple knots in edge segments that should have no fully multiple knots.
I think David already explained this but the steps are to use mergersrf [smooth=No] and then remove 2 knots in each direction where there are fully multiple knots. After knot removal the shape may stay the same or it may change.
1 Like
Hi @jim,
Lol, yeah. If I were just trying to avoid dealing with the problem, then I have ignored this topic altogether.
The reason I am asking questions is because I want to understand the problem. I should probably know about it, but I don’t…
– Dale
If you had done nothing from the start, I would have no objection
But this business of asking nit picking questions and then when the user goes to great lengths to explain and then doing nothing that I find annoying.
1 Like
The fact that Rhino does not have this capability is a serious enough flaw that it could be called a bug.
Is there another software that does? Would good to see how alternative options work and how well they are at identifying needed vs not needed knots.
Actually, it is easier for a software developer to ask a question, than it is to spend weeks or months trying to code something for a problem that does not affect the larger user community.
So, lighten up a bit, and appreciate the fact that you are actually talking to the people who make this thing work, do show some respect, and keep focused on the topic at hand.
To Michael Pryor’s point, how does other software handle what is, to me, this obscure issue?
Most other software are not tripped up as badly by this as Rhino is.
That’s probably mostly because Rhino geometry has no history. But
it is also because they deal with it rather than persistently ignoring it.
Let me give you an example. Look at the thread “Fillet edge behavious”
That’s an example of Rhino adding an unwanted fully multiple knot to an edge definition and then the result is naked edges when join is attempted.
Pascal offered the solution to basically start over and recreate all the trim boundaries.
Other programs can always do as Pascal suggested (use history to start from scratch)
Rhino does have the capability of removing unnecessary fully multiple knots from lines and arcs. That’s good, but it should be expanded to include spline curves and surfaces . There should also be the added capability of identifying and removing any knot that is not needed to define the shape.
You couldn’t be more wrong. This is something that has plagued the larger user community for years. And it would not take very much developer effort to fix it .
And as I pointed out it is completely irrelevant how unwanted and unneeded knots got there. McNeel has all the info they need to fix this. What it seems they don’t have is anybody that cares.
Which software, I’d like to test it.
Hallo @dale @pascal and @Michael_Pryor
I am not a developer, so I can’t imagine what problems in detail are to be solved.
But I believe that It might not be that difficult to detect fully multiple knots. If these would be color coded one could see, which knots have to be removed by removeknot. As an Alternative: wouldn’t it be possible to script a command that detects all fully multiple knots and deletes them?
best regards
Hi Jeorg - this is no solution but it might be of some use, if I’ve done it right - this script marks discontinuities and fully multiple knots - it is not a display thing, it drops a colored isocurve at the locations it finds.
To use the Python script use RunPythonScript, or a macro:
_-RunPythonScript "Full path to py file inside double-quotes"
@JoergH - there were a couple of typos in that script; it is updated now.
MarkSurfaceDiscontinuities.py (11.3 KB)
-Pascal
Great,
I’m not in front of my computer right now, will try it later.
Thanks a lot Pascal. Much better than expected. More than great customer service here.
Even better. Could you explain what samples does?
Hi Jeorg - if you ask for Details=Yes, then the thing tries to locate the maximum discontinuity and mark it - but it needs to do this at discreet locations - hence the samples - more will find a more accurate location for the maximum, but be slower.
-Pascal