The perfect simple character modeling & animation workflow

Hi everyone. I wasn’t sure where to post this, I hope it fits in this category. I fear this post will turn out a bit long, opaque and convoluted but I will do my best to make it as clear as possible.

To put it in a nutshell, I have been trying to find the holy grail of a quick and simple workflow to create and animate characters. But while I made a lot of progress, I feel like I get tripped up by constant little workflow inefficiencies.

I would love to discuss how others would approach or do approach this and, perhaps, where I just do it wrong.

A little context: I’ve been dabbling in CGI for over 20 years and have been a Blender user for about 10. I’m interested in many many things, including the technical side of CGI (and many other things) but at heart, I’m an artist and just want to create without any hurdles in my way. My biggest love is film and I love creating stories, from personal to escapist action/sci-fi stuff.
TL;DR: I’m still a kid at heart and want to play with my digital action figures. (make short films)

To this end, what I really want is to easily and quickly create scenes, put some characters in them, animate them and churn out a cool little short film in a month or a feature film in a year or so. I’ve gotten quite fast at creating scenes with ready-made assets, but characters (to the surprise of noone) are still really hard.

This is the sort of stuff (and quality) that I’m making currently:

Not super realistic, older game-engine kind of aesthetics are really enough for telling stories, the way I want to tell them.

Okay, now to the hurdles: Lately, I have gotten a good workflow going. I create characters in MakeHuman, import them with the (older) MakeHuman plugin and the MHX2 rig, which gives me phonemes (making lip sync pretty easy). I animate by hand or use tools called Glycon VR or APS Luxor to get motion tracking from my VR headset. Then I use Auto Rig Pro Remap to apply the motion capture to the MHX2 rig. Voila! Pretty fast and efficient.
BUT: In actual production I find switching between all these tools a nightmare:

  • If I need to make a change to a character I often need to make it in MakeHuman, re-export it, then re-apply changes I made in Blender.

  • For every remap I need to fix the rest pose. ARP makes that really quick, but it becomes a pain to do it every time. At this point, I just want to drag&drop Mixamo-style

  • To make new clothes, I need to use yet another tool, which also has its own very precise workflow I need to keep in mind (previously I even needed to use Blender 2.79 for that, but now it’s 2.8/9 compatible)

  • The MHX2 weight painting isn’t perfect and sometimes I need to fix things.

  • Etc. etc.

Of course, I have looked at Character Creator 3. I do not mind the cost, if it delivered all that I need. But it seems to me, rigging and animating a CC3 character still needs a ton of manual work to make it usable. Alternatively I could use iClone to prep the animation and export to Blender for every scene, yet another software. Getting other clothes and models is expensive, making them yourself is (once again) cumbersome. To me, a lot of this just seems completely unnecessary. I feel like I got some good ideas and solutions here to get really fast at creating, rigging and animating characters at a whim. But I die the death of a thousand papercuts.

If anyone is still here after my rambling:
What kind of workflow do you use?
What kind of suggestion do you have to my workflow (as far as I was able to explain it here)?
What other solutions could you think of?
Do you use Unreal for this kind of quick production? How is it?

I would love to read some of your thoughts on the topic!

1 Like

Hi Unluccy

That’s a really solid and nicely done animation. Great cinematography too.

I think the reason there are not any comments yet for this post is that this is a large and complex subject and ask with no simple go to answers. There is no magic make character button in any software and I doubt there ever really will be. This will always be the cutting edge of the industry and always getting re worked, changed and improved upon.

Character creation and design rigging and animation is commonly the most complex CGI part of a production. This is why there is the need for technical directors on studio productions.

Its sounds like you have system going that is working for you at least. My advice would be to focus in on that and try to find as many efficiency savings as you can in the pipeline and workflow you have already settled on.

Both Blender and Maya have very robust and efficient character rig referencing systems for production to allow for the making of changes to rigs and character models along the way.

I would try to integrate your current character pipeline to work as efficiently as possible with Blenders library referencing system. Try as much as possible to anticipate any changes that might be needed and ensure that flexibility is factored into your animation pipeline.

In this way you could be able to proceed with a project into block out and animation and if something is not quite working then you should be able to update your rig in the library file and make incremental changes to design and performance along the way. This is the way it works in professional production. Normally rigs are constantly re worked tweaked and fixed all the way through. There is never time to extensively test animation rigs fully in today’s production schedules and I don’t think there has been since perhaps the first Jurassic Park. The ability to make changes and fixes along the way most often is factored into the pipeline.

I hope this helps. Character work is a big challenge. Well done for diving in so tenaciously and ambitiously. That’s the most important thing.

All the best

1 Like

Hi Toka, thank you so much!

About an hour after posting my little rant, I already knew that it was way to vague for anyone to reply with anything meaningful. Although I would have to say, you have kind of disproven that theory. :wink: I actually find your input very valuable. Focusing more on each problem at hand instead of throwing up my hands in despair is certainly a better strategy. :smiley: I also think the asset manager will be great and help smooth out my workflow.

Ultimately, the part of me that is a storyteller (writer/director) is constantly butting heads with the CG artist (and editor, sound editor, etc. etc.) in me. When trying to tell a story I need to work quickly and creatively. And perhaps a simpler tool would be the answer there or just the acceptance that I need to make sacrifices in quality for a simpler workflow, when I’m working in Blender.

Anyway, I hope this will be helpful to someone out there in the future and not just for me. Thanks again!

Hey thank you as well. I hope this helped.

This is a problem common to any of us who have needed to make a complex creative project from start to finish. The trick to the one man band approach is to compartmentalise those different sides of your personality. So you balance out the free flowing creative phases with the heavily technical phases. By keeping them separate you can stay clear headed efficient and most of all sane.

In the initial creative burst phase I would say stick to writing and storyboard sketches. Perhaps some animatics too. But I would always try to stay clear of a computer or other technical distractions in this initial phase if possible and just work on paper. The most direct link from mind and imagination on to a tangible readable medium. I would commonly like to be outside somewhere. Most often in a cafe with a coffee with other stuff happening around me. Not so easy in these times, but hopefully again soon and there is always the park bench.

Then when you start blocking out pacing and designing your shots in 3D forget completely about the complications of character rigs and detailed lit environments. Block out and time the shots using simple geometry and bold simple abstract shapes. Focus on composition pacing and storytelling.
Use simple FK linked block figures and environment props that can be made in minutes or seconds. This way you will not be distracted at all by any other issues and can create in a free flowing relaxed state of mind.
This can also be a time to start on a bit of rough edit sound design and music cues too using a temp track. Getting a real sense of the whole thing together.

Next will come the character creation and rigging of course, environment lighting, and props. But if you wanted to emulate the multi layered schedule of a studio production you could overlap this with the 3D creative block out phase. So perhaps have creative days and then days just focusing only on the detailed technical stuff. Which is it’s own type of creativity as well of course. By blocking out your shots early on using simple shapes you will not be wasting any work when you come to this phase and will know exactly what you need and no more. It’s a very efficient way to work.

Then of course you can bring it all together. You will probably be more excited thinking about your next project at this point if you are like the majority of creative types so will just need to stay disciplined and knuckle down and get through finishing this one. Perhaps take a cafe break when the animation is in full swing to clear the head and start writing and sketching the next project idea.

This compartmentalised process works I can assure you. And is the best I ever found.

Again I hope this helps. Your work is good already so it’s just needing to smooth out the workflow a bit more. Hopefully it could help others too because these are universal common problems when working alone. Unless you have had any direct experiance of a studio production large or small it’s often hard to conceptualise how these elements and phases all work together to make a whole.

All the best

Thank you! That is some really solid advice. Having worked in film production professionally for many years, I usually structure a project this way, of course. But for some of my very small personal projects I want the creative part and the production part to flow together more. Like I wrote before, more like playing with action figures. :slight_smile: Separating the different tasks becomes a chore too quickly for me. I want to pick a set, a character and just make a scene come alive.
For my professional work, I’d love to perfect this to the point, where I’m so quick that I can use my approach for previs, which would be really useful. But mostly, I just want to be able to do it for the inner kid in me. :smiley:

In the end, it all comes down to accepting quality limitations or finding new ways of doing things, which is why I started this thread. Perhaps, there is a better way of doing what I’m going for, for example, working with Unreal Engine and Metahuman, or Character Creator and iClone. But that takes time, of course. For now, I would rather stick with Blender and, as you said, try to optimize my own workflow.

I’m starting to think that accepting a “rougher”, less professional look and just getting things done quickly could be my priority moving forward.

Thank you again for helping me put these things into perspective!

P.S.: As a side note, one of the things that got me into filmmaking and CGI was a game/software called 3D Movie Maker:

I suppose, to some degree that is what I’m chasing, just that today it would need to be a billion times more advanced and customizable but still quick and easy to use. I’m looking for something that cannot exist. On another side note: I’m also sad that there does not seem to anything out there that is equally easy to use and powerful for my son to play with. Good thing there are mods that make it work on Windows 10.

Here comes the very vague and meaningless (hopefully not) comment from somebody who doesn’t even do what you do, so keep your salt-shaker ready.

Have you considered using UE because of the obvious availability of the Metahuman library?
While I cannot vouch for it, I remember seeing a tutorial of an export of the Metahuman rig into Blender which included the facial rig (and therefore making it a valid appearing workflow).
There are multiple solutions available for facial tracking running on smartphone and PC.
I know that learning UE might be appear daunting, but I know (and that is from experience) that learning it and starting to do stuff with it is comparable easy, especially for a 2 decade 3D veteran.
It is extra work and it would change your workflow, but I think the advantages would create a worthwhile payoff in the long run.
While I know this recommendation is very vague and not solid, I would encourage you to go to the UE forums for example and seek out people who have done it successfully and can say more about it.
I thinks its absolutely worth it to check it out since UE and Metahuman will be developed further and while it may not be for you right now, it might be soon.

To add some more meaning to this not really very helpful post I am going a little offtopic in an attempt to help you out in more general and artistic aspects of your work.
I think your work while good could stand out some more by being more bold as well as nuanced and deliberate in almost all aspects.
I would recommend you this Youtube channel about filmmaking techniques:
Ignore the product they try to sell - it is choke full with very deep analysis of movies, directors and techniques. I found it extremely helpful, more so than any other YT channel I found over the years.

Secondly, I see you are a man of culture, doing Action movies and especially the now ubiquitous dual gun wielding / gun fu style that originated from early 80s Hongkong cinema, made famous by the director John Woo (which I absolutely love).
If you want to create more of it in the future, I would recommend to learn the history of it and some tricks along the way.

Hope this is helpful.

Hi there, sorry for the late reply! Very busy work week.

Working with UE and Metahuman definitely seems like an option! Although I’ve found that perhaps Metahuman is actually too close to realism for what I “should” be aiming for. The Uncanny Valley look of even some of the better facial mocap you can do at home is something that I still find offputting (which is why I also spent tons of time working on 2D and stylized animation styles). I will have to get into Unreal for work pretty soon though, so we will see if that changes my mind or gets me to a new place. Thanks for nudging me back into that direction! :wink:

That video was actually a remake of an experiment I did 10 years ago. :slight_smile: This is the original:

I’ve actually written a seminar paper on Sam Peckinpah and John Woo in the meantime. :smiley: I created my character before John Wick existed, even though he seems like a rip-off now. He was actually inspired by Johnny Hallyday’s character from the Johnny To movie “Vengeance” (another Hong Kong film). The remake was a bit limited by what I did 10 years ago - which isn’t to say even after all this time studying and working professionally there isn’t still tons to learn. Those are good resources!

True, but this is the nature of facial mocap, you have to be really good and put a lot of effort in to make it look great (or having a team) and/or very expensive equipment.
One of the reasons I shy away from attempting to do it myself.

Cool, is it available publicly?

I don’t think your character seems like a rip-off. First, its more in line with the classic heroic bloodshed dual gun style and not the upgraded martial arts influenced gun-fu of Wick.
Second, Woo himself took from many that came before him and his style became part of the language of action movies. I’ve read somewhere that he is one of the most cited Directors after Orson Wells and Hitchcock, so you are in good company.
Third, in order to be a rip-off you have to steal more elements that are unique.

Yeah, I remember that movie. Hallyday was great in it.

1 Like