Some animation experimenting

[NiranV Dean]

I've been experimenting some with animations to see whether my plans for a full animation editor would even work starting with what i consider is probably the most important feature, a timeline allowing you to go through the animation and jump to any point of the full length of the animation. You can see that its working and even that as you'd expect only the actual preview animation (the dance) is being manipulated here, the tail does not continue to move. For the actual animator tool this timeline feature will be essential to jump to and from keyframes and anywhere inbetween and see the results live in Second Life, obviously now that this is done it will also be kept as a baseline extra feature in the preview window just like the timescale option.

Since i brought up the Poser and committed it to LL for official implementation there was always the question for an animation exporter to save the poses or animations you create and show them to others. Since LL is not willing to work with me on adding a live-sync for posing i'll have to work on an exporter sooner or later. I've been pushing the exporter back for a long time for multiple reasons.

1. An exporter would suddenly require limitations for the Poser that i'm not willing implement, such as not being able to edit your pose in the first place. You'd have to start from a T pose, defeating the reason the Poser was made (which was fixing and tweaking already existing poses, which would automatically include making new ones of course)
2. Imposing limitations on the Poser would mean i'd have to split the Poser into multiple modes, one for "create-a-new-pose" starting you from T pose and another one similar to what we've got with EEP which will allow editing your current pose but not saving it and we've seen what a *****storm this caused. This would unnecessarily up the complexity of the tool even further and the Poser was made as easy-to-understand tool to quickly tweak or create poses.
3. The exporter requires rethinking most of the already existing animator system and basically scrapping it all together which i wasn't willing to do at the time.
4. The exporter was never planned and the "i cannot see you posing" issue was always meant to be fixed via live-sync, not via exporting it as static animation, which would have circumvented all concerns about copybotting.

The exporter however wasn't something i just wanted to glue onto the Poser, if anything i want it to be a full animation editor which would also trivialize the exporting feature since one already exists in the Viewer but functions with already existing animations. With the Animator being able to edit existing animations LIVE this problem wouldn't even exist, it would also make the process of creating new animations inside the Viewer a lot easier since everything is already in place and used for other things too.

My biggest issue with the Animator however has always been the UI. An animation tool is no simple tool, its a very powerful tool with a lot of potential functionality that requires a sophisticated UI to house it all, not to mention you need to be able to see your Avatar as preview for your changes which imposes limitations on how big the UI can be. The next big problem is making a UI that doesn't require rocket science to operate. It should be something anyone can use at any time without having to study and with all the features i plan to add to it, this is going to be a big challenge. Another thing will be the editing itself, i will require to implement the Poser into the Animator because an animation tool wouldn't be one if it wouldn't allow you to create pose keyframes to interpolate between, the Poser is already big in itself and adding a whole new tool on top while keeping the Poser intact will make the window quite big. I have several solutions to this issue but they are invalidated by the fact that i'm planning to bring this to the official Viewer and i'd want to keep the official implementation as close to my Viewer's implementation as possible while making it consistent with the rest of both UI's which sadly rules out a Sidebar-like toolbar approach and forces me back to static windows. I have no idea yet how i'll  work out the UI, probably like always, just add add add until everything is done and then work on compacting the UI and making it easier to work with.

I have no idea yet when i'll have something usable ready since i'm not fully working on it yet, its still experimenting so far but given how long the Poser took it's probably a month or two from the point on i actively start working on it, not counting further iterations.

I'm however looking for feedback on the feature set of said animator and possibly the implementation of the UI, QoL and so on. An in-world animator is a huge thing with the potential to outright obsolete all third party animation software depending on its feature set and implementation. Ideas for features, UI mockups, that kind of stuff would be very welcome.

Edited July 20, 2020 by NiranV Dean

[Lucia Nightfire]

I need to output from a marker based MOCAP unit live in SL upon my avatar with capture and frame editing and .anim save, basically no back and forth between MOCAP to bvh to animator to SL.

[Profaitchikenz Haiku]

As a workaround or an experiment., I'd like to see the joint details (X-Y-Z pos & rot) chatted out frame by frame as a stream of chat so I could plug them into Qavimator. I appreciate this isn't an elegant method or ideal for a long-term solution, but it would be a lot better than my current method involving an in-world bvh generator using a couple of mannequins. (It has a sort of uneasy similarity to the use of dolls in police investigations - "using the dolls show us exactly what happened ")

Edited July 21, 2020 by Profaitchikenz Haiku

[NiranV Dean]

53 minutes ago, Lucia Nightfire said:

I need to output from a marker based MOCAP unit live in SL upon my avatar with capture and frame editing and .anim save, basically no back and forth between MOCAP to bvh to animator to SL.

First we'd need live-mocap for SL. Long ago there was this project that allowed the Xbox Kinect to be used inside SL to live-animate your avatar, this could have been used in combination with a record feature that automatically creates keyframes when bones rotate/move too far from their previous rotation/position to create what you are describing here. The project would need to be extended to work with MOCAP equipment of course.

But this is far out of the scope of the project, though if someone manages to allow live-posing via any MOCAP or lets say VR equipment, it could be easily hooked up to the animator to allow it to record you.

27 minutes ago, Profaitchikenz Haiku said:

As a workaround or an experiment., I'd like to see the joint details (X-Y-Z pos & rot) chatted out frame by frame as a stream of chat so I could plug them into Qavimator. I appreciate this isn't an elegant method or ideal for a long-term solution, but it would be a lot better than my current method involving an in-world bvh generator using a couple of mannequins. (It has a sort of uneasy similarity to the use of dolls in police investigations - "using the dolls show us exactly what happened ")

I'm not sure how this would help the Animator since the Animator would completely replace QAvimator. You wouldn't need to pull any information out anymore since you can simply export (as in save to disk or directly upload) the animation you are creating/editing inside SL. Everything QAvimator can do would be possible inside SL directly including rezzing and attaching objects and that would only be the baseline feature set, QAvimator cannot animate bento or fitted mesh bones, the Poser can and so would the Animator, the Poser can rotate, position and scale all available bones including even Attachment Bones and Collision Volumes.

All other information would most likely be shown similar to the Poser in a list if you toggle a debugging option in the Animator (i don't want it to be visible without specifically toggling debug mode on because its a lot of unnecessary information that would just bloat the window)

Edited July 21, 2020 by NiranV Dean

[Profaitchikenz Haiku]

1 minute ago, NiranV Dean said:

QAvimator cannot animate bento or fitted mesh bones

QavimatorBeta can. But the interface just isn't good enough for all the extras, there's not enough space in the field for bones, it's a bit tedious.

My suggestion was to have an interim solution until the Animator is ready, it's obviously not as good as saving the animation direct from in-world.

[NiranV Dean]

1 minute ago, Profaitchikenz Haiku said:

QavimatorBeta can. But the interface just isn't good enough for all the extras, there's not enough space in the field for bones, it's a bit tedious.

My suggestion was to have an interim solution until the Animator is ready, it's obviously not as good as saving the animation direct from in-world.

Ah.

Well the problem is that i need a working animator in the first place to add a feature that spews out all rot/pos in chat, since you'd need to have an animation loaded that you created unless you want to take an existing one and simply export it which the Viewer can already do without the animator which would make chatspamming yourself obsolete... however the viewer can only export as .anim , not as .bvh and unless the Beta can also import and export .anim files you're sol

Edited July 21, 2020 by NiranV Dean

[Profaitchikenz Haiku]

I was thinking about your poser, it currently works just for the local client which is great for taking a snapshot but having got a successful position I would like to then build it as a pose in Qavimator so that it becomes visible to others.

The next step using that method is to create several poses that form the sequence in a dance or complex movement, and put each pose into Qavimator at different steps, using the natural transition of the bones from one pose to the next to create the final movement. It's obviously not a patch on mocap but it's quite successful if used carefully. Currently I use the inworld tool to create these poses which are exported as bvh files, but qavimator does not allow the import of poses at different stages, so each pose has to be laboriously written out as the joint details and then equally laboriously entered at the appropriate stage in the working file.

[NiranV Dean]

2 minutes ago, Profaitchikenz Haiku said:

I was thinking about your poser, it currently works just for the local client which is great for taking a snapshot but having got a successful position I would like to then build it as a pose in Qavimator so that it becomes visible to others.

The next step using that method is to create several poses that form the sequence in a dance or complex movement, and put each pose into Qavimator at different steps, using the natural transition of the bones from one pose to the next to create the final movement. It's obviously not a patch on mocap but it's quite successful if used carefully. Currently I use the inworld tool to create these poses which are exported as bvh files, but qavimator does not allow the import of poses at different stages, so each pose has to be laboriously written out as the joint details and then equally laboriously entered at the appropriate stage in the working file.

Ah. Well now that makes sense. The poser could have a chatspam option added yes, although you can also take the saved poses and open them with a text editor, all rotations/positions/scales are in there. This would replicate keyframes at least no in-between frames though although that wouldnt be necessary since QAvimator generates transitions between two keyframes automatically and allows creating new keyframes inbetween to manipulate.

[Profaitchikenz Haiku]

1 minute ago, NiranV Dean said:

although you can also take the saved poses and open them with a text editor, all rotations/positions/scales are in there

Doh, I never thought of looking in them, just assumed they'd be binary files  

[EnCore Mayne]

while your Poser application is superb (do you have a Patreon account we can support your development?) i'm worried about copyright infringement. as it is now, as far as i understand, there is no ability to export .anim or .bvh. are you working on getting it to export poses/animations based on other's work?

[NiranV Dean]

15 minutes ago, EnCore Mayne said:

while your Poser application is superb (do you have a Patreon account we can support your development?) i'm worried about copyright infringement. as it is now, as far as i understand, there is no ability to export .anim or .bvh. are you working on getting it to export poses/animations based on other's work?

Copyright infringement is the bane of all good features.

It's the reason the Poser isn't yet in the official viewer and its the reason the Poser has no export. Export means you could copybot animations/poses, which in turn requires limitations, which in turn limit the potential, which in turn make an otherwise great feature less great to the point you start asking yourself why it was made in the first place.

The animator will face the same issue. Half of what i'd like the Animator to do (live editing already existing animations) won't be possible because due to the whole content theft thingy. It's a shame. Ultimately i see both the Poser and the Animator in their official implementation being completely neglected or ignored by users because they simply can't offer what makes them truly good. Which is why i haven't spent any more time investigating an export for the Poser: It's going to require limitations i'm not willing to add to the Poser.

[EnCore Mayne]

setting aside all of the imposed limitations for an official implementation into the SL viewer, have you considered creating a stand alone pose/animation editor (similar to Qavimator) made available for purchase so we could use it to make live poses/animations inworld? as it is for moi, i'm using the Blender/Avastar rig and would die to create on an inworld avatar. of course, this would be starting from a Tpose, so no "tweaking" other's poses.

[Mollymews]

12 hours ago, NiranV Dean said:

Copyright infringement is the bane of all good features.

It's the reason the Poser isn't yet in the official viewer and its the reason the Poser has no export. Export means you could copybot animations/poses, which in turn requires limitations, which in turn limit the potential, which in turn make an otherwise great feature less great to the point you start asking yourself why it was made in the first place.

if the Export was restricted to the creator/uploader of the animation file I wonder if that would satisfy Linden compliance

[Lucia Nightfire]

2 hours ago, Mollymews said:

if the Export was restricted to the creator/uploader of the animation file I wonder if that would satisfy Linden compliance

Probably a moot point with creator/uploader checks since 99.99% of the user base are in possession of animations they didn't upload and the same amount don't even make animations themself to use the feature, so it would become a animator's tool/toy, of which, they more likely will just use whatever existing animator program they normally use instead.

[Mollymews]

edit add this part as I never explained myself fully in my other post

my thought is that people could upload a 1 frame T-pose animation file of which they will be the SL named creator, and then use this with Niran's  tool to make exact fit animations for their avatar inworld, which for Export purposes they will the named creator, and of any other animations that they might derive from their previously created animations from that T-pose animation file

i would use this tool if was available, as is a bit tedious to create exact fit animations for my own avatar when I have to use an external animation editor tool

Edited July 22, 2020 by Mollymews

[Chic Aeon]

25 minutes ago, Mollymews said:

edit add this part as I never explained myself fully in my other post

my thought is that people could upload a 1 frame T-pose animation file of which they will be the SL named creator, and then use this with Niran's  tool to make exact fit animations for their avatar inworld, which for Export purposes they will the named creator, and of any other animations that they might derive from their previously created animations from that T-pose animation file

i would use this tool if was available, as is a bit tedious to create exact fit animations for my own avatar when I have to use an external animation editor tool

Along with that exporting only YOUR work (I am for that of course) it might be possible to make a new animation start (T pose) much like you would make a new shape. That way you could make your animation inside the viewer and export out and it would be all your work.    I am used to working in Blender now so likely I wouldn't use the feature for making poses and animations, but  those not friendly with Blender or other animation software would likely enjoy being able to make their own poses and animations. 

 

[Mollymews]

1 hour ago, Chic Aeon said:

Along with that exporting only YOUR work (I am for that of course) it might be possible to make a new animation start (T pose) much like you would make a new shape. That way you could make your animation inside the viewer and export out and it would be all your work

yes. I think this would be the extra things that Linden would add to Niran's code should Niran submit it.  As you say a new T-Pose asset can be created. And when we Export then Save to File, or Save To Inventory (pay the upload fee) same as Snapshot photos are down now

i think the main thing for Niran to know is that before proceeding further with a exporter then would the uploaded T-Pose animation file naming the creator, suffice for Linden compliance requirements.  Only Linden can say for sure, but I think it most likely would as this is how compliant TPV prim-to-mesh exporters work now

[NiranV Dean]

14 hours ago, EnCore Mayne said:

setting aside all of the imposed limitations for an official implementation into the SL viewer, have you considered creating a stand alone pose/animation editor (similar to Qavimator) made available for purchase so we could use it to make live poses/animations inworld? as it is for moi, i'm using the Blender/Avastar rig and would die to create on an inworld avatar. of course, this would be starting from a Tpose, so no "tweaking" other's poses.

No. A) I want this tool to be free, not another avastar everyone is forced to use due to a ridiculous snowballing system. B) Making it a separate program is basically exactly the opposite of what a in-world Animator is about. The inworld animator's big upside is being able to directly manipulate your avatar inside SL and match your animation to your avatar and the surrounding you are creating it for. That was the premise of the Poser too, otherwise you could just use one of the many other programs such as Poser, Qavimator, Blender and so on for animating/posing. C) I'm literally incapable of making a third party program for this, that's far beyond my paygrade, so to say.

14 hours ago, Mollymews said:

if the Export was restricted to the creator/uploader of the animation file I wonder if that would satisfy Linden compliance

 

12 hours ago, Lucia Nightfire said:

Probably a moot point with creator/uploader checks since 99.99% of the user base are in possession of animations they didn't upload and the same amount don't even make animations themself to use the feature, so it would become a animator's tool/toy, of which, they more likely will just use whatever existing animator program they normally use instead.

Exactly. This limitation would rule out basically 99.99% of all assets making direct animation editing an useless feature.

11 hours ago, Mollymews said:

edit add this part as I never explained myself fully in my other post

my thought is that people could upload a 1 frame T-pose animation file of which they will be the SL named creator, and then use this with Niran's  tool to make exact fit animations for their avatar inworld, which for Export purposes they will the named creator, and of any other animations that they might derive from their previously created animations from that T-pose animation file

i would use this tool if was available, as is a bit tedious to create exact fit animations for my own avatar when I have to use an external animation editor tool

 

11 hours ago, Chic Aeon said:

Along with that exporting only YOUR work (I am for that of course) it might be possible to make a new animation start (T pose) much like you would make a new shape. That way you could make your animation inside the viewer and export out and it would be all your work.    I am used to working in Blender now so likely I wouldn't use the feature for making poses and animations, but  those not friendly with Blender or other animation software would likely enjoy being able to make their own poses and animations. 

 

 

9 hours ago, Mollymews said:

yes. I think this would be the extra things that Linden would add to Niran's code should Niran submit it.  As you say a new T-Pose asset can be created. And when we Export then Save to File, or Save To Inventory (pay the upload fee) same as Snapshot photos are down now

i think the main thing for Niran to know is that before proceeding further with a exporter then would the uploaded T-Pose animation file naming the creator, suffice for Linden compliance requirements.  Only Linden can say for sure, but I think it most likely would as this is how compliant TPV prim-to-mesh exporters work now

You're thinking way too complex about this! It's much easier. Also it wouldn't work like this. You cannot create an empty 1-frame animation inside the Viewer on the server and then edit it. Animations cannot be "replaced" or "changed" like notecards or objects, you'd have to re-upload them knowing LL you'd pay 10L$ for the 1-frame animation creation because its technically an upload, and another 10L$ to upload the completed animation. Starting from T-Pose is pretty easy, its a simple iterate through all bones and reset all rotations, the Poser can already do this. The animator creates a local "temporary" animation inside the Viewer that only the Viewer knows about in order to display it, doing so comes with the upside that everything a normal animation does can easily be done with this preview animation as well, ease in/out, transitions, loop etc all of which automatically works and can be previewed. The export would immediately and automatically work too.

Edited July 22, 2020 by NiranV Dean

[Candide LeMay]

Once upon a time we were hoping to get this http://wiki.secondlife.com/wiki/Puppeteering

I'm not sure if the code still lurks around somewhere.

[Lucia Nightfire]

SL needs procedural animation capability, but it scares @Vir Linden to death, heh.

[Wulfie Reanimator]

1 hour ago, Lucia Nightfire said:

SL needs procedural animation capability, but it scares @Vir Linden to death, heh.

It already has it. Motion without explicit animations, and IK. 

Edited July 22, 2020 by Wulfie Reanimator

[Lucia Nightfire]

2 hours ago, Wulfie Reanimator said:

It already has it. Motion without explicit animations, and IK. 

The only procedural animations the viewer has are look at alt cam point, turn head to alt cam point, turn head to speaker, face hand/arm to selection, foot/leg IK's based on terrain slope you're standing on, etc.

What I'm talking about is the ability to "build" temp/assetless animations by script, similar to how you can build a keyframe motion list for objects.

I'm also including the ability to insert IK's into the keyframes for any bone with weighting and distance referencing to other bones on the same user, bones on a partner or other user, or rezzed object.

Edited July 22, 2020 by Lucia Nightfire

[animats]

Here's the state of the art: the UE5 engine demo. There's narration explaining what the animation system is doing.

How the pros do it.

Foot placement and hand placement are all implicit here. They're all derived automatically from the big body motions. This is not inherently impossible for SL.

(An SL Viewer 3, based on the UE5 engine, would be a big job, but not an impossible one.)

Edited July 22, 2020 by animats

[Lucia Nightfire]

1 hour ago, animats said:

Foot placement and hand placement are all implicit here.

The problem is, in SL, some applications use "unorthodox" bone usage. This is even moreso within Animesh applications.

Hand and feet IKs alone are not going to cut it.

[NiranV Dean]

I've been thinking of making the foot IK play over animations, just like i did with the head lookat, which improved the look a lot, i haven't managed to get that to work yet though.