Some questions about the Symmetry Mode

#1

Good night all. So, here I was “building” a model from a box primitive, and with the Symmetry Mode on, whenever I clicked on a point, its twin was also selected and allowed me to manipulate them both at once for some fast pacing modelling.
Then, all of sudden, after splitting a polygon on both sides, I wanted to manipulate some points (wich were far from the splitted poly) and the Symmetry Mode althought highlighted, it wasn’t working. The twin point wasn’t selected and pushed me to cut my model in half, work on one side and to mirror it (wich it shown me lots of mistakes I wanted to polish with the Symmetry Mode, but yet again, despite it was a perfect mirror twin of one side, the points weren’t selected in couples - and I already merged the “axis points” to make it a whole and single piece).

And from that, my question rose:
-What are the criterias of working for the Symmetry Mode? When it is active? When it disables itself from working? How can I activate it again when it stops to work?

I like the Symmetry Mode because if allows me to build more accuratelly my models, but, since I don’t know what I’m doing wrong, this tool disables itself at its whim althought it is already highlighted, and I’m unable to get it back to work when I’m advancing on my model :frowning: Any clues?

#2

From what you have described it sounds like you are working correctly with symmetry mode from what I can tell. I always make a habit of splitting my model in half so if my model becomes asymmetrical for any reason I can delete one side and mirror it just like you said you have done. Doing the mirror operation should always restore broken symmetry. You may want to try selecting each of the twin points that are suppossed to be symmetrical and hit i for info on each to check that they are the same distance from the center axis. If they are the same distance then the only other thing I can think of is that symmetry is active on a different axis than the one you are working on.

#3

On a different axis? How can I check that?
Someone told me that I was doing something wrong while creating the mirror image, but if I use the Y axis as the center of the mirror (then merge the points 8mm near of each other for further correction in the middle of the model) but still it says it is asymmetrical because the points aren’t choosen nor pulled in couples. What I’m doing wrong?
The “i” tab says that the point in this shot is 10mm ABOVE the other point in the X axis! IMAGE
How can this happen if I didn’t rotated the starter primitive box I used? It beats me… (and I am getting tired of redoing the box and start again and again, getting the same result… points that aren’t symmetric after mirrowing the piece). I even push the CTRL button to prevent tilting during the mirrowing process, but I guess that’s pretty useless.

Thanks for the drop by NicolasJordan.

PS: I guess I found out my problem. The main points in the middle of the piece aren’t aligned with the Y axis since the start. I don’t have a clue how to align the inner side of the half with the Y axis so the mirror image starts on the Y=0 and then merge there with their proper twin points.

#4

I’ve recently discovered that 9.3 constantly loses symmetry for no apparent reason. All you can do is delete one half of the mesh, ensure that the center is at 0 on X and re-mirror the mesh. It seems that using any of the slicing/cutting tools, even if it is exactly mirrored on +/- X will screw up the symmetry.

I’m not sure if this is a known problem with Newtek?

#5

biliousfrog, it has been always this way… :wink:
Either Knife nor Add Edge don’t support symmetry…
The only cutting tool that’s actively supporting symmetry is EasySplit (added after Warner’s article on SpinQuad http://www.spinquad.com/forums/showthread.php?t=7583&highlight=easysplit)…

#6

Try this: make plain, one edge on x=0.0. Triple it to have two triangles. Mirror and turn on symmetry. Then select triangles (other side will also highlight), and use shift-z for merging polygons… :wink:

It’s really hard to find tools in LW which support their symmetry… :wink:

#7

No I know that, but you used to be able to use those tools on both + &- axis without screwing up the whole mesh. As an example, I’ve been using bandsaw a lot on a recent model. Even if I used exactly the same settings on +X and -X, the symmetry would be lost…I would have to delete half the mesh, run bandsaw and then mirror again. It would even seem to lose symmetry for no apparent reason, I found myself getting really frustrated with it.

#8

It could be bugged dependent on the version you’re using. I too use 9.3, and for no reason the symmetry always flops on me. It’s annoying as hell, but thank God for mirroring.

#9

Then it is a program glitch. Hope it gets fixed on a new version if we clog Newtek’s HQ with mails. Thanks for the data :wink: I suppose is time to get used to this (cannot purchase EasySplit online, sorry with that too but thanks for the add).

Oh, I also use 9.3.1 vr.

#10

unfortunately , simmetry modelling in Lw does need some attention, and yes many tools don’t exploit this mode at all!

. however here ares ome tricks

1) draw your cube primitive and hit  f2 to center it on  world axis( could sound obvious but sometime's not)

2) when i model something, i often try to avoid using a centerline , a row of polys along centerline both for the reason modelling is easier  and because i can insert it later however, with a simple bandsaw command, delete half ,  mirror, etc.

this is just me, btw.

3) to align your points to y axis - (a good method to mirror better)

a) turn off simmetry b) select half of your model c) hit [b]set value [/b]command and set x= 0 d) with polys of your model now still selected hit cancel and erase them d) mirror happily . e) re tunrn on simmetry.

simmetry should work fine again now. if it doesn’t its a bug

4) in previous versions of lw, smooth shift used to break simmetry alot. extender plus seems to work really better. a good trick is however to hit the command and then use move tool to move the new polygons

hope these little hints can  be of help
#11

It is the symmetry tool that’s screwed, or at least a lot of the tools don’t work symmetrically.

Yesterday I needed to create a donut shape so I made a circle (cylinder with no depth) on 0 in Z and offset on X, I then used the lathe tool on 0,0,0 which created a perfectly centered donut shape. But turning on symmetry and selecting polys showed that either the symmetry was off or the symmetry tool doesn’t work because there were several missed polys on -X.

So that’s without any slicing, even a numerically created shape isn’t symmetrical. Try it yourself and see whether it works.

#12

biliousfrog, there is no such thing like Symmetry tool in LightWave… When you press button on bottom panel, it’s turning on/off flag-simple information whether it’s on/off, and EACH individual tool must be rewritten to support it… Basically everything is done in tool which is active, and nothing is done automatically…

Only VirtualMirror does all automatically…

#13

yeah , simmetry in Lw is a mode, not a [i]tool.

[/i] its not a tool like it happens to be in Modo, where you can fix your model if it lost simmetry.

so,in Lw, (but in modo too) the model must be simmetrical before tuning simmetry on

the problem is, many tools don’t work with simmetry (so u have to apply them 2 times , one for each side)
and that some other tools tend to break simmetry in your model so that you’re at last forced to delete half/mirror them to restore them.

I like simmetry mode, because i find it quite intuitive for modelling organics and characters, but i’d personally like that simmetry to be rock solid and that all modelling tools would exploit it.

this would actually speed up modelling process and make it funnier too.:slight_smile:

#14

I understand how the symmetry mode works, my point was that you’d expect to be able to create a symmetrical object using a standard Lightwave tool such as lathe.

Also, it shouldn’t matter whether I knife a perfectly mirrored object in exactly the same +/- X co-ordinates or cut one side then mirror it at the origin, the points and polys should still be symmetrical and in exactly the same place whatever way you do it.

The point that we’re trying to make is that the symmetry “mode” doesn’t work as well as it did or something else has gone wrong.

All solutions so far have just been to mirror everything…which takes us back to before the symmetry tool was introduced (6.5?).

#15

yes you’re right. i guess that Lw produces small errors when making lathe ?. remirroring the mesh does work, but’s not a solution. it’s clear that a numerical generated object should be actually perfect and simmetrical.

#16

This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.