Wish - variable rad surf fillet - G2 blend option

I’m enjoying the new interface and interactivity of FilletSrf but wish this interface and options could be copied over to the VariableFilletSrf command. Most notably a G2 blend option.

Many thanks in advance!

1 Like

Mathematically, a ‘Fillet’ is circular, so the best blending that can be calculated is Tangency (G1). I
f you want G2 Curvature continuity, you need to use a ‘Blend’ instead of a ‘Fillet’.

Indeed, thanks John. I fully understand that.

Take a look at the fillet surf dialog box though. There appears to be a G2 blend option in there. Was this a mistake?

Hi Rob - VariableBlendSrf may do what you need - the Variable*Srf are different from plain FilletSrf in that they require intersecting surfaces - in order to use the FilletEdge style handles etc.
It was not a mistake to put G2 in among thje FilletSrf options, I’m not sure that fillets are as rigidly defined as John suggests… if so, we are breaking the rules.


Thanks Pascal, that’s what I was after!

Can’t find it in the menus though (want to add it to a custom toolbar), or am I not looking hard enough?

The distinction between fillets and blends in Rhino was clear until the introduction of G2 fillets in FilletSrf. Fillets were rational with circular cross sections (constant curvature) and G1 continuity. Blends were non-rational with variable curvature cross sections independent of the continuity.

I tried to argue that introducing G2 as an option in a fillet command was a mistake and would cause confusion but was unsuccessful.

Thanks David.

But I’m not confused about what a fillet is, just confused that the variable blend command had seemed to have disappeared (from the menus that is, forgot that I could type stuff in and search…).

Well done.

I mean they have introduced a hundred new commands for users that want to rearrange the deck chairs on the Titanic and make pretty pictures, but heaven forbid they create any new commands for users who want to create or improve good geometry in Rhino . So yeah. I regard the addition of the G2 option in filletSrf as a prime example of degrading an existing useful geometry creation tools.
But compared to all the other degradation of useful NURBS geometry tools in the last decade, this is not that big a deal. You can always create a macro and put it on a button or alias that avoids the g2 option altogether. Yet another example of user having to be being smarter than the developers.

edit: To be clear. I have no problem with adding the functionality in Rhino that substitutes a G2 blend for a true fillet, but why does that have to be included in the FilletSrf?
There are tons of things that could have been added to FilletSrf to improve it but this is more of an insult than an improvement.

Strange, I find it useful to be able to swap quickly between a fillet and a blend in the same command.

Great job McNeel team!

Really? How does that help you?

If you didn’t want a true fillet then wouldn’t it be more convenient just to have a separate command that would do that straight out? Instead of you having to pick it as an option in some other command?

And before you answer that consider that as a separate command it could have even more options. By cramming it into the filletsrf command they have excluded the possibility of giving you more options so that it would do what you wanted.

Oh! I didn’t realise we’d reached the maximum features that can be squeezed into a dialog box.

Funny, I’m sure ICEM commands have much more functionality.

1 Like

In my opinion it’s all redundant. Two types of command would be enough: surface fillet/blend and fillet/blend edge, within which you can find all the various options (fillet, G1 and G2 blend, with constant and variable progress), as many solid modelers are organized (NX, to name one) .

This issue of fillets should not only be better organized, but it should be developed better (let’s say that the fillet srf tool has significantly improved the fillet capabilities).
Rhino is still too far behind many Cad in this aspect. The “blend edge” is in the stone age. It’s time to give this feature more weight (hopefully this will happen in Rhino 9).

Start rewriting good fillet/blend edge code above all. A couple of developers who work assiduously on this tool, and in three years we will also take out Plasticity & Co.:relieved: :innocent:


Thinking a little more about the new FilletSrf dialog - it’d be even more efficient if we were to have a button at the bottom that keeps the command active but prompts you to select two new surfaces.

Currently it’s just a UI efficiency (one button to remember the location of) but this feature would actually speed up the workflow, as I think Jim was alluding to above.

If you are making 100’s of fillets, a dialog box is not really much of an improvement its an impediment. Its a time waster. Fortunately, the dialog can be turned off. One common complaint about FilletSrf command has been that the geometry creation is too slow, tedious and puts too much of a burden on the user.

As I said before, I have no problem with adding more NURBS geometry creation functionality to Rhino. Lord knows there has been very little of that added in the last 20 years.
So yeah there is plenty of room for new geometry creation tools that work well. And if the only way to add new geometry creation function that works well is to pack it into some other tool then why not shoehorn this functionality into the BlendSrf command?

The Blendsrf command has always needed what the FilletSrf command has - the ability to figure out the best place to put the rails. The Blendsrf command has always needed a way to find a pair of rails that work well with the surfaces being blended and the blend surface. Shoehorning a G2 blend onto a pair of rails that were design specifically for a G1 circular blend is the kind of lazy thinking that the Rhino developers are famous for. Using the rails that were designed for a G1 circular blend only works well in limited circumstances, like when there is symmetry in the shape and curvature on both sides.

The users have always been able to delete a fillet and replace it with a G2 BlendSrf, and when you do it that way you get more options from BlendSrf to control the shape of the blend - so I really don’t see this as breaking new ground. Its just the same old same old that has never worked particularly well.

You have some very valid points here Jim, but saying that, none of them have convinced me that I need multiple buttons for these operations.

That is of course based on my own particular niche usage of Rhino.

It sounds like you also have a particular niche usage which is focused on true radius tangent fillets. I get that, and appreciate it, but I guess McNeel have to try to accommodate us both along with everyone else… that must be tough for them too.

I have no idea what your comment about multiple buttons means. Its you that is advocating for a pop-up dialog loaded with “multiple buttons” for you to push. How did anything I said get interpreted by you as meaning I was the one telling you that you need “multiple buttons”?

Think deeply Jim… maybe not buttons within the pop-up…

You don’t have think much at all to see there is not much difference between multiple buttons in or out of a pop-up dialog. Well except for the fact that the pop-up dialog takes up a space and obscures the view and if the button you want is not in the pop-up dialog you have exit the dialog and find it elsewhere.
I still can’t fathom how how you managed to interpret anything I said as trying to convince you that you need “multiple buttons for these operations”.

It can always be better. Here is a comparison between “Blend surface” in Rhino 8 and Rhino 28. :smiley:

Rhino 8:

Rhino 28:

For more details, read post #61 (58) here:


Very nice!

Shame I won’t be alive to use it… :rofl:

1 Like