Donate!BOOKMARK our Homepage!
VOLUNTEERASIFACONTRIBUTEASIFAEXPLORE
LINK TO USASIFAJOIN ASIFAASIFAThanks!

Monday, August 13, 2007

Opinion: Bakshi Speaks To CGI Animators

rotoscope
In 1914, Max Fleischer invented the rotoscope as a time and labor saving way of producing animation. He soon came to realize that although the device was a great aid in effects and technical animation, it was a poor substitute for character animation.
motion capture
In 1986, engineer Ernie Blood developed motion capture techniques as a time and labor saving way of producing animation. A decade and several mocap features later, many CGI animators are coming to the same realization that Max Fleischer and his staff had more than a half century ago.

The ASIFA-Hollywood Animation Archive documents the golden age of hand drawn animation, but it isn't intended as a resource exclusively for 2D animators. I encourage CGI artists to think of themselves as animators and build upon animation's rich history instead of reinventing it bit by bit. Animation is animation. Pencils and computers are only tools.

No one today has as much experience with hand drawn animation and rotoscoping than the legendary director, Ralph Bakshi (See related article Bakshi Phone Doodles). I've asked him to speak "animator to animator" to CGI artists and pass along his observations about the things that really count when it comes to animation. -Stephen Worth


Ralph BakshiRalph BakshiBAKSHI SPEAKS TO CGI ANIMATORS

Frame to frame animation eventually came to a grinding end. I'm not sure which generation of young animators were at Disney redoing and relearning the tradition of making boring films and recreating cliched motion when it expired. Except for Jim Tyer, "Modern Animation" and Ralph Bakshi, animation was dying- while doing the same old thing. Big money and animators never really followed Bakshi, "Modern Animation" or Jim Tyer. They just rehashed its past.

Engel at UPA
(Read Chuck Jones' article on the failure of "Modern Animation")

UPA failed because it was nothing more than elitist designers trying to animate on museum walls. Content was unimportant to them, really. Matisse or Picasso were more important. Bakshi was hounded out of the business by controversy. And you'd be surprised how many animation directors at Terrytoons disliked Jim Tyer's work because it didn't look like Disney- or anything else for that matter. Terry kept him on because his weekly footage output was so large.

Bakshi's Lord of the Rings
(See the gallery of images from Lord of the Rings on RalphBakshi.com)

Lord of the Rings was done in rotoscope animation because rotoscope made it physically possible to do it. You couldn't do Lord of the Rings in less than 25 years using traditional animation. Thirty years later- Wow! Along comes the computer... "We can do Disney story animation with another look and sell it back to audiences." Of course, I would have used computers and motion capture if they had been around during my day. But I turned to Tolkien to try to change the kinds of stories animation told. My city films were being thrown out of theaters.

So, what's the argument here? Unless hand-drawn animation finds new creative story approaches and new creative drawn motion exaggerations, it will look as it always looked at the end- faded and drawn. There'll be no great interest for it either. Computer animation has the exact same problem. Computer animation will eventually grow old, just like hand-drawn animation, unless something new happens. It will fall into manneristic boredom if it continues to endlessly redo what's already been done before. The success and the money will always follow the creative artists who take either of these two mediums and do something different with it.

A lot of people remember and love Jim Tyer's animation today because he really did something different with hand-drawn animation. He didn't follow the crowd.

Jim Tyer Animation
(See Jim Tyer's work: Terrytoons: Barnyard Actor / Funny Animal Comics Part One and Part Two


Ralph Bakshi 2007

Buy Me At AmazonUNFILTERED: The Complete Ralph Bakshi isn't one of those "art books" with postage stamp sized pictures floating in oceans of tasteful white space and huge text blocks of scholarly blather that crowds out the images. It's just pictures, pictures and more pictures... along with just enough text to put them in context. The book is organized to show Ralph's career from his earliest days at Terry-Toons, to his groundbreaking features, to his revolutionary TV work, to his most recent fine art paintings. Even if you think you know all there is to know about Bakshi, this book will grab you by the lapels and shake you and show you things you've never seen the likes of before. Click through the link to pick up the Bakshi book at Amazon.
.

Labels: , , , , , , , , , , ,

StumbleUpon Toolbar Stumble It!

9 Comments:

At 1:44 AM, Blogger Mr Atrocity said...

This is an interesting area and one which is not especially well understood outside "the industry". I have worked in CGI effects for ten years and, though I am an artist who specialises in lighting and texture rather than character animation, I have worked on many films that feature character work. I can tell you assuredly that the extent of motion capture's use is greatly overplayed, mostly by the publicity departments of the studios who would like everyone to believe that everything is achieved with the latest technology rather than by talent and hard work.

I cannot think of a film where mo-cap has been used and the un-edited mo-cap data is seen on screen. At the least the data will be augmented with keyframe animation (i.e. by hand). Oftentimes the mo-cap merely serves as reference for the animator and the final performance that the audience sees on screen is in fact entirely done by hand.

I think much of this lack of emphasis on the hard work of VFX artists stems from the desire of PR people to show that a single perrson is the author of a body of work - it's a concept the public finds easier to deal with than a whole team. Artists are supposed to be solitary types after all. Therefore the idea that actor X "played" the CG character so-and-so and the performance was captured with mo-cap and that's what you see has appeal for marketers. Sadly it is a long, long way from the truth and the art of animation, even in VFX heavy summer blockbusters, let alone CGI animated films is far from dead.

 
At 12:22 AM, Blogger Broadening Horizons said...

Great article. Can't wait to read the rest.

You know Mr. Bakshi has been talking for awhile about having an online teaching course. Could this potentially evolve into that? (It would be fitting as the $100,000 course is sponsored by John K)

For those of you who want to read more of Bakshi's thoughts don't forget he's got a book coming out.

Also, I do an interview with him regularly. Linked Below.
http://www.ralphbakshi.com/forum/viewtopic.php?t=1042

Sincerely,

Brother Rabbit
www.RalphBakshi.com
www.myspace.com/RalphBakshi

 
At 1:14 PM, Anonymous Anonymous said...

I still think that comparing motion capture with Rotoscope is bullshit. Yes the basic idea that the animator doesn't come up with the performance is the same in both but that's really where the similarity ends. In rotoscope, you draw over the top of a live performance, so from the top you have a person interpreting and possibly charactering the movement. In motion capture you have a computer track dots in physical space, you then can attach models to those dots and then you should have your performance. The problem with motion capture is that the computer ends up loosing about 80% of the performance. If you took an actor and stripped away all their flesh and just had their bones moving around, then that's what you would get.

The animator then has to go in and try and recreate the 80% that the computer didn't capture. Your job ends up being more technical then creative. You have more in common with a plumber then an artist. So instead of charactering a performance, like you would in rotoscope, you are working as a technition trying to fix what's broken. It's a miserable and expensive way to make a film.

 
At 7:46 PM, Blogger Stephen Worth said...

Ask someone who has done hand rotoscoping. The inherent problems are exactly the same as you describe for motion capture. When you mechanically translate complex shapes and motion blur of the live action into simple lines, it's inevitably that technical errors in interpretation creep in from drawing to drawing. The roto artist has to fix these errors at the same time he is trying to breathe life back into a live action performance that has had the life sucked out of it through tracing.

Instead of building and creating a performance, the rotoscope artist spends most of his effort trying to prevent what little there is in the scene from slipping away. Sound familiar?

See ya
Steve

 
At 2:47 AM, Blogger Mr Atrocity said...

I'd just like to add a couple of points to what "Anonymous" said. The analogy of stripping away an actor's flesh and having the mo-cap just capture skeletal motion isn't quite correct. Generally many points will be sampled for areas of interest, so shoulders and other areas that have very complex motion will have densely sampled data and areas where the range of possible movement is limited, like a knee joint, perhaps only one. This ensures that the maximum possible range of motion is sampled.

What happens to that data afterwards depends upon the project. As I mentioned before, oftentimes it is used only for reference, like a mood sheet, and if the data itself is used it will have a great deal of hand animation added to craft the performance that the director wants.

Motion capture is a very useful tool and it is a lot more sophisticated than merely creating a representation of skeletal movement but eventually it is the director's decision how much of the sampled performance should be kept and how much animated by hand. The best performances that I've seen usually come from a combination of the two.

Hope that provides a little more info. Thanks,

Mr. A

 
At 1:07 AM, Anonymous Anonymous said...

One other technical point regarding traditional 2D rotoscope is that it's usually shot on twos, meaning every other frame of live reference is traced. Were it on ones, it would jitter like crazy. CGI and mo-cap is all on ones because that's how it mechanically works. There was an Osterizer commercial in the early 1980's in which the client insisted that the rotoscoping of the "Johnny O" character be on ones and it jittered like hell. Only nerds will ever read this.

 
At 5:26 PM, Blogger appleanimator said...

RB and SW make very good points. Rotoscope and Mocap are tools, and it's up to the artist to build on info what the tools offer. Most of the time, they do. But in some businesses, such as video games, time pressures pretty much force animators to use the data as-is.

Meanwhile, the crutch so many producers rely on of "rotoscoping" old animation to reuse in a new project often results in worse results than just making the animation from scratch.

 
At 7:30 AM, Blogger Steven E. Gordon said...

Good response Steve...look at that ...we're agreeing again;)
Mo-Cap and Roto are almost exactly alike especially in the fact that it allows less qualified animators to create B and A work on a schedule and a budget. It also alolows really good animators to crerate superior work. In both cases there are those that use it and make it look as good (or better) as non-roto/non mo-cap and those that can't.

 
At 6:34 PM, Blogger Weirdo said...

Mr. Bakshi preaches the truth. They've been telling the same stories for years in animation. Fantastic post.

 

Post a Comment

Links to this post:

Create a Link

<< Home