Sunday, October 12, 2014
Saturday, October 11, 2014
Screen Space vs. Physics
It's always interesting when something works in animation that doesn't abide by the laws physics. Here's an example from Chuck Jones' "Dover Boys of Pimento University".
The boys are tiptoeing over from one place to another.
First Frame
Second Frame
Third
Fourth
The strange thing is that between drawings 2 and 3 the character actually moves up in screen space. But that shouldn't happen. His feet aren't touching the ground yet. There's nothing to propel his body upward, and gravity should still be pushing it down. But if his body kept moving down there would be too much of a spacing difference between drawing 3 and 4. So the animator decided he needed to defy the laws of gravity to make the walk work smoothly. And it does. Watching the cartoon, the walk is funny, and there doesn't seem to be anything wrong with it.
The boys are tiptoeing over from one place to another.
First Frame
The strange thing is that between drawings 2 and 3 the character actually moves up in screen space. But that shouldn't happen. His feet aren't touching the ground yet. There's nothing to propel his body upward, and gravity should still be pushing it down. But if his body kept moving down there would be too much of a spacing difference between drawing 3 and 4. So the animator decided he needed to defy the laws of gravity to make the walk work smoothly. And it does. Watching the cartoon, the walk is funny, and there doesn't seem to be anything wrong with it.
Dean Cornwell Study
Value study of a painting by Dean Cornwell. The artists back in the Golden Age of Illustration were on another level.
Friday, October 10, 2014
Animation History
There's a lot of good blogs out there for animation history.
Jerry Beck has "Cartoon Research" :
http://cartoonresearch.com/
Michael Barrier has a lot of great articles on his site, as well as a number of long books:
http://www.michaelbarrier.com/
And the recently departed Michael Sporn kept a great blog with lots of animation knowledge over here:
http://www.michaelspornanimation.com/splog/
Reading animation history is useful to deal with frustration of how time consuming and difficult doing animation is. People in the past were struggling with the same issues of how to plan their shots, properly pose their characters, time out actions, make lipsynch work, deal with administrative demands, and figure out how to hit deadlines. Many CG animators get frustrated when looking at the graph editor and realizing how analytical animation can be. But take one look at a traditional animation exposure sheet or the effort Chuck Jones spent timing actions to music, and you'll see that animation has always been about careful, meticulous planning with heavy doses of logical thinking. It has to be. The process is too labor intensive and expensive not to carefully plan out. The trick, I think, is finding out how to internalize the planning so it becomes an organic part of the creative process. But knowing this on a theoretical level isn't the same as putting it into practice. This is something I'll only gain over time.
All of the above listed blogs mostly deal with American and European animation, with a bit of attention paid to Miyazaki. For a much more in-depth look at contemporary and historical Japanese animation, Ben Ettinger has a great blog called Anipages:
http://www.pelleas.net/aniTOP/index.php
Anipages is probably my favorite animation site because of how few sources exist in English on anime production. There's so many movies Ben talks about that I wish I could track down and watch, though most don't exist in digital form. And the forums are a great place to get feedback. Every so often someone like Peter Chung or Bahi JD will stop by with some amazing knowledge about a production or how to make better animation. Highly recommended.
Jerry Beck has "Cartoon Research" :
http://cartoonresearch.com/
Michael Barrier has a lot of great articles on his site, as well as a number of long books:
http://www.michaelbarrier.com/
And the recently departed Michael Sporn kept a great blog with lots of animation knowledge over here:
http://www.michaelspornanimation.com/splog/
Reading animation history is useful to deal with frustration of how time consuming and difficult doing animation is. People in the past were struggling with the same issues of how to plan their shots, properly pose their characters, time out actions, make lipsynch work, deal with administrative demands, and figure out how to hit deadlines. Many CG animators get frustrated when looking at the graph editor and realizing how analytical animation can be. But take one look at a traditional animation exposure sheet or the effort Chuck Jones spent timing actions to music, and you'll see that animation has always been about careful, meticulous planning with heavy doses of logical thinking. It has to be. The process is too labor intensive and expensive not to carefully plan out. The trick, I think, is finding out how to internalize the planning so it becomes an organic part of the creative process. But knowing this on a theoretical level isn't the same as putting it into practice. This is something I'll only gain over time.
All of the above listed blogs mostly deal with American and European animation, with a bit of attention paid to Miyazaki. For a much more in-depth look at contemporary and historical Japanese animation, Ben Ettinger has a great blog called Anipages:
http://www.pelleas.net/aniTOP/index.php
Anipages is probably my favorite animation site because of how few sources exist in English on anime production. There's so many movies Ben talks about that I wish I could track down and watch, though most don't exist in digital form. And the forums are a great place to get feedback. Every so often someone like Peter Chung or Bahi JD will stop by with some amazing knowledge about a production or how to make better animation. Highly recommended.
Monday, October 6, 2014
Koike Studies
Studies of major poses in the running animation from "World Record" by Takeshi Koike. One of my favorite animated shorts from Japan, though all the Animatrix shorts are pretty excellent. They make that whole franchise worth it.
Sunday, September 28, 2014
CGI Anime
A number of Japanese studios have been experimenting with cell-shaded 3D in the past decade. It's useful, saving time and allowing more distinct animation frames to be created with less time spent on inbetweening and coloring each drawing. In my opinion there are two really standout studios doing this kind of work, Sanzigen and Kamikaze Douga.
An example of Sanzigen's work:
An example of Kamikaze Douga's work:
Looking at Kamikaze Douga's work in particular, they use CGI technology to add texture to their rigged models. And they use traditional FX animation to provide some visual variety. Sanzigen uses the great Pencil+ plugin for 3DS Max to be able to explore a range of styles with lines and shadows that stay visually consistent and appealing. It's light-years away from the toon shader in Maya. The team that programmed Pencil+ made sure that it attempts to capture good shape design in the shadows. It adds a sense of intention to the CGI "drawings" that computer animation usually lacks.
One thing I don't understand though is why the recent series "Knights of Sidonia" and "Ronia the Robber" are being done in CGI. Both have relatively simple character designs and would look much more appealing if they were drawn. And neither series is taking advantage of the Pencil+ plugin, so shadows sometimes get that jagged, unintentional look, making the jerkiness of the animation seem like a mistake. I wish the studio, Polygon Pictures, would at least get the Pencil+ plugin, but it won't fit in their pipeline unless they're using 3DS Max.
Here's an example of an ideal form of CGI/cell-shading. This Gobelins short uses texture combined with hard and soft shadows and great art direction for a distinctive look. Plus a really funny script.
An example of Sanzigen's work:
An example of Kamikaze Douga's work:
Looking at Kamikaze Douga's work in particular, they use CGI technology to add texture to their rigged models. And they use traditional FX animation to provide some visual variety. Sanzigen uses the great Pencil+ plugin for 3DS Max to be able to explore a range of styles with lines and shadows that stay visually consistent and appealing. It's light-years away from the toon shader in Maya. The team that programmed Pencil+ made sure that it attempts to capture good shape design in the shadows. It adds a sense of intention to the CGI "drawings" that computer animation usually lacks.
One thing I don't understand though is why the recent series "Knights of Sidonia" and "Ronia the Robber" are being done in CGI. Both have relatively simple character designs and would look much more appealing if they were drawn. And neither series is taking advantage of the Pencil+ plugin, so shadows sometimes get that jagged, unintentional look, making the jerkiness of the animation seem like a mistake. I wish the studio, Polygon Pictures, would at least get the Pencil+ plugin, but it won't fit in their pipeline unless they're using 3DS Max.
Here's an example of an ideal form of CGI/cell-shading. This Gobelins short uses texture combined with hard and soft shadows and great art direction for a distinctive look. Plus a really funny script.
Tuesday, September 16, 2014
Animation Bits
I mostly do CG but have been trying some hand-drawn stuff that has a minimal number of frames. Many of these are based on reference, though none of them are rotoscoped (if they were the forms would stay more consistent, lol).
Monday, September 15, 2014
Blue Blazes
Blue Blazes ("Aoi HonÅ", also a pun on the main character's name) is a Japanese comedy-drama show about animators and manga authors. Specifically, one aspiring manga-author/animator and his competition with the future stars of Gainax like Hideaki Anno. It's an interesting, dramatized look at Japanese art/film-student life, and demonstrates the importance of being in the right place at the right time. If all these guys hadn't gone to school together and formed these relationships and rivalries, would they have ended up with such enduring/successful careers? It reminds me of the CalArts class that produced John Lasseter, Brad Bird, Henry Selick, Tim Burton, and Chris Buck. Sometimes you get the right convergence of incredible talents, who all work off each other to get better and better, and are able to create the right opportunities for themselves. Especially if they're mentored by the right people.
The series is a lot of fun because of its exaggerated acting. I haven't seen much live action Japanese comedy so I don't have any idea if the stylistic conventions this show uses are normal, or are meant to represent that the characters share more in common with cartoon reality than the ordinary world. Its clear the show runners have done their research, and are committed to demonstrating the trials of an aspiring creative professional. Here is Hono externalizing his sense of not being good enough:
Also, the actor playing Hideaki Anno has one of the most threatening smirks I've ever seen. I might have to borrow that expression when animating a character filled with both satisfaction and disgust.
I recommend the series to any anime fans or aspiring animators. Especially because it contains some animation production history, such as how animation cells work. Thank god we don't gotta paint those no more. Maybe they'll go into x-sheets in a later episode.
The series is a lot of fun because of its exaggerated acting. I haven't seen much live action Japanese comedy so I don't have any idea if the stylistic conventions this show uses are normal, or are meant to represent that the characters share more in common with cartoon reality than the ordinary world. Its clear the show runners have done their research, and are committed to demonstrating the trials of an aspiring creative professional. Here is Hono externalizing his sense of not being good enough:
Also, the actor playing Hideaki Anno has one of the most threatening smirks I've ever seen. I might have to borrow that expression when animating a character filled with both satisfaction and disgust.
I recommend the series to any anime fans or aspiring animators. Especially because it contains some animation production history, such as how animation cells work. Thank god we don't gotta paint those no more. Maybe they'll go into x-sheets in a later episode.
Sunday, August 31, 2014
Twist Distribution Method
I've been experimenting with twist distribution lately in my rigs to get nice deformation. This is crucial in CG because a rig without good deformation is like a bad drawing. And in CG a bad drawing doesn't even have a childlike charm. It just look like someone put a magnet too close to the computer.
There are a bunch of methods for twist distribution. You can do it with locators, expressions, constraints, or math nodes. One method on Creative Crash got my attention. It uses an IK Handle to prevent one joint from twisting, and then you can apply your twist distribution down the line. Then someone in the comments pointed out a better way to do it using math nodes instead of a locator & constraint. But going back and forth between the comment and tutorial itself can be annoying, and the guy in the comments doesn't include pictures, so I've created a condensed explanation. The original tutorial is here:
http://www.creativecrash.com/maya/tutorials/character/c/advanced-non-flipping-upper-arm-roll
First create a normal IK Chain arm or leg. Then attach the deform chain roll joints to the shoulder or, in the case of a quadruped, to the torso. Apply an RP IK to the first roll joint and zero out the pole vector so it won't twist.
Then create two Multiply/Divide nodes with the operation set to "Multiply" and the 2X value set to -1. Plug rotate x of the bind bone (the one that is part of the twist-distribution system) into the first Multiply/Divide node, on input 1X. Then plug the rotate X of the driver bone into Input 1X on the second Multiply/Divide node.
Then create a Plus/Minus/Average node, set the operation to "Subtract" and attach the bind bone Output X into input 1D[0] and the driver Output X into input 1D[1]. Now attach the output of the Plus/Minus/Average node into a new Multiply/Divide node, into both input 1X and 1Y. Then divide each of those numbers by whatever value you think will create good twist distribution. I chose 3 for Input 2X and 6 for Input 2Y. Then plug Output X into the first bone that will twist, and Output Y into the second bone that will twist. If you have more twist bones, create more Multiply/Divide nodes as needed.
The node network will look like this:
The bones in the viewport look like this:
If anybody reads this and has questions/suggestions, leave a comment. This is far from the only way to do twist distribution, but it's a method I've found to work.
There are a bunch of methods for twist distribution. You can do it with locators, expressions, constraints, or math nodes. One method on Creative Crash got my attention. It uses an IK Handle to prevent one joint from twisting, and then you can apply your twist distribution down the line. Then someone in the comments pointed out a better way to do it using math nodes instead of a locator & constraint. But going back and forth between the comment and tutorial itself can be annoying, and the guy in the comments doesn't include pictures, so I've created a condensed explanation. The original tutorial is here:
http://www.creativecrash.com/maya/tutorials/character/c/advanced-non-flipping-upper-arm-roll
First create a normal IK Chain arm or leg. Then attach the deform chain roll joints to the shoulder or, in the case of a quadruped, to the torso. Apply an RP IK to the first roll joint and zero out the pole vector so it won't twist.
Then create two Multiply/Divide nodes with the operation set to "Multiply" and the 2X value set to -1. Plug rotate x of the bind bone (the one that is part of the twist-distribution system) into the first Multiply/Divide node, on input 1X. Then plug the rotate X of the driver bone into Input 1X on the second Multiply/Divide node.
Then create a Plus/Minus/Average node, set the operation to "Subtract" and attach the bind bone Output X into input 1D[0] and the driver Output X into input 1D[1]. Now attach the output of the Plus/Minus/Average node into a new Multiply/Divide node, into both input 1X and 1Y. Then divide each of those numbers by whatever value you think will create good twist distribution. I chose 3 for Input 2X and 6 for Input 2Y. Then plug Output X into the first bone that will twist, and Output Y into the second bone that will twist. If you have more twist bones, create more Multiply/Divide nodes as needed.
The node network will look like this:
The bones in the viewport look like this:
If anybody reads this and has questions/suggestions, leave a comment. This is far from the only way to do twist distribution, but it's a method I've found to work.
Thursday, August 28, 2014
Introduction
Hi! My name is Chris Scott and I animate things, rig stuff, and do all the other elements of CG with varying degrees of skill. I recently graduated from Sheridan College's Digital Character program, and figured I'd start an actual blog instead of just posting random bits on facebook.
So a little about me. I started learning CG a little over 2 years ago now. I graduated from Rutgers University in 2012, after doing a thesis project on T.S Elliot's "The Love Song of J. Alfred Prufrock". It was a two part project, involving a written thesis to appease the English department, and an animated, visual component to please myself (and by doing the film I got to write 30 less pages). It was done in Flash, and was my first attempt at film-making. It actually amassed a surprising number of views on youtube, probably because it concerns one of the most famous poems in the western world, and less because of the animation quality (there's hardly any real animation in the entire thing.)
After graduating I knew I didn't have the kind of skills people would pay me for, so I spent the summer learning Maya and made this film. It's rendered in an odd aspect ratio, I don't know what I was thinking at the time. Also, check out that herky-jerky character motion and demented camera movement. Fun stuff.
But hey, it got me into Sheridan's Digital Character program, so there ya go. While I was waiting to go to Sheridan I made another film, this time with sound, about monkey-creatures summoning hammers from an alternate dimension and fighting with them. The storyline is unclear, the ending is ambiguous, and the computer's attempt at cell-shading is...eh. The only thing I still like about it is the lightning and smoke effects which I did frame-by-frame in Flash. Another good learning experience.
Then I went to the Sheridan program where I made this film. Mike and his Grizzly Bob accent is the funniest bit:
And then after graduating I applied to some jobs, didn't get them, took an iAnimate class, rigged some more things, and am now once again on the job hunt! This is my latest reel:
So on this blog I'll post animation tidbits that interest me, as well as any new stuff I make. And then maybe one day, once I actually get good enough, someone will read this. And that'll be cool.
Subscribe to:
Posts (Atom)