Jump to content
Hash, Inc. Forums
itsjustme

v18 Smartskin Mirror issue

Recommended Posts

I've been having a very hard time getting Smartskin to mirror correctly. So, I ended up making a simple model in v16, v17 and v18 to see if and/or when things go wrong for me.

 

On my end, the only way I can get Smartskin to mirror correctly is to make the model in v16...unless I'm missing something.

 

Here is a quick video showing the included Project with similar models made in v16, v17 and v18. Each model has a Smartskin on the right side with an Action for each bending the bone that is affected and the left side version of that bone rotated similarly. I right mouse-click on each model in the PWS and mirror the existing Smartskin, then check each Action to see if the mirroring worked. Only the model made in v16 worked for me...it seems like the spline ring with a 50/50 CP Weighting isn't working. Is anyone else having this problem?

v18_SS_07_27_2016_test.mp4

v18_SS_test.prj

Share this post


Link to post
Share on other sites

When I try your test project, the Smartskin does indeed fail to mirror correctly.

Share this post


Link to post
Share on other sites

When I try your test project, the Smartskin does indeed fail to mirror correctly.

 

If you make a simple model in v18, do you have similar results, Robert? Or does it only happen with this test project?

Share this post


Link to post
Share on other sites

 

When I try your test project, the Smartskin does indeed fail to mirror correctly.

 

If you make a simple model in v18, do you have similar results, Robert? Or does it only happen with this test project?

 

 

 

Yes it happens in my case too, and I found another bug while I was making it.

 

It is interesting that even if i resave your v16 and v18 models from v18 and reload them, then do the mirror smartskin, the different results persist.

 

 

I notice that your v18 model has a "CPs without mirror" group that the others do not.

 

 

An interesting experiment I have not tried yet would be to carefully remake the models using snap to grid so that they could be absolutely identical in CP and bone placement even though they were made in different versions... then compare the two in a text editor to see what is still different about them.

Share this post


Link to post
Share on other sites

I also tried to "Paste Mirrored" to fix the mirrored Smartskin, but that didn't work either.

 

I just forgot to delete the "CP's without mirror" group in the v18 model...the center spline rings aren't assigned in any of them and a group like that is made for unassigned CP's after using the MirrorBone plugin.

Share this post


Link to post
Share on other sites

It looks like it may be an intermittent problem. I decided to model the test model in v16 and v18 again so that I could make them as close to identical as I could, but mirroring Smartskin worked without a problem on both models. It's still a problem with the original test models and the version of Sam I'm working on though.

Share this post


Link to post
Share on other sites

Hello,

I do not know if it will help: I compared the files and found that the is not in the same . This does not seem normal, and i don't know why this differences.

weighted.jpg

Share this post


Link to post
Share on other sites

It looks like it may be an intermittent problem. I decided to model the test model in v16 and v18 again so that I could make them as close to identical as I could, but mirroring Smartskin worked without a problem on both models.

 

 

Yeah, I just tried that also and both worked. :facepalm:

Share this post


Link to post
Share on other sites

My current theory. It happens when i do this...

 

attachicon.gifclip3996IncorrectSmartSkinMirror.mov

 

 

 

 

Yessir, the only thing I do differently is I don't have the MirrorBone plugin clone the weighting...I just weight everything once the bones are in place.

 

I'll have to try things in a few different orders.

Share this post


Link to post
Share on other sites

Here's another weird bit I found. I took a model that I knew would break when I mirrored the right Smartskin to the left side, deleted the Smartskins, Smartskinned the left side and mirrored the Smartskin to the right side...the problem jumped to the other side.

 

 

 

SS_test_07_30_2016.mp4

SS_test_07_30_2016.prj

Share this post


Link to post
Share on other sites

I did some more messing with the last posted project and realized that it's the 'X' axis that is the problem with the Smartskin mirroring here. In the attached video, I move the affected spline ring "-2" on the 'X' axis, but the other side was only moved "1" on the 'X' axis for that Smartskin, so the left side moved the wrong direction when mirroring. I still haven't found what causes the problem though....I'll probably mess with this some more because the character I'm working on (Sam) has this problem.

SS_test_08_01_2016.mp4

Share this post


Link to post
Share on other sites

I found this old thread the other day while looking for solutions to this issue in version 19.0e2. I've been able to reliably reproduce the issue with weighted cps and mirrored SmartSkins on a character I'm rigging right now. Do we know any more about what's been causing this behavior and how to avoid it? I have a forearm SmartSkin that this is affecting with about 13 different rotations keyed. Some rotations are working okay, some are definitely not, and it is always centered on the handful of cps that I assigned weights to. Fan bones? No problem!

 

Bother. <_>

 

Chris

Screen Shot 2018-07-04 at 7.22.08 PM.png

Share this post


Link to post
Share on other sites

I found this old thread the other day while looking for solutions to this issue in version 19.0e2. I've been able to reliably reproduce the issue with weighted cps and mirrored SmartSkins on a character I'm rigging right now. Do we know any more about what's been causing this behavior and how to avoid it? I have a forearm SmartSkin that this is affecting with about 13 different rotations keyed. Some rotations are working okay, some are definitely not, and it is always centered on the handful of cps that I assigned weights to. Fan bones? No problem!

 

Bother. <_>

 

Chris

 

 

Just to try... you might experiment with upping Tools>Options>Modeling>Mirror Mode Tolerance

 

I'm sure I don't need to tell you to save your work beforehand and to not save over it with experiments.

Share this post


Link to post
Share on other sites

Good suggestion! I made a couple of tweaks to the Mirror Mode tolerance, ramping it up from 0.05cm to 0.15cm and also bumping the Group Snap tolerance from 0.5cm to 0.75cm. I deleted the left side SmartSkins that had been created by mirroring and tried the Mirror All SmartSkins operation again but got the same results.

 

...it's at this point that I discovered something that made me feel pretty foolish: the issues I'm having are not entirely the fault of SmartSkins. I had a rogue fan bone in the elbow that was storing the roll value of the forearm when the forearm was twisting. :facepalm: I unticked the Store Roll option for the fan bone's "orient like" constraint and found that things were behaving a LOT more predictably even before SmartSkins came into play.

 

To test it, I manually created a SmartSkin for the left forearm and began copy-pasting mirrored keyframes from the right side. PRESTO! Mirroring the SmartSkin via the "Mirror All" command now works as well. Now if I can just remember to build my fan constraints consistently from right to left, it looks like I'll be all set!

 

Thanks, Robert!

 

Screen Shot 2018-07-05 at 9.40.21 PM.png

2ArmsSmartSkin-test.mov

Share this post


Link to post
Share on other sites
Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


×
×
  • Create New...