Houdini Rigging Toolset

[MrScienceOfficer]

Hello friends,

 

So I just wanted to show some of the work I've been doing for the better part of this year now.  

 

I figured I would put it up here now as lately I've been working on taking the core API and testing out scenery tools and game asset creation.  I'm thinking I need to write a GPU tool set to build the tools I need for that and the next slew of character tools, but that'll take some time to get off the ground.

 

https://vimeo.com/139775212

 

 

There's quite a bit more going on then the video shows, but I don't really remember what's going to crash and I didn't want to have to do a second take.  The end game of this is creating a simplified pipeline for FEM or possibly PBD solved muscle systems, then being able to take those results and "bake" them into a linearly skinned rig with no extra step of creating two rigs and really great support for LODs.  Seamless interpolation and cycling is gonna be a **** to work out though.

 

The biggest tool the video doesn't show is the volume based bone weighting node, it lets you mix Houdini capsules and arbitrary volumes you "extract" off the mesh, then you can use a ramp parm and basic skin weighting tools to really fine tune the weights, then it uses a different method for normalization that guarantees all points get weighted logically but no points have there "user input" weights changed.  There's still a bit to get that working, I need to implement single bone weight visualizations as the full skinning is slooww(you'll probably notice the hang up when I entered into the capsule) even being multi threaded there's just a lot of work going on there.  

 

 

Thanks for listening to me yammer on though guys

Edited September 19, 2015 by MrScienceOfficer

[Alexey Vanzhula]

awesome!

[MrScienceOfficer]

Thanks

[MrScienceOfficer]

I want to get some opinions on something here...  Although the animation forum, thread, whatever its called is probably the wrong place to do this but ehhh whatever...

 

After likely more then 2000 hours of work on the API for tools, and not even three weeks of work on the actual tools themselves (it's actually a bit less if you only consider the tools in the videos, and that includes the time it took to make them robust enough too handle basically every situation, particularly the bone deletion code which was more complex then I initially considered, and all the work with easily changeable naming conventions and always naming the right and left sides of the characters bones properly but that's all a story for another day) I've realized how powerful this API is.

 

Not to mention the fact I didn't even get into the ability to interface directly with geometry with custom nodes, which is what this system was designed for to begin with.

 

 

So my question here is what do you guys think of me releasing the API as Open Source or as binaries?  Its certainly a possibility as its genuinely an API so the tools are compiled separately.  But what I want to know is would there be interest in a such a thing?  

 

And just generally what people think of such an idea, as I've never had a job in this industry or any formal training or anything outside of a high school education, in fact, I haven't ever spoken to anyone who knew how to rig a character, of course switching to Houdini didn't really help me out there...  But I guess I'm really asking for a bit of career advice here, particularly being someone who wouldn't mind just getting a job at a Burger King, and continuing with my tools until I can pull things together on my own, what kind of opportunities, if any, could the few tools I did show afford me.

 

I mean I have probably have quite a bit over ten thousand hours of experience mostly between Maya, Houdini, ZBrush, and C++, all focused towards character animation, but I don't think that wins me anything with a studio whose main concern is fast turn around time.  My works is always a means to creating impossibly high quality animation, with little to no resources. there's really very little time for hacking together work just to show off, as most Burger Kings don't require a reel, and caring what people think is tragically very low on my to-do list.

 

But rambling aside... I really feel I should be better informed, and ask people with professional experience in this industry before handing out over two thousands hours of mind draining work.

 

Thanks for everything guys and thanks for all the likes too, I felt so special today...

[Guest tar]

Concerning the career advice; I know of a programmer back in the day who turned up at vfx facilities and said I can do X, now I can work at a pizza place or you can pay me the same wage and I can do work for you.  In a few years they wrote some software that became part of Maya and then they ended up at ILM.   This was many years ago but something similar may work these days.

Edited September 20, 2015 by tar

[MrScienceOfficer]

Oh undoubtedly, but in today's world getting in touch with someone, even at studio that specializes in character animation, who knew a character rig from a ham sandwich seems impossible, I don't think you can just go knock on the door anymore.  I could be in New York City first thing Monday morning on the door step of quite a few studios, but I bet good money no ones gonna let me in.  I would gladly work for peanuts, almost literally, but I won't get the time of day without a highly polished reel, I've tried...

 

Making enough money, building these tools, to not starve and die, which for me isn't very much money at all, would be a dream.  To do the math here, I work about 70 hours a week when I'm being rather lazy and I could live pretty okay off of 40 bucks a day so that's 4 an hour.  I have a rent free house and/or a fiance who doesn't want kids, I'm pretty set in life...

 

But... I wouldn't know where to start.

Edited September 20, 2015 by MrScienceOfficer

[jrockstad]

Don't sell yourself short! You seem to have a strong technical background, and are obviously not lacking in motivation to devote so much time and effort to developing these tools - those are the kind of traits that will make you a very valuable employee to some studio. The issue is that what you are seeking is a rather specialized position - everyone is always going to be advertising for junior animators / riggers / compositors because those are the fields with high turnaround, whereas with something like technical rigging / tool development, it can be tough to find an open position because there are both less of them to start with and they also tend to be much longer-term. The flip side is that once you do land one of those spots, you're pretty well set. 

 

Best of luck!

[Mandrake0]

what you also can do is ask SESI for advice they have got very good connections to studios maybe they can help you.

[Guest tar]

Oh undoubtedly, but in today's world getting in touch with someone, even at studio that specializes in character animation, who knew a character rig from a ham sandwich seems impossible, I don't think you can just go knock on the door anymore.  

 

In general people in STEM fields can benefit a lot from understanding the all aspects of this business. i.e. sales and marketing:

 

David Ogilvy – Confessions of an Advertising Man'

http://www.gamasutra.com/blogs/JohanToresson/20140929/226651/Books_for_Creatives_David_Ogilvy__Confessions_of_an_Advertising_Man_Part_1.php

 

'How to Win Friends and Influence People'

https://en.wikipedia.org/wiki/How_to_Win_Friends_and_Influence_People 

[MrScienceOfficer]

Thank you link Marty and for all the other links as well, always informative and quiet memorable.

 

Thanks John for the kind words

 

Now that you mentioned that, I might actually apply for their Games Software Developer position, Francis.

 

 

I think in a way this is a way you can "turn up" in   https://forums.unrealengine.com/showthread.php?85357-Character-Artist-Software-Developer-with-a-unique-proposal in today's internet only world.

 

I'm really leaning towards just getting a job at a Burger King though, just until my tools are really off the ground then, charge an Enormous fee as a Character Pipeline Consultant.  A set of character tools for Houdini more powerful then what you can get with Maya is pretty cool, but making it more intuitive is what a company needs to smoothly transition their pipeline to new software.  

 

And the maintenance on the API is like 8000 lines of code and can handle any tool you can think up, and it's constantly being improved by SideFX, this is really good stuff.  That's why I'm going to release the code under a dual-licensing structure, similar to how Qt is licensed.  And once I finish reworking the Ui drawing code, it'll look amazing too.  You guys are gonna love it, you'll all be using it, soon enough.  I think real time interactive tools with a full set of GUI controls and the ability to easily define obscure GUI elements, will replace simple python GUIs for the expected standard of a tool.  At least for Houdini users. They are far more powerful, more intuitive and just as easy to write, way easier in my honest, completely biased opinion considering what you can get with the C++ tools.

 

If anyone's interested in the so called StickyAPI, just PM me, if people are interested I won't procrastinate setting things up for that long.

[old school]

I am very impressed with the StickyAPI toolset.

This is not trivial work. It is inspiring.

 

 

The robustness of the Sticky Character state/tool looks very good. You have pushed this so very far. It also looks like you don't accidentally bail from the tool, which is a bad behavior of many of Houdini's default tools.

 

Hotbox like control (although it doesn't have the necessary ui ques to make it sing but you know that by your self-criticism) = Not a single shelf tool was used. YES!

 

Directly capturing text input in to fields right in the viewport.

 

Bone creation finds centre of manifold geometry to ease creation.

 

Use of custom HUD's to do spline falloffs, text capturing, showing feedback, etc. makes modifications in your face and effective.
 

I like how you stay inside the Sticky Tool for the entire time. Create a viewport state that divorces itself from actual nodes. If only we had a Modeling state like this.

 

Very well done. I agree with your comments.

 

Looking forward to your toolset being made available to the community.

[Guest tar]

Brilliant!

 

Let all of us know when there is a Pay-pal/donation link I'm now looking for...

[jordibares]

Sorry I jump to this thread to late.

 

Tom, you work is very impressive and although it is true big studios have this impenetrable force field unless you have been working for X it is also true that a smaller place will die for your skills.

 

Thinking of which, why don't you come and visit us at Glassworks? I am based in the London office but we have also a lovely office in Amsterdam and Barcelona and as we are very keen in finding people with your profile.

 

hope to see you soon

jb

[MrScienceOfficer]

 (speechless)

 

 

I'm going to be setting a website for the StickyAPI, and I'll put stuff up for donations and what not, once I can get the API out there, hopefully soon.

 

 

Thank you for all the kind words guys, it means quite a lot to me.