How can we have a better Blender Character Animator userbase?

@D_structorr
While Modularizing some aspects of tutorials would help for some things “like the diffrent constraints” the problem is that new users often dont know what they need, they only know what they want to achive.

A new user would more likely search “Leg Setup Tutorial” rather than “IK Constraint Demonstration” especialy if they dont know about IK constraints and thats why most tutorials are goal oriented rather than feature oriented.

As for tutorials often referring to things not covered in the tutorial, I have mixed views on this.

I think that its detramental to ristrict ALL tutorial’s so that they are accessable to entry level users and cover all the basics, this is becuse it prevents you from covering complex things. It also increases the length of the tutorial meaning you have to sift through a 20min video to find the 3mins you were after. Its also frustrating to the person creating the tutorial as they have to spend thier time talking about things that have already been done to death before they can talk about whats important.

When I created a tutorial on IPO Drivers I got alot of complaints becuse I did not cover the basics, but if I had the video would have been over 2 hours. You need to be able to assume a certain level of knowlage when creating a tutorial otherwise you are limiting the scope of tutorials to cover the basics only.

That said, I always try and let the viewer know were to find that relivent infomation even if I only say what its called.

Animation as a skillset is teachable through video tutorials, I learned how to animate at AM watching them. The problem that the blender community has is that often the line between rigging and animation is blurred and this is especialy true for tutorials.

For an excellent introduction to animation, check out Character Animation Fundamentals on the Blender Cookie site.

Speaking only for myself, I started modeling first because I want to animate my own characters. I’m willing to wait until I understand modeling, texturing, and rigging before I dive into animating. But that’s just me.

personally, i’ve been feeling pretty exhausted by so many beginner-to-intermediate animation and rigging tutorials “out there”. these days, i prefer books just because it’s so much easier to jump ahead and find information that’s relevant to my level, whatever level that is. i’d like to see some much more advanced character animation tutorials. i enjoy animating immensely, and i practice and study it, but i’ve never had any formal training in it - i’d like to get more of that from people with experience, but i find that too often that experience isn’t being fully appreciated because the tutorials are too introductory.

Books are good becuse they can start at the begining with entry level basics and end with more advanced infomation. They can contain enough infomation to keep you reading for years. Video tutorials are too short to contain all that infomation alone, but thier is no reason why a database of video tutorials cant contain the same infomation. This is why its so important to alow video tutorials to not start at the beginning, it alows for more advanced concepts to be covered.

Let me append something:

If blender had a collection of said feature tutorials, an author making a video tutorial could conceivably refer to it for further explanation.

That would lead to a form of standardization and more people would be reading from the same page. I’m not up for restricting, just a “fall-back” base for when users get stumped on something about a tutorial (an advanced user’s tutorial). Hey, if on the comments someone doesn’t understand something, another poster could post to the relevant feature tutorials which would be compiled (supposedly by us generous blender users) since it’s easy to make entries of these reference materials one by one and refine them while keeping them modular and to the point.

The feature tuts would be safety nets to keep users from staying confused when they want to move forward. References that’s what they are intended to be and stated above is their purpose.

I think a big problem that leads to many people not going into animation is a lack of good learning resources.

I brought a Blender(2.5) book on animation and I have read it and let me tell you its crap because 50% of the book has nothing to do with animation. Now that I have read this book it leaves me with only one other Blender related animation book to read. Am I going to buy it? No because I fear more of the same.

My biggest problem with said book is a lack of focus on the topic of animation itself. Modelling, texturing, rigging, weight painting, rendering using the VSE, useless interviews all that padding that has nothing to do with animation.

But this problem is not isolated to this book alone, take a look at old summer of documentation pdf on character animation where do you start modelling, texturing… you will cover 60%-70% of that pdf before you get to the actual animation part. I have looked at Animation Mentor curriculum that is the laser like focus on animation I went from a book myself.

I want to see a book where you are animating on chapter one. Don’t bother with modelling, rigging, texturing just provide the rigs and an explanation on using them. I would rather spend my time reading about arcs, overlapping action, timing posing than read pages on UV mapping or setting up particle hair.

Animator’s Survival Kit is hands down the best animation book. Disney’s Illusion of Life and Cartoon Animation by Preston Blair are two other good ones I have. If you were just going to get one book it’d definitely be Animator’s Survival Kit though.

For Blender specific I still think the “Character Animation in Blender” DVD is the best even though its for 2.4x. It’s up on BlipTV here. Lee has some good animation vids here too. I haven’t checked out the CG Cookie DVD yet.

Great to see few positive responses here. Initially, I thought my words wouldn’t reach so many of you. Thanks for this conversation.

While replying on some of the great comments (and few very common notion in our community) which I see here, I may take few names of our talented artists/devs/authors with due respect. No offense intended.

[Click on the images to get the full view]

@rontarrant: Advanced rigging is in my course plan and I will eventually try to master it and provide some useful rigs to the community. Thanks for your wise words.

Are you referring to Red-Nelb_v08? If that’s the one, I will have to try that out.

As Chris L said, auto-rigging system never gives you a complete and final solution. It would be rather good, if we can use it as a base and able to add required features on top of it. We are also looking forward to Bassam’s rig-retargetting/ autorigging system: Rigamarule and lets see what it turns out to be.

About the ‘The look of the character is unimportant if it’s a motion’, I agree if it’s pure body mechanics without any facial gesture. But if it’s going have facial expression, Character Design and Appeal is quite an important factor. Have a look at this.

I will give another example of appeal in later part of this post.

I don’t have 100s but just about 10 Maya rigs, out of which 99% are excellent rigs. I do have plenty of Blender rigs as well but they are not as powerful as they have on Animation Buffet or Creative Crash. I will give you 2 simple examples;

  1. I want to do a simple bouncing-ball exercise and want maximum control for my desired extreme poses. Here is the layout plan for my squash n stretch effect.
    http://dl.dropbox.com/u/13535210/11%20Second%20Club/TLC/ball_bounce_sns_small.png
    Do I have a Blender rig that can give me such squash n stretch feel? I am sorry, I couldn’t find any. But for Maya, there are numerous such rigs already available (even more advanced and complex than the one I have shown below)
    http://dl.dropbox.com/u/13535210/11%20Second%20Club/TLC/bouncing_ball_an_small.png

I was trying Beorn Leonard’s rig, then just for simple ROT mechanism and I have had to counter-animate!

http://dl.dropbox.com/u/13535210/11%20Second%20Club/TLC/cons_bouncing_ball_small.png

Here is the GIF demo.
http://dl.dropbox.com/u/13535210/11%20Second%20Club/TLC/bouncing_ball.gif

  1. Another example of simple generic rig, yet having those extra controls for pushing your poses further.

http://dl.dropbox.com/u/13535210/11%20Second%20Club/TLC/flour_sack_an_small.png

http://dl.dropbox.com/u/13535210/11%20Second%20Club/TLC/flour_sack_posed_small.png

Some great points there, also liked the idea of your monthly animation challenge. If you need some assistance, do let us know; how can we help and make it happen.

@xrg: Oh, that’s a cute little animation, but why I see too much overlaps on the left hand? :eyebrowlift:

This thread is mainly focused for the professionals (or hobbyists, who wants to be professionals in due time) and serious about their character animation as a specialization than just being 3D generalist for the rest of their carrier.

Is it possible for the riggers, who worked on such Open Movies to update their respective rigs for major Blender releases at every 6 or 12 months, and/or add requested/required features to make it more flexible for any genre of CA? Few rigger do update their rigs for every annual release of Maya or Softimage XSI.

Oh, this is of great concern and I guess, this is being done by so many authors since Tony Mullen released his first book. Most people who are introduced to 3D CG world through these books, implants an idea of creating a movie/game of their own from scratch. This approach is really not good if you want to have strong base/foundation in any of the disciplines that needs mastery not jacks.

However, I feel most of the training materials from Blender Foundation is of top quality and mainly focus on specializing one particular area than just putting irrelevant and redundant topics.

Lastly, I want to show something more about the appealing thing (which I apparently came to know some time back).

Here is a little overview:

If you’ve liked that, you would definitely love to get more insight of these:

Part I:

Part II:

But then, appealing alone is not going to make your animation super awesome. There is a general opinion that human characters of DreamWorks Animation mostly lack that appealing part which Pixar have taken advantage of it since their early days. Anyways, DWA’s animation is rock solid and so much fluidity, irrespective of these failures. Similarly, Sintel is a very appealing character but animation at some part in the film (@04:43 - 05:49) was a let down.

If a character come with those required appeal, it will impel animators and whisper in your ears “Come on, animate me!”
:wink:

although I have to admit I’ve skipped and scanned through this thread and I don’t have much animation experience.
One thing I have learned though is that animation (and also modelling for that matter) shouldn’t be regarded too much as a package dependent thing. I think a lot of beginning users start out with well beginning tutorials to learn the tools of the program. But once you get passed a certain skill level you don’t have to rely on package specific tutorials. You can usually apply fundamentel techniques of more advanced books/tutorials/dvds with other packages in mind in blender.
Especially animation is something that has a lot fundamentel things, that you don’t (have to) learn from package specific tutorials. So maybe in that regard, educating beginner blender users/hobbyist that even though a tutorial says maya, one should not dismiss it immediately, might actually help as well?
anyway just putting my 2cents in.

I think this is the major reason most people start with modeling. I have projects that I want to work on and those projects look a certain way in my head. I cannot start animating until I have the object that I want to animate.

I’m also curious. I like to know how an entire process works from start-to-finish. I would not have much fun starting with animation because I would immediately start asking, “where did this model come from?”, “what if I need to change it?”, “Why does moving the model in certain ways work better than others?”. Building a model is the first step in the 3D art process.

While you are obvously right, would you not reconsider if a character rig existed that was customizable enough so that you could create your character without modeling or rigging it yourself?

This is an interesting thread, to be sure. And, I am happy to see some of the people who have jumped in here, because ultimately this problem can only be solved by animation talent taking their spare time and doing something with it.

We had a really cool rig started by Ivo, but he left for greener pastures when he came to the conclusion that he couldn’t make money with Blender. He took the Norman rig from Animation Mentor and converted it, which was looking really good before he left (Norman Rig Blender)

Let me add some other issues:

  • Lack of a World Class Render Engine. And, in general, I don’t have a lot of hope for Cycles to address this issue, because I have a hard time believing that an unbiased renderer is the solution for animation. Unbiased Renderers really aren’t good at generating thousands of images quickly… it is more suited to working on one image, passing over it 1000 times to get it right. BI is too quirky and non-standard, and Cycles is really best at generating jaw-dropping stills. So, why should I spend time in Blender to create cool animations when I am going to end up fighting to get it into another package where I can render it?

  • And, the well documented feud between BF and AD keeps the pipeline from being everything it should be… so this is an issue.

  • The other thing is that Blender suffers from an ambitious release schedule which quickly makes any tutorial out-dated…

  • Finally there is the community itself which seems to want everything for free. There are some exceptions, but mostly people want to get something for nothing – the GPL becomes a way of life. The problem is that it is hard to get professionals to be willing to donate the time necessary to build quality training when they know at the end of the day, they will recoup nothing for their troubles.

(PS: MCHammond – sorry to hear that you got bad feedback on your IPO tuts because of their level. I think you are absolutely right… we need intermediate and advanced tutorials. And I learned a lot from yours!)

@ @MarkJoel60
Interesting you should bring up the render engine, the last few days I have been looking into it and it is a problem. BI is good and can do it, but when you start looking at more complex scene’s BI gets too complex too quickly, this is especially true for hair/strand render.

Cycles on the other hand is better but its still not ready for production, I think that an unbiased approuch while it might not work at the moment will work in 3 to 5 years time as GPU power increases and thats the time scale of a program like this. Cycles will be around for atleast that amount of time and it is right that the BF should get into the game early.

What we need is a good renderman exporter, that would solve the problem right thier.
Blender + a choice of Industy proven render engines = no brainer

Well, except I don’t know many Blender users who would plop down $3500 for a Render Engine for an animation program that was free… I think that could be a solution, but there needs to be a more economical path as well…

There are Open source renderman compatable engines like Pixie and Aqsis that could do with the love that a blender exporter would bring them.

3Delight is free “for a single machine” and is a pro level renderer and is very cheap to upgrade:
http://www.3delight.com/en/index.php?page=3DSP_pricing

But there are also loads of other Renderman compatible engines to choose from, thats the point. At the end of the day Blender needs to be compatable with industy standards if we wants it to attract industry veterans.

Perhaps, but only if I could tweak the look of the character(s) to suit the scenes in my head. Sort of like iClone, but with better animation controls.

Trouble is, the ideas I get don’t lend themselves easily to stock humanoid characters. So, the amount of tweaking I’d have to do would result in learning most of the modeling, texturing, rigging etc. anyway. Either that or I’d end up waiting (or begging) for someone else to build characters closer to what I was looking for.

I decided it would be far less frustrating to just get on with it and learn enough of each aspect so I could do the stories I have in mind with the characters who are already running around inside my head.

Something like the morpheus rig?
http://www.joshburton.com/media/images/work/MorphyTeaseWip_web.png

Cycles is not “unbiased” exactly. True unbiased engines like Luxrender are nowhere NEAR as flexible and controllable as Cycles.

Cycles is a brute force unidirectional path tracer. It’s design philosophy takes a lot of cues from Arnold and Sony’s Open Shading Language, making it VERY similar to Arnold in practice.

Arnold is used exclusively now for Sony Imageworks. Cloudy with a Chance of Meatballs was rendered entirely with Arnold.

When Cycles matures, it has the potential to be one of the BEST all around modern renderers on the market and has the potential to turn a lot of heads.

So, in short, I disagree. Cycles is on track to be an incredible renderer for animation. It just needs more time to get there.

That looks like a very flexible system, sort of iClone on steroids and if I were planning a bunch of biped humanoid characters, I’d likely want something along those lines for speedily creating characters. However…

Unless you’re satisfied with the way the character is dressed…
Or what if the character you want to work with has four arms?
Or two heads?
Or twelve fingers? (don’t laugh; I have a 12-fingered hand already modeled for an alien character I have in mind; you should see their counting system; it’s base-24: vegiquatral?)
Or is a cyclops?
Or you don’t want noses?

Well, you see my point, right?

Well, Messiah uses an early cut of the Arnold Renderer, and I am very familiar with that… I guess opinions vary, but I think to call Cycles “Very Similar to Arnold” is a huge stretch (and at this point, a huge leap of faith.) If you feel, like MCHammond, that it will get there in three to five years - OK, I’ll concede that, because as a developer, I know that a lot can be accomplished in 3-5 years. I think Brecht is a talented coder, and given four years or so, he can accomplish a lot. However, I have seen too many things with “potential” fail to deliver, so forgive me if I remain skeptical.

One of the things that concerns me about Cycles is it really isn’t supposed to be much more than a hobbyist’s tool. In his interview in BlenderDiplom, Brecht said this:

“The target audience is individual artists and small studios. It’s not meant to be a render engine for Hollywood. It’s meant to be something that is easy to use…”

I think MCHammond’s idea of a bridge to Renderman makes a lot of sense… And I wonder what could have been accomplished if some of the effort being thrown into Cycles had been thrown into one of the Open Source projects instead. Imagine, a free Renderman that got you 80% of the features… and for the studios who really needed it, they could buy the regular studio and get that (or use the new Cloud Render from Pixar.) That would have been a great approach to the problem, IMHO.