Jump to content

Project Bento Feedback Thread


Linden Lab
 Share

You are about to reply to a thread that has been inactive for 1276 days.

Please take a moment to consider if this thread is worth bumping.

Recommended Posts

Hi, Medhue;

I wonder if having bones at LipLowerCenter and LipUpperCenter would make it easier to get more decent mouth animations. It appears to me that the mouth is a super important animation spot and you can easily get from acceptable to creepy if you do not take care.

For example it appears to me like the mouth tends to become a bit "squarish" even when you move the upper/lower lip bones closer to the center.So i propose to add 2 lip bones on the vertical center line of the face.

And when thinking of creatures then i believe that adding 3 extra structural bones to the head gives some interesting freedom:



[Edit:] (I replaced the above image because in the first one i unintentionally parented the lower lip bones to the wrong bone) The intermediate structure bones (lower,center,upper) also clean up the head structure a bit. But my favorite change is still to add 3 extra bones on the face vertical center line:



 

From all what i tried so far i found that having those center bones in place has 2 advantages:

  1. The automatic weight creation creates much more useful results
  2. The squarish mouth animations can be fully avoided

So if you'd ask me, i would do the following 3 changes (ordered by "most wanted"):

 

  1. Add visual center bones (mFaceUpperLipCenter, mFaceLowerLipCenter, mFaceForeheadCenter
  2. Add Extra bones for the ears (flapping ears)
  3. Add structural BoneRoots (mFaceUpperRoot, mFaceCenterRoot, mFaceLowerRoot)

When asked which bones could be removed i tend to say, the lower cheak bones might be less important (could be replaced by the lip corner bones if needed)

  • Like 1
Link to comment
Share on other sites

  • Lindens

I'm not sure how the new lower face root would work with the jaw. Originally we had jaw as the parent of both tongue and lower lip. Is it now split so tongue is attached to jaw, and lower lips are attached to lower face root? If so, why would that be preferable?

Link to comment
Share on other sites


Gaia Clary wrote:

Hi, Medhue;

I wonder if having bones at LipLowerCenter and LipUpperCenter would make it easier to get more decent mouth animations. It appears to me that the mouth is a super important animation spot and you can easily get from acceptable to creepy if you do not take care.

For example it appears to me like the mouth tends to become a bit "squarish" even when you move the upper/lower lip bones closer to the center.So i propose to add 2 lip bones on the vertical center line of the face.

 

We've talked about it but, I'll just say here that I think this just comes down to how far LL is willing to go. Of course I want center lip bones, but I'm kind of still in shock I have face bones to play with in the first place. lol

 


Gaia Clary wrote:

And when thinking of creatures then i believe that adding 3 extra structural bones to the head gives some interesting freedom:

[Edit:] (I replaced the above image because in the first one i unintentionally parented the lower lip bones to the wrong bone)
The intermediate structure bones (lower,center,upper) also clean up the head structure a bit. But my favorite change is still to add 3 extra bones on the face vertical center line:

I can definitely see how this can be handy, but at the same time, we have bone translation, and we are Blender users, so with all the tools we have, this can be mimiced. At the same time tho, I will never turn down more bones. I'll find something to do with them.

On LL's end, I think your thoughts on Bone Constraints within SL is something that should be persued. It adds so much functionality, especially considering how we can use SL internal animation with our creations, like making the head move with the mouse. Who knows what else it could lead to?

 


Gaia Clary wrote:

 

 

From all what i tried so far i found that having those center bones in place has 2 advantages:
  1. The automatic weight creation creates much more useful results
  2. The squarish mouth animations can be fully avoided

So if you'd ask me, i would do the following 3 changes (ordered by "most wanted"):

 
  1. Add visual center bones (mFaceUpperLipCenter, mFaceLowerLipCenter, mFaceForeheadCenter
  2. Add Extra bones for the ears (flapping ears)
  3. Add structural BoneRoots (mFaceUpperRoot, mFaceCenterRoot, mFaceLowerRoot)

When asked which bones could be removed i tend to say, the lower cheak bones might be less important (could be replaced by the lip corner bones if needed)

I pretty much agree with you, except the part about the lower cheek bone. From the original set, I don't think I could do without any. I think it was a well thought out, least amount of bones possible to get anything decent. If we got even just the first in your list, heck, I'd be happy. Plus, I'm also a business man, and I want things done quickly and out the door for people to start using things, creators making money, and LL making money on it all as soon as possible. So, all that being said, I'm ready to go with what we have if that means we can start working on actual things that make money. lol

 

Here's a tangent, but relevant. The facial bones parented to the skull bone, I really don't have any issues with. Wish we didn't have to, but I do understand the issue fully, and why. Again tho, that said, will I be updating my wolf to utilize that extra head bone? No, I will not be, because I do like to make all my avatar's heads move with the mouse, and then try and blend in the body movement, when that works best. The neck on my wolf is not really an issue, at all. I have rigged a few horses in my time tho, so I do understand the issue pretty well. 1 tip for horse riggers, which I learned in the process of rigging them, is that, on a horse, it is best to run the neck bone near the underside of the neck. This produces better horse neck movements, and is actually how a real horse's neck bones are, giving them that range to reach the ground.

Link to comment
Share on other sites

Hi, Medhue;

Of course the cheeck bones are of value. My reasoning was more like this:

First: I believe the mouth and the eyes are the main focus on the face when we take animation into account. So i love the idea to get more bones for the lips to get the ability to make realy amazing animations there.

Then: The structural bones in the middle of the head are more than just cosmetic. Think of a creature where mouth and eyes move independently. Yes we can use translation to move the bones wherever we want, but having the ability to animate a complete subgroup of face bones sounds more practical to me.

Finally: I know that the number of bones can not be infinite. And we must keep the balance. And since we actually can move the bones around in animation AND by using joint positions, the cheeck bones and the lip corner bones could possibly be candidates for reuse.

However, all of this is still playing with ideas. I hope that we can get something more substantial out in the next few weeks :matte-motes-sunglasses-3:

oh, btw: If you are playing with new bones then Avastar-2 has a avatar_skeleton.xml exporter

File -> Export -> SL Avatar Skeleton (xml)

as far as i have seen the workflow is this:

 

  • Create a new Avastar (rig only is sufficient)
  • Take care to set the appearance to SL Rest pose (the white stickman icon in the Appearance panel)
  • Edit the (green) control bones
  • Snap SL to Control Rig
  • Export as modified avatar skeleton file

And for creating a test character from a modified skeleton file:

 

  • File -> Import -> SL Avatar Skeleton

I might look into that later and add some better documentation to this. Some parts of this importer/exporter  pair seem to be not optimal yet :)

Link to comment
Share on other sites

 Hello,

I made more experiments with the new translations posibilities on my "in progress" mesh head and here is the result in video ;-) :

https://gyazo.com/75182df6d763c8c4043b02a32c0056b8

I tend to agree with Medhue about the mFaceCheekUpperOuter bones. They would be much more useful if placed much lower (as suggested).

As Gaia indicated, the mouth and the eyes are the main focus and the two additional bones proposed for the lips would for sure allow more munances in the mouth animations. 

But while experimenting, I rapidly understood that one big difficulty would be the weight limit to 4 bones !
This drastically increases the complexity of the rigging (with some many close bones to deal with)...
So I would suggest to raise or best remove this limit.

Link to comment
Share on other sites

Are you guys sure that you want the facial bones parented to skull rather than head bone?

Remember that mSkull bone moves upward with some sliders. I dont think people would want their eyes, mouth and such move upward while the rest of head remains still lol. I think that just having another bone that could holds the facial bones just as mHead children could be the best. It could also work as extension for neck while not breaking anything else.

  • Like 1
Link to comment
Share on other sites


Gael Streeter wrote:

But while experimenting, I rapidly understood that one big difficulty would be the
weight limit to 4 bones
!

This drastically increases the complexity of the rigging (with some many close bones to deal with)...

So I would suggest to raise or best remove this limit.

Honestly i do not exactly understand the problem with the limits. Having more bones does not imply that you have to use more weights as well. Actually the vast majority of vertices in a typicall rigged mesh has typically only 1 or 2 weights. A good example is the default Avatar. (And yes, i know what some will answer now. But the weight problems of the default Avatar are not related to the number of allowed weights.)

Well, when you also use Fitted mesh bones then the weightcount can double (depending on how you actually do your weighting). However when your typical maximum weightcount is 2 for the classic mesh weighting (without collision volume bones), then with fitted mesh it becomes 4 and that is still in the limits...

Ok, here are issues when you take fitted mesh into account. But those issues are mostly related to the physics volume bones on the upper body. There we quickly can get out of weights.

Since neither face nor hands have fitted mesh bones, i see no serious issue there. However i learned to never become too strict as there is always an exception for everything :matte-motes-sunglasses-3: So please can you give some real examples where the weight limits are too small on the face?

Link to comment
Share on other sites


Gaia Clary wrote:


Since neither face nor hands have fitted mesh bones, i see no serious issue there. However i learned to never become too strict as there is always an exception for everything :matte-motes-sunglasses-3: So please can you give some real examples where the weight limits are too small on the face?

From my experience, a raise to 5 is not without warrant. You have talked about when Fitted mesh is involved, and the complexity there. Yeah, we don't have collision bones for the face or fingers........yet. Vir has already stated tho that this may be in the works. For the most part, I agree with you Gaia, that 4 is good most of the time. There are quite a few times tho, that I have run into verts where I have no real good option which bone I remove to get to that limit of 4. When this happens, I cringe, because I know that vert is going to have issues, and not move the way it really should move. I'm definitely not for unlimit bones weights, but I think a 5 bone limit is much more reasonable, considering what we are dealing with.

Link to comment
Share on other sites


Kitsune Shan wrote:

Are you guys sure that you want the facial bones parented to skull rather than head bone?

Remember that mSkull bone moves upward with some sliders. I dont think people would want their eyes, mouth and such move upward while the rest of head remains still lol. I think that just having another bone that could holds the facial bones just as mHead children could be the best. It could also work as extension for neck while not breaking anything else.

Well a......, if that is the case, then we can't do it that way. I really have no idea. I don't have unlimited time to test everything, and I'm already neglecting my clients enough because of my obsession with the bento bones. I do know that someone else talked about how moving the skull bone moves where the bubble chat and name hover box above our heads, which is handy to know, and moving that is needed in some cases. So, yeah, I'm for accommedating for the extra neck, but not if it isn't fully tested, and interfers with other things.

Link to comment
Share on other sites


Medhue Simoni wrote:

Here's a tangent, but relevant. The facial bones parented to the skull bone, I really don't have any issues with. Wish we didn't have to, but I do understand the issue fully, and why. Again tho, that said, will I be updating my wolf to utilize that extra head bone? No, I will not be, because I do like to make all my avatar's heads move with the mouse, and then try and blend in the body movement, when that works best. The neck on my wolf is not really an issue, at all.


The plural of anecdote is not data. Humans, canines, and many other creatures have short necks without many vertebrae, and therefore look perfectly normal with only a single neck bone. But that doesn't mean that other creatures would not benefit from the option of an extra bone. It's dangerous at this stage of the project to proclaim that anything isn't needed or won't be used, especially when you can see what value it might offer to someone else. Obviously any new bones that come out of the bento project, including the bones already in the bento skeleton, will be properly tested to assure that they don't break existing content or in any way affect those creators who choose not to use them.

  • Like 2
Link to comment
Share on other sites

Hehe, Im sorry but the theory doesnt work here.

I can see why you think that but its very far from reality.

As I mentioned before, Im professional rigger (I would better use skinner as rig is something totally different but we know how this workwithin SL lol) and been working for Blueberry since more than a year. I have also worked for others popular brands like Reign, Addams and others. I have even made the official dev kit for Maitreya Lara so I can tell you pretty well how the limit affects the rigs.

Lets first state something. Yes, "technically" 4 is a good value. The problem comes from avatars and fitted mesh. Fitted mesh doesnt really act like it should when you move sliders. If you increase, lets say, legs witdh certain amount, you get a totally different value making use of only fitted mesh bones. This situation forced mesh body creators to naively make use of both, fitted and normal bones, to achieve good deformation results.

Unfortunately, this leads to too much weights everytime to want to rig something to those meshes. While this isnt really an issue from LL, its a reality that we cant ask to those creators to reduce the weights of those bodies that already are top on sales. So we need a solution to this, even more if you plan to rig faces as 4 is really too short limit.

If you want an example, try to draw this in your mind, then you will see it clearly.

Imagine you try to rig a skirt for a mesh body. Its legs uses 3 to 4 bones each (L_UPPER_LEG. PELVIS, mHipLeft, mPelvis, BELLY,etc..) then try to smooth those weights so the skirt acts like a skirt instead of like a glitched shorts and you will see how you need a MINIMUM of 6 bones per vertex.

We really need to introduce this change and now is the best oportunity before people start to realize and complain about it once Bento is realeased since you will see even more openly this scenario when rigging faces.

  • Like 2
Link to comment
Share on other sites

Well, i realy want to understand what the problems are. But i can't do that only in my mind. It would be very helpful if someone provided a working example of a setup where more than 4 bones are unavoidable.

Maybe this is asking for too much. But i believe that Linden Lab wants to see such examples anyways if we want to get the limits raised. So if you(whoever is concerned) don't want to answer here, then there is still the Bento Jira where i am sure any suich input is welcome...

Link to comment
Share on other sites


Teager wrote:


Medhue Simoni wrote:

Here's a tangent, but relevant. The facial bones parented to the skull bone, I really don't have any issues with. Wish we didn't have to, but I do understand the issue fully, and why. Again tho, that said, will I be updating my wolf to utilize that extra head bone? No, I will not be, because I do like to make all my avatar's heads move with the mouse, and then try and blend in the body movement, when that works best. The neck on my wolf is not really an issue, at all.


The plural of anecdote is not data. Humans, canines, and many other creatures have short necks without many vertebrae, and therefore look perfectly normal with only a single neck bone. But that doesn't mean that other creatures would not benefit from the option of an extra bone. It's dangerous at this stage of the project to proclaim that anything isn't needed or won't be used, especially when you can see what value it might offer to someone else. Obviously any new bones that come out of the bento project, including the bones already in the bento skeleton, will be properly tested to assure that they don't break existing content or in any way affect those creators who choose not to use them.

Notice that I was advocating for the odd parenting, not against it. I was simply pointing out that depending on what the creators requirements are, the extra bone won't always be utilized. As I state later in the comment, I completely understand why a creator would want another neck bone, and there are many different creatures that could utilize it.

It's also worth noting, that if LL considered bone constrainst in SL, as Gaia has suggested, then we could parent those facial bones wherever we wanted.

Link to comment
Share on other sites

  • Lindens

There have been several messages in this thread related to increasing the number of skin weights allowed per vertex beyond the current limit of 4. We understand that this would be useful, and might consider it for future work. Unfortunately, it would not fit within the Bento project for scope reasons. Currently the skin weights are packed in 128-bit entities that we treat as optimized vectors of 4 floats, so changing the limit would require reworking a bunch of the underlying data structures, and would have unknown performance impact since various vector-optimized operations would no longer work. That's not to say it's completely impossible, just that it would be too big an additional task to undertake for Bento at this stage. We are eager to get Bento out for all our main grid users as soon as we can, which means at this stage we have to focus primarily on getting the final skeleton done, and fixing high priority bugs.

Link to comment
Share on other sites

Sure, rig it to another bone. But what's wrong with the tongue bones since you can unlock them and move as you wish? If you don't need it to be parented to jaw, rig to something from upper face. If it's in blender and you use for it not a center bone, but left or right bone, and if you have mirror modifier, disable there vertex groups mirroring, otherwise it will move only half of the tongue.

Link to comment
Share on other sites

 

Gaia Clary wrote:

 {snip}


 

Only 1 wing root but 2 tails

To allow a second pair of legs without need to create odd animations:
  • mWingsRoot
  • mTail1Left
  • mTail2Left
  • mTail3Left
  • mTail4Left
  • mTail5Left
  • mTail1Right
  • mTail2Right
  • mTail3Right
  • mTail4Right
  • mTail5Right

This change allows to add a second pair of legs which are attached to mPelvis instead of mChest.

This leads to 11 bones compared to the 9 bones we have now.

 

 



 

Also Creatures with tail and 2 pairs of legs could reuse the wings as tail and possibly get a wiggling tail "for free" (because of the automatic breath animation for the chest)

 

Alternative to the 2 tails solution: "constraints"

This idea came to our minds when we where working on the bone constraints for the Avastar IK Rig. The idea is:

Parent the Wings root to the Spine, but also add for example a (configurable) constraint to the local rotation of mChest to mWingRoot. If possible make the constraint target configurable, so that users can "clamp" the wings to mChest, or mTorso or mSpine, or mhead, ...

To explain this a little better: This idea does not change the topology of the Skeleton. It only adds bone movement constraints which make one bone (the driven bone) follow another bone (the controlling bone). Bone constraints are very common in animation programs, so maybe it is also doable in the Second Life Viewer. We use this technique in Avastar to separate the deform bone animations from the Rig (so we could add IK based animations)

 

Generalize constraints

If constraints can be implemented in general, then this concept could be generalized to work for all limbs, then users can for example constrain the root bones of the legs, arms, wings and tail to mSpine and create a 7 legged creature that can be animated without adding odd rules to the animation.

 

Thinking even further, if location constraints where allowed, then this idea allows to constrain any limb to any bone. Wings could then be constrained to eye brows for antennas, tail(s) could be constrained to chest for 4 winged creatures, etc.

 

I would like to voice my very enthusiastic support for either of these ideas, but more specifically the constraints to make use of the bones we already have.

Allow me to domenstrate a usage. For bento I am focusing on developing my own centaur body (human upper half, horse lower half.)

Now you'd think that using the extra limb bones (the wings) would be usable, and i shouldn't be having an issue. However this is not the case, because if i repurposes the wing bones as leg bones, which i could, i would be forced to apply a full-body animation to the centaur avatar because of the wing's parenting to the chest. This would mean that users could not animate the upper half with their own animations and would be stuck using whatever i made for them, meaning it would not be comaptible with a large amount of content they already own. Like for example, anything that would animate the arms like props and weapons, as well as pre-existing human poses to give unique personality to the upper human body.

Something that go quite well with a centaur is a bow and arrow set! and there's plenty out there, and i doubt my customers would want to be forced to use a limited amount of animations I could provide for them. That breaks the overall versatile second life experiance that draws so many users to begin with.

So this brings me to using only bones that parent to the lower body, or the pelvis, such as the leg bones, the tail, and the groin. I split the tail in two so each extra back leg would have 3 tail bones each, and moved the unused toe bones from the front legs to be a fourth bone for each leg, allowing me to animate all nessecary joints of the horse's legs. I also repurosed the groin bone as a very simple tail bone.

Now, this rig will function in second life (thanks to translations! :) ) if it's provided with the proper animations. Which is all fine and dandy, but since the toe bones are connect to the front legs, when I try to move a back leg I run into this issue:



The hoof stays behind.

The same thing happens when I move a front leg, but in reverse:



The back hoof moves with the front leg it's connected to.

Also, since the tail bone is all connected, when i move the leg that has the tail's parent bone, both of them move together.



In order to make animations for this bone setup, with current second life restrictions, I would have to manually, by hand. move every bone back in place when I try to animate every frame of the animation. That is extremely tedius, painful work, especially considering the solution would be simple, as Gaia has suggested in her post. The solution would be to allow to add custom constraints to these new limbs so we could use preexisting bones for new uses, or to change the tail to be two tails with 4 to 5 bones each. Unfortunately, the extra tail only benifits users like myself who would find use of it in a similar fashion, and thus the constraint concept would function much easier and kinder on all creators as a whole as it would allow anyone to repurpose any bone for any usage in their projects.

  • Like 1
Link to comment
Share on other sites

Thank you Vir for your answer and your explanations regarding this request and the reasons for this limitation.

I am really disappointed if this limit can not be raised in parallel to the Bento project.

I think this means I will have to make the choice to not make my heads Fitted as before...

 

Link to comment
Share on other sites

Indeed, the problems with weight limit to 4 bones come when you are dealing with Fitted Meshes as very well explained by Kitsune Shan.

And this exactly my case on my head that I want to be Fitted, so also rigged to the HEAD bone.

As very simple example, let's take the Bento Template of Avastar.
Most of the face vertices are already weighted to 4 bones (as the one selected on the picture).



So now, it becomes very difficult to keep all the smoothness of the rig if you need to add the HEAD bone to the weights...

 

 

Link to comment
Share on other sites

You are about to reply to a thread that has been inactive for 1276 days.

Please take a moment to consider if this thread is worth bumping.

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share

×
×
  • Create New...