Bug in SubD Reflection when mirroring

Hi,

I just encountered a problem with SubD-Reflection several times:

I have a SubD-Box with a Reflection around the Z axis, it’s offset along X one Meter.
If I now mirror that SubD-Box on X, I get a SubD object that shows the reflection of the original object visually, but the reflection does not work.
I have to remove and re-apply the reflection on the mirrored object for it to work.

What I expect to happen is either the reflection to be removed from the mirror or preferably the reflection to work after the mirroring on both parts, the original and the mirrored one.

And I want to +1 the requests for a two-way reflection: I’m working on the design of tables ATM and would love to be able to have a two-way reflection for the 4 feet.

Also, it’s weird that the left and right mouse do the same with the Reflection button in the SubD Toolbar. The left mouse button behaviour is fine, but the right should be an instant removal of the Reflection without further ado if the object is already selected.
I saw that this was commented on before but wanted to +1 it. :wink:

Windows 8.1 x64, Rhino 7 SR5 2021-3-15 (Rhino 7, 7.5.21074.17001)

Thank you!

Cheers,

Tom

I have another report of this, but here, the macro is correct - can you please run ToolbarReset (with only one Rhino open) and then close and reopen Rhino - does that work better?

The toolbar macro should be

! _Reflect _Pause _RemoveExistingReflectSymmetry

-Pascal

1 Like

Thanks, here it was “! _Reflect _Pause _RemoveExistingSymmetry”
Your version works, thanks a lot!

Cheers,

Tom

Hi Tom - ToolbarReset should sort that out - you may want to run that anyway, if you are using an un-customized default rui (toolbar) file, just to make sure you have the most recent.

-Pascal

1 Like

I did some customisations so I was reluctant to run the reset, but good to know about it.
I assume that at one point since 7.0 this was changed? I installed the Demo version rather early after release and only bought the update recently, so I may have had an old version there.

Would you have any comment on the problem with mirrored reflections or is that another department?

Thanks and cheers,

Tom

This department, I just need to test it and see exactly what you are asking…
@Thomas_Helzle - there can only be one reflection on a SubD at a time, is what you are saying, correct?
-Pascal

1 Like

Yes, but that is more the FR part of my post, I’d like to be able to reflect both the y and x axis for instance to be able to edit the four feet of my table at once.

The bug I see is shown in the file attached.
I have this SubD object with a reflection:

I then mirror this object like so:


It looks as if both object have the reflection, right?

But it isn’t working. If I select the three points at the back edge on both objects and drag up, only the original object reflection works, the mirrored object - although showing the reflection mark - does not work and the edit is not reflected.

So in the not-so-nice - but somehow correct - case, the mirrored object should have no reflection marking.
What I would expect and want though is, that the visual result is correct and both objects have the reflection working after the mirror.

As it is now, it’s neither here nor there.

Here is the file:
SubD_Reflect_Mirror_Issue.3dm (125.2 KB)

I hope it’s clearer now :wink:

Cheers,

Tom

Hi Tom - yes, I see, thanks - the mirrored object should not show mirroring. Mirroring with History is what I think will help in your example.

RH-63415 Reflect: Reflect + Mirror

-Pascal

1 Like

Okay, will have to look into history then - never used it before. I hope it works, I have to do this all the time ATM since I can’t reflect in two directions at once…

But in above example I can easily remove the reflections from both objects and then apply a reflection to both at once and it works, so I don’t think it’s unreasonable to expect that the mirror keeps the reflection?

Cheers and thanks,

Tom

Edit: Interesting. History seems to work, have to see how robust it is for my case when editing further. Thanks!

@pascal While I have your ear on reflections:
Is there a specific reason why I can’t have creases along the reflection line?
I’m working on a figure where the central line should be creased, but each time I create a reflection on the model, I lose the crease along the centre line…

That’s a bummer, since I have to remove the reflection, apply the crease to see how it looks, re-apply the reflection, rinse and repeat…

I can’t see a reason why that should be the case?

Thanks and cheers,

Tom

@pascal Sorry to bother you, but above problem with not being able to have sharp creases on the mirror line is a major annoyance here, since my current model relies heavily on exactly that.
Is this an easy fix maybe? It would make my life so much better! :slight_smile:

Thank you,

Tom

I still have the problem in Rhino 8 WIP that I can not have a sharp edge crease on the mirror line.
This is really bad since many of my designs have such sharp creases exactly there.
Any way this can be fixed finally?

Thanks!

Tom

Thanks, added here - RH-73147 Reflect: Support creases on the reflection plane

-Pascal

1 Like

Thank you very much.
Just to add to your bug:
There are two scenarios I fight with:

a.) The first is, if I have a sharp crease on the symmetry line BEFORE I create the mirroring, that crease is removed (both on points and edges).
So it’s an actually destructive behaviour.

b.) When I want to create new sharp crease on a point or edge that is part of the symmetry plane while a symmetry mirror already exists, I simply can’t.

ATM my only option is to either not use symmetry at all or work around it with adding more edges to create sharper creases, like we did it in Lightwave 6 in ~2000 before there was edge creasing/weighting - but this defeats the purpose of the edge creases option and is much more troublesome to work with, since the very-close-together edges are quite hard to integrate without getting nasty corners somewhere else.
Or I have to work with symmetry, remove it, create the creases to see how it looks, re-symmetrise and lose the creases, continue modelling, rinse and repeat, which is super frustrating and feels idiotic, but which I do sometimes anyway, since the alternatives are often even more tedious.

What I would like to see is that it’s basically transparent.
The sharp creases should work, no matter if there is symmetry or not and should survive the toggling of symmetry as well, both on edges and points.

I do not know if this is a bug/oversight or if it points to some kind of problem.
I’d be happy to discuss this further in detail if the latter.

Thank you very much.

Tom

P.S. … and I REALLY hope that we will see edge weighting rather sooner than later. That is such an integral part of what makes SubDs useful for organic design!