Download: A.T.S

I originally developed this script with Adrian Fjellberg for one of his early projects. In that scene the team wanted to shoot a fireball down a temple hall. The temple floor was to crack up an break as the fireball progressed. The technique was envisioned by Adrian at the time although he didn't quite know how to attach the myriad of broken bits to a plane without having to do every single one of them manually.
After a few hours we had nailed the effect and the setup was more or less painless.

A few weeks ago I got the idea to use a similar effect in my Advanced Visual Effects module at Teesside University. Unfortunately I didn't have the script we wrote back in the day so I had to buckle down and figure out all the maths again. 

Here's a breakdown of the effect:

1: The ball rolls over a plane

2: A set volume selection is used to deform the plane; in this case a cylinder

3: Using Fracture Voroni to fracture a box I created a broken up surface. When rendered this looks smooth.

4: The Attach To Surface (A.T.S.) script is used to generate a point at every fracture, parent the fracture to it's respective points and then attach the points to the nearest triangle in the plane.

5: The result can be controlled by the size of the cylinder, the falloff used by the volume selection and the distance the affected faces in the target plane is moved

If you don't know how to install a maxScript properly please read this.

The script is used as follows:

 - Press the "Pick EditableMesh" button and choose the surface you are attaching geometry to. 

 - Choose every object you want to attach.

 - Press the "A.T.S" button.

This effect can be used on any geometry not just a plane. Here is an example of more or less he same effect on a human mesh.

Any questions/suggestions/bugs you can either drop a comment here or send me an email

First of let's get the obvious out of the way. If there is some problem opening the file containing the rig you've made wrongo number one. The rig should just work the moment you open the file. This brings me in to user friendliness. The animator is an animator, not a TD. Well obviously but what does it entail? Well it means that most likely the animator has replaced the space I use for technical howto with Disney's twelve animation principles. It also means that if they wanted to spend hours upon hours figuring out how to get a file to work they wouldn't have become animators in the first place.

So make sure every plugin and texture or any other external file the end user (in this context the animator) needs is automatically  added for them.That way the animator just get's going and the TD can focus on his/her next task and you'll save everyone a ton of time.

Once the file is open there are other problems to tackle. How will the user know how to use the rig? The most important thing is to try and place the controls on the character in logical places and make sure it’s visually clear what that control does. In addition all the controllers of the rig should be of the same type. If they are not then it’s hard to filter for them and the Animator might overlook the one point helper that does something useful if all the other controls are spline shapes. 

Another thing that helps with understanding how to operate the rig is to release an instruction video on how to use the rig. If none of these things are present the user will not be able to jump in and animate straight away and might choose to use another rig.

Going into more details of what makes a bad rig I need to mention not being bled to turn of automated systems to get full control, only having one way of controlling something or having to many. A lot of rigs have fancy controls that technically do impressive and cool stuff but in the end they end up being useless for the animator. In most cases you are actually better of not spending time trying to implement that particularly cool way of controlling something if you can do it in a simple fashion. If you really have to add it make sure it’s possible to turn it off or work with simpler and less fancy controls as well.

Lastly I’ll talk about skinning. Be really thorough when you test your skinning solution. It doesn’t matter at all that the rig is good if the skinning is bad. Not only should you test every extreme pose you can think of but you should let an animator do some quick animation tests to see if he/she can push it over it’s limits.

So to sum it all up here’s a quick list of what the should have.

1. The file must work

2. All assets should be added by the TD not the animator

3. User friendly interface

4. Video explaining the interface

5. Simple instead of complex solutions is often the best

6. Good and thoroughly tested skinning

What did this do to the animation, the production and with the animators?
In some places I had not set up the spring controller properly and so if the character moved to fast the wires or other things with spring on them would go haywire. This was fixed in some instance by going in and tweaking the numbers individually for each scene before render and in some other cases in post production. 
Production-wise it saved a lot of time since the animators didn't have to animate a lot of small wires and antennas. What the animators really thought of it I don't know but the impression I got was that they didn't like the lack of control it gives but they did like the fact that it saved them some work.

In the future I need to device a system which allows for both autonomous  animation as well as hand animation on top of that.

That aside, how can I "boss" people around in an informal group where no one has elected anyone as leader without pissing anyone off? I won't say I've got a final answer to this, but I have made me up some thoughts.

For instance we had a situation where a few of the group members were really hellbent on going in one direction while me and my co-director/producer Adrian immediately saw that it was the wrong way to go. Unfortunately there was no way of turning them without using a direct order. And that was something I wasn't prepared to do. So instead of battling these guys further we actually went along in their direction until another more concrete and better solution could be worked out. In the end everyone was happy and we avoided any bad feelings during production.

We also had an issue with people coming in late or not turning up without letting anyone know what was going on. So I made it a rule that if you weren't going to be on time you had to report it in our forum online. This actually proved to be very effective as I now suddenly knew where everyone was and what resources I had at any given day. It goes without saying that being on time and calling if you aren't is the norm in a studio but this isn't always the case at a school. 
In this case I used a direct order even though as a informal leader of a group I really shouldn't do this. Nevertheless I got away with because it was a reasonable demand that everyone either already agreed with or felt like they had to anyway. 

Bossing informally is hard because you can't always act as the boss and demand things without explaining why and your authority is constantly challenged. You have to stay at the top of your game at all times.

We were quickly joined by Xia Yefu, Remi Strand, Einar Martinsen and William Christofersen.

We are currently working on a few different things. Einar and Remi are working hard to produce as much concept art as possible while at the same time working on our storyboard. Adrian is working on a communication/sharing/collaboration platform, William is running animation tests, Xia is learning Vray and Nuke at the speed of light while I am trying to organize it all while rigging some of Einars early robot models.

It's still early days but we all think we'll be abled to pull this project off.

Edit:
The movie is now called Lights Out 

Living City

A lot of these guys could be scampering 'round town.This is more of a art film idea. The basic idea is that for whatever reason the city is alive and sometimes parts of it spring to life. What I would like to explore is how would a "newborn" part of the city react to it's surroundings.

For instance one could see a mundane part of the city and suddenly some part if it comes alive and starts exploring. Snail?Almost any little story could be incorporated into this setting but what I initially thought of was some doglike creature coming alive and after wandering aimlessly around just experiencing the bustle of the city it discovers another creature like itself and falls in love.

Then you could have some sort of obstacle stopping our creature from meeting the other. I was thinking in the lines of a tram or

bus stopping between them and when it has gone the other creature might be gone.

After a few more obstacles they could find each other again.

What creature could this turn into?This would all be filmed live action on location throughout Oslo and afterwards we'd add the 3D creatures.

Machine with design by Kjell Aukrust

I propose to make this machine in 3d, then animate it andincorporate it all in flash to make an interactive web application where the user can explore the many functions of this machine.

pose&render by Sverre Kvernmo

Finn_Rig_Download

Feel free to use this rig in any project you want, but do remember and give proper credit. It wouldn't hurt with a link to this website either.

After some work tonight I've also put together a quick video walk through of the rig. The project I was given was to rig a simple bipedal character without a face. As I've stated earlier I also had a control group consisting of Sverre, Remi and Knut. Knut also made the model.

They've all made valuable suggestions and corrections during the course of this project. In addition Sverre ended up swapping his lip-synch assignment with a fuller review of the rig which can be found here.

Things Done

I believe I've been able to make a stable and intuitive rig with a lot of freedom both in the way it is possible to pose it and how you work with it. There aren't as many "special" and "fancy" functions as I would like, however all of the important ones (or at least almost all) are present.

One of the biggest issues I had was getting the knees to function the way I wanted them to. After trying 5 different methods for a week I decided to go back to my original design and let the foot control the alignment of the knee.

I also added four different types of eyes that you can combine in any way you see fit. I know I wasn't supposed to add a face of any kind but i just couldn't resist. Anyway you can hid parts or the entire face if you so desire.

reverse foot rig

One of the coolest things I've learned to do properly is setting up proper IK/FK switching on the arms. I haven't set up an IK/FK switch on the legs because I use what's called a reverse foot rig setup and that isn't really doable with an FK system. If anyone has any ideas on how to merge the two, please leave a comment. If not, I guess I'll have to use another foot setup to get the functionality that I want.

pose&render by Sverre KvernmoNot Included

Both I and the members of my control group came up with many things that would've been nice to implement. Unfortunately a few of them were either too time consuming or I simply didn't know how to do it. A few of the ones I know how to do but didn't have time to add are;some controller to adjust the with of the hips and the shoulders, pose saving for the entire body, a better and more flexible spine and a better stretching solution for the arms.

In the end I feel like I've learned a lot and I can't wait to get started on a new rig so I may add the things I didn't do know and maybe even more. I might even try my hand at a quadruped or some other crazy stuff.

Walkthrough

The last thing I'll add is the walkthrough on how to use the rig and a final thanks to Paul Neale for making a great rigging dvd for max, and for putting out many useful scripts for free on his site(one of which is heavily used in this rig).

press the image for the tutorial

At first it may seem like the ball rotation doesn't work after all, however it needs you to animate it as the script looks for where the ball was in the previous frame and where it is now and from that decides on how to rotate it. More details on how to use it can be found on Bruce Lin's site.

So check it out and have fun with it!

not looking quite right

Gears

On the cogwheel side I have yet to get the bevel effect on the teeth themselves working properly. Other then that they're done. As long as the module value is the same the cogs will fit together no matter what radius they have.

Lego

The Legos aren't quite there yet. I'm strugling with getting the knobs in the right place but other then that they to work just as I want them to.hopefully I'll get this in the end

IK/FK controls and the stretchy system

I defined "fun to animate with" as such; if you don't have to fight the rig, find the control you're looking for rather quickly and you're not limited in the ranges of motion you want to achieve, then it is fun to animate with. I do believe I'm closing in on somethin similar to this, but it still has a few bumps that needs smoothing.

After setting up these things I sent the rig to my select few and they tried it out. From the imput I got back I got some new ideas. Some of these included to give the user the ability to either control everything in viewport or go in the modify panel and adjust things with sliders. The squash and stretch of several parts of the body were tweaked and the head was shrinked down, as mentioned in the previous post.

Another thing I wanted to adress was the fact that my IK/FK aligning was name dependent. This has been fixed with the help of Paul Neale's helpful hints and tips on several forums. The same technique was used when I set up the ability to toggle the face on and off.

Anyways this going back and forth, me fixing things and people testing the rig has been going on for some weeks now and I feel that It's about ready for a first release. I do hope I'll have the time to go back and add a few functionalities that I've thought of, such as; saving reusable poses for the entire character and adding FK to the legs.

The last thing I want to add for now is the usage of the controllers you see to the right. The top one switches between IK and FK by either sliding or pressing the on off button. This also hides the controllers in viewport not being used. The snap moves one set of controllers to the other. The stretch has been devided between FK stretch and IK stretch as I wasn't abled to get excatly the functionality out of the IK as I did with the FK. With the FK you simply turno stretching on and bove the controllers, but with the IK I had to use a slider. More on the why later.

That's all for now.