Posing Library and animations workflow

I have watch quite a few tutorials and have seen many processes for doing Blender animation but have a few loose ends I am wondering if someone could help me understand, basically the Pose library in an animation workflow

I have a fully rigged character and ready to animate and understand the base concepts of establishing poses and betweens. I am a little confused about walk cycles and the pose library. I know what they are for but is it advisable to say build all the poses in a pose library and create walk cycles in advance and then animate a scene for the film that you have, using the pose library and walk cycles already stored? or should these be built as the scene progresses?

It would be helpful to know what you are exactly doing. This little bit right here:

and then animate a scene for the film that you have

Makes me think you are inserting cgi character into film footage.

Anyway, here are my thoughts on pose libraries and animation in general. Walkcycles are good for learning, but not much good for anything else. It’s a way of learning the basic 4 poses of a walk, but can’t really be used for animation of a main character that is walking around a room. The main problem with a walkcycle that is created with repeating poses (be if poses from a pose library, or just copying and pasting the poses) is that it repeats, exactly. The end result is a robotic looking animation. If this is for the main character, it just won’t work. You need variation to a walk, because that’s how people walk. People don’t walk like repetitive machines. Then again, if this is for hords of backgound people, use pose librabies. 5 different walkcycles spread across 40 people, and you would probably not notice the constant repetitiveness…

The only thing I really use pose libraries for is mouth animation, and even then, I adjust poses as needed to match the voice.

Just my 2 cents…


Hi Randy,

Thanks for writing and I think you answered my questions. Sorry if it seemed vague. I am a self taught Blender artists as most are I guess. I have a degree in Art and using my skills in 3D. I have read numerous books and have a character I have completed and fully rigged. Did some small animations but now going through the full animation process.

In reading books and tutorials they always talk about walk cycles and the pose library but it is out of context and does not fully explain how they fit them in their workflow. I think you helped me put it into perspective. I agree that walk cycles have their place but to me in a animated film it would just be too mechanical.

I know there is no set way one follows to animate, sort of. There are basic rules we all follow but in terms of workflow I guess it is different for everyone. I think I am settling on possibly creating key poses for each of my animated films in the pose library for convenience and then establish timing through animatics and placing my character at various points in the timeline using poses I have in the pose library as a rough guide and then fine tuning using the pose to pose method. Not sure if this is the best way but I guess we all have to start somewhere in establishing a workflow.

I’m guessing that because you are (or were) an art student, you are not opposed to spending a few dollars on training material. I can suggest several products that will help you put together a good animation workflow using Blender. I feel they are worth every penny and will speed up your learning. Both are available at cgcookie.com

The first is the Character Animation Toolkit, by Beorn Leonard and Nathan Vegdahl. This tutorial series will go through every tool that you would want to use for animation in Blender. Nathan was the main rigger on many Blender open movies (and is also the original author of Rigify (pre 2.79)) and Beorn was an animator on many of those projects. Beorn has also worked on many feature films. The series will cover animation workflows from two experienced animators from start to finish.

The second series is Animation Bootcamp by Wayne Dixon. This series also covers using the animation tools and good animation workflows. Many of the same ideas are present from both series. I would recommend this series for anyone wanting to learn from the ground up.

Between those two series, you will have access to three talented animators and should be able to find a good system for your own work.

For your original question about using the pose library, I find it useful to copy hand or facial poses to speed up pose to pose animation. Because posing a rig can be time consuming, having a few stored hand poses or facial expressions can really speed up the workflow.

Some people use the pose library for lip sync, but that method does not give the best results. I find the relative shapes method (taught in both series mentioned above) to be far more convincing and easy (quick) to animate when the principles are known.

I hope that helps. Good luck!

Thanks DanPro. I will check on those course. Yes, most of the training has been online through Youtube and I did have 2 books on Blender. Some dont go that route but I can be sort of methodical sometimes. I will check out those courses.


So I got a better idea of what you are doing by the posts so far.

I really think what you are looking for you’ll won’t find on a forum.

Not sure if this is the best way but I guess we all have to start somewhere in establishing a workflow.

I’m not sure what the best way is, other than jumping in head first and go at it. There is no best way to learn, cause everyone is different, and learn differently.

It’s like painting, or anything else really, no two artists are the same. So there’s no best way for every artist to learn, each has to learn in their own way.

Forums are useful for solving problems, learning new methods, and gaining knowledge, but they won’t help you develop you own workflow…

@ DanPro - speaking of learning, what’s the lipsynic method described in the videos you mention. I do use pose libraries for this purpose, cause I don’t know of any better way to do it.

But I’m always interested in learning.


It’s called the Relative Shapes Method.

Here is a quick summary:

First plan and plot what the character is saying. Import the audio into blender and add markers on the timeline with the sounds being made. It’s important to plot the actual sounds, not the correct spelling.

Next, do a pass keying only the open/close of the jaw and narrow/wide controls of the mouth based on those markers. It’s important to have a starting expression for the character at this point. Also, if the character talks out of the side of his/her mouth or has an accent, keep this in mind. All mouth shapes for each new sound are relative to the shape before it. More or less open and more or less wide.

For example, if you say “Ohh ,Pee” and “Cee, Dee”, the “ee” in “Pee” is narrower after “Ohh” than the “Dee” after “Cee.” There are all “Eee” sounds, but the width of the mouth depends more on the sound that precedes it. Hence the name, Relative Shapes.

If you are pasting “E” phonemes from the pose library, all the "E"s will be the same width and will not account for the difference. People are very efficient when they talk. Small differences in narrow/wide and open/closed can make a huge impact on how natural a lip sync animation looks.

I like to think of this as the “muppet pass”. Muppet mouths can only be open or closed, yet the puppeteers can get a good results just by varying this one parameter. For each sound the amount of change between open/closed, narrow/wide is what is most important.

At the end of this pass, 90% of the lip sync should be completed. I was always amazed at how much could be accomplished just by varying the mouth opening/closing and the lip corners changing from wide to narrow.

Next, do a pass on the tongue. Remember that the tongue moves very fast and often flicks from one state to the next. Set up the tongue poses for T’s and S’s. Be careful not to overdo the S’s. Showing too much tongue for too long, can make your character look like it has a lisp.

Finally, do one or more polishing passes based on the quality you are going for and the time you have to invest in the animation. This can vary greatly. If you are pressed for time, the lip sync should look decent at this point and you can stop. If you are going for movie quality animation, more polishing passes can be done to fine tune details.

That’s it. It’s a very effective and surprisingly fast way to create decent lip sync.

Good luck!

You are right revolt_randy and yes, it seems we all have a slightly different approach to it all. So like you said jump in and do it. I do have the basics down and have worked hard on modeling and rigging everything in my movie so I should just do it. I like DanPro’s approach to just lip syncing with the pose library. I was going to build full poses but Like DanPro was saying it can be rather restrictive and walk cycles can be very robotical only to save time. I guess Blender offers so much you basically have to decide what works for you and go for it.

I got the Animation Toolkit DanPro suggested and it offers a great workflow concept that I will probably alter a little to fit my needs. I have been building this on a Mac laptop due to no money basically but have just built a kicking Hackintosh with a multi boot Wondows 10 setup on a new computer with a new GPU I built myself and transferring my files to this to begin animation. When I am done I wlll post the link on this site.

This will be my first full 3D animation. I work in the media field in various positions in my life as director camera, editor etc. But have not tackled animation before so might as well get my feet wet. I so much appreciate everyone’s help through the process and hope I can pay it back one day as well.

Ubisoft showcased their pose library they used for Mario & Rabbids: https://youtu.be/qxLR8qbD5JE?t=2103