Tuesday, May 29, 2007

Screen Business

In Mark Mayerson’s recent post about Pinnocchio,
he offers this neat bit of observation…

One thing this sequence excels at is the use of stage business. In too many modern animated films, characters stand around yakking with nothing else to do. The animator is stuck trying to find arm gestures and head bobs that go with the dialog. In this sequence, Stromboli is working with a prop in almost every scene and using the prop to perform specific actions.

That nails it pretty well, don’t you think?

So as usual, I got to thinking, well, why is it this way? The best answer I could come up with is the heavy production process that every CG asset labors under. If it exists on screen, then it has to have the following tasks completed-

  • designed
  • color designed
  • modeled
  • rigged
  • UV’s
  • textured
  • render tested
  • animation tested
  • pipeline asset coordination & compliance
  • multiple version publishing (render version, layout rig, animation rig, etc.)
  • placement & tracking within the scene as it transits from dept. to dept.

And that’s the basic minimum for everything you see on screen, from the smallest ball point pen or coffee mug to a complex machine. The task list is the same, and each task is the domain of a different department, usually. And each task needs to be tracked for schedule, modification, approval, etc. In other words- props in CG are expensive, even “simple” ones. With such a heavy price to pay, naturally the prop list will take a hit when producers are looking for ways to maximize their production pennies. Not to mention that CG character interaction with other objects is still one of the biggest technical pains in the gluteus maximus that exists in the animator’s workday. It’s just a beating. So we take stuff out of the character’s hands and leave the animator to try and find some interesting way to combine hand gestures- of which the general library tends to be somewhat shallow.

Prop interaction (or ’screen business’ in old days speak) allows for a greater variety of actions for the characters to perform, immediately adding a sense of depth and believability to the milieu onscreen. And when you can draw you props as easily as anything else, well, it’s no wonder that 2d films have more of it. A fine example in a recent traditional film was the sequence where we see the introduction of Mr. Silver in Disney’s Treasure Planet. The stuff they have him do in that 2 minute section of him cooking and then sharing his stew would cost many, many millions of dollars in CG.

As in other areas, a good place to look for comparison to CG in this regard seems to be puppetry or stop-motion animation. But even there they’re ahead of CG on ease of implementation. A physical model of a clay coffee cup takes about 15 minutes to make and bake and the animator can put it in the character’s hands easily with a bit of sculpy on the backside. Puppeteers using the glove hand can just pick up the object after they come back from buying it at Target or Walmart. Compared to the flaming technical hoops that CG animators need to jump through just to have something for a character to pick up, and then have that character pick that something up- it’s child’s play.

It’s just another example of how the technology of the medium can get in the way. What would revolutionize things is to come up with an asset production paradigm that was as intuitive and simple as it is in 2d films. Unfortunately the massive command and control requirements of a large scale CG film kinda preclude such low level action.

Seems to me there could be some innovation of thought that would streamline the process and open some doors creatively. Maybe there’s a technological solution, but I think CG practitioners tend to look to the Technology Genie to give them their magic wishes a little too quickly. I’d be interested in hearing if some production of size had managed to find a way to quickly and effectively introduce props or assets in a way that still allowed it to get onscreen with acceptable quality without such a heavy task/workload toll? Do any commenters have any examples or anecdotes that might be had by way of improvement?

Saturday, May 26, 2007

Whoa…very cool lipsync lesson- Aardman style

7.jpg
This is like finding a $100 on the sidewalk. How is it that I never saw this when it first came out?

Stefan Marjoram offers a quick tutorial on doing lipsync in Maya in the Aardman style. It includes his incremental Maya work files and the sound track, too!! Of all the studios in existence today the one whose work I enjoy more than any is Aardman. Yes, even more than Pixar, though I’ve heard they’re pretty good, too :) (StudioAKA, home of director Marc Craste, would probably be my third choice. After than I probably stop keeping track). Second, after seeing his short film The Deadline I’ve been a fan of Stefan’s approach to animated acting. He has the impeccable ability to find those little subtle things and bring them out of the audio track in a fun and unique way (something he talks about briefly in his tutorial, by the way). His characters have such amazing sense of depth and life to them and he uses the simplest of methods to pull it off. He’s one of those super talented guys who can’t seem to make anything I don’t like. So to be able to poke around in some of his work files- well, I kinda geeked on it. His method of manipulating the puppet is rudimentary compared to higher end GUI’s and controls, but to see the principles of how he builds his performance under the hood was like a small treasure. My guess is this tutorial was first published a few years back. Anyhow, if you want to try a different, fun, interpretive approach to facial animation with an Aardman flair, then check this tutorial out. You won’t be sorry.

Tuesday, May 22, 2007

Following Mayerson

Mark Mayerson has decided to post his master’s thesis (in chunks) on his blog. The paper is title Six Authors in Search of a Character: The Collaborative Nature of Performance in Animated Films. He is exploring the challenges facing animators in the creation of a cohesive character performance. The prime obstacle to doing this well is the fact that animators rarely ever have 100% control over a character for the entirety of a film or project.

Mark has been a thoughtful commentator on animation for a good long while now and I’m always interested in his writings. He doesn’t usually get too caught up in navel gazing- a common problem with animation writers. I especially like his intelligent analysis of previous animation efforts. The stuff he’s posted about Pinnocchio recently has been wonderful and insightful. I’m willing to bet that this effort will be no less interesting.

Monday, May 21, 2007

APT Session #2 Student Reel

ClickToPlay_APT2Reel.jpg

Well it’s two weeks late, but it’s done- I am proud to present the APT Session 2 Student Highlight Reel.

Session 2 was a great session in ye olde’ APT. I want to express a huge “Thank You!” to all of my APT session #2 students. You guys were awesome!

I enjoyed it a lot, the students were great and they made some wonderful progress. In doing the student reel for this session I decided to skip the “before & after” approach and just go for the “after”. There are a few reasons behind that decision, none worth going into here. But if you look at the “previous experience” info for each student you’ll see where they’re coming from. We had a wide range of folks- absolute newbies, old pros, students of varying degrees of advancement, young pros and a Flash animator who had only tinkered in CG before. And while I don’t think even the students themselves would say any of the pieces on this reel are “perfect” - every piece could use a little tweaking here and there (who wouldn’t say the same about their own work?) every one of them would agree that the work here is much better than what they were doing before they started the APT. Everybody had their own unique curriculum and set of exercises and they all responded great to the teaching. I always find it rewarding to see that folks are showing greater confidence and ability after one of these APT sessions. Putting the reel together certainly reminds me that the hard work of running the program is well worth it.

APT Session #3 is sold out and I’m already excited to get going with this new crop of students. If you missed out this time or are thinking about taking part in the APT keep checking back here to find out when APT Session #4 will be. And once more, a big Congrats to all of my APT Session #2 students- you guys did wonderfully.

Sunday, May 20, 2007

Spring Cleaning for the Site

Last night I finally got fed up with how bloated and slow Firefox had become as a browser. Remember the early days of Firefox? It was like a breath of fresh air compared to Netscape and Internet Explorer. Light, fast, simple. Ahh, those were the days- a mere 3 years ago, fer cryin’ out loud.

Anyhow, I was looking around for a faster, less bloated feeling browser. So I tried a few different flavors. I found something though- my site looked like poo on a platter in more than a few different browsers. Of course I knew why- when I updated the style of the site last year I wasn’t too up on CSS and PHP and all that jazz. So I hacked and sawed and filed the edges, taped, glued and nailed some pieces into place and voila! I had a new, nice looking site held together by the webpage equivalent of duct tape & bailing wire. Ahem.

So after finding my new browser to mess with (I’m going with Opera, in case you were wondering. I’m liking it a lot so far) I decided to get a new template for the site/blog that was a lot more compliant, flexible, easier to maintain and update, blah, blah, blah. I went with Mandigo and just tweaked a few things here and there- certainly far fewer tweaks were necessary this time because this new theme supports sidebar widgets for ye olde’ blog.

Another nice thing is the theme handles some nice features internally, which allowed me to turn off a few plug ins that I was running. So we should see an overall speed boost as well. The site was a bit too pokey to load before, for sure. This will be much easier for me to maintain moving forward. And it will look better across the browser spectrum. I just need to twiddle with the color of the text a bit- it’s a little too electric for me right now, but that should be an easy fix.

Not that anybody cares, but I feel better about it.

Saturday, May 19, 2007

APT Session 3: Sold Out!

The title says it all- we filled up all of our spots for APT Session #3 for this summer- and this without me even getting session #2 reel done. Oops! (It’s coming by Monday for sure!)
Thanks to all the students who signed on. I’m eagerly looking forward to getting going in early July.

For those of you who are interested in another APT session, good Lord willing I hope to have APT session #4 before the end of 2007. However I know that life- especially my life- is very unpredictable. So if we’re not able to have session #4 before the end of 2007 then I’m pretty confident in projecting an early 2008 start date for APT Session #4. As always keep an eye out here for dates, info, etc.

OK, back to work for me!!!!

Housekeeping: Links, tutorials, rigging and such

Reader Patrick Barroca felt he would offer another attempt at translating my old Pose-to-Pose Organized Keyframe tutorial into French. Now I’m no French scholar, but he felt there were issues with the first translation. So I’ve put Patrick’s French translation alongside of the French translation done by Pascal Hang. I’ll let all you francophiles decide which is better. All I can say is a big thanks to both Pascal and Patrick for their kind efforts to translate the tutorial into French.

Another reader, Dwayne Elahie, is a technical director. He has created a video tutorial (WMV or QT) that adapts a version of my facial rigging into 3dsMax using bones. So for you Max folks check it out to see the concept applied in your particular flavor of CG-ness.

I’m almost done rigging another new character. He’s about the 4 or 5th different R&D rig I’ve made in the last year or two- all in an effort to keep refining what I want to use in my own rigs with an emphasis of reducing the technical distance between the animator and the CG puppet. At times I have folks write me and ask if I’d be willing to rig their characters (for hire, that is. I don’t rig characters for free unless I REALLY like you and I have known you for many years. Heh. All I can say is that I’m not a terribly good production rigging TD. I’m more about R&D, exploring new ideas. The practical necessities of production rigging are too exacting for my tse-tse-fly mind to keep track of. Production rigs tend to need watertight solutions- be as close to unbreakable as possible. My rigs, much like my MEL scripts, are buggy and leaky affairs. Fine for me- I understand their quirks and problems- but a sure fire recipe for frustration and pain for anybody else. So I’ll take this as another opportunity to suggest that Maya users who need solid, production ready rig making capability to spend $99 and get Anzovin’s The Setup Machine for Maya. It does a fine job and it does it affordably, quickly, reliably and has a built in level of consistency that all production environments require. Not to mention it’s a pretty nice rig. Last I heard Raf’s team is developing a facial auto-rigging tool as well. I don’t know where that’s at, but you can check out some early R&D videos and such here. Rumor is it may ship soon.

Tuesday, May 15, 2007

Behind on stuff

Good things are popping up in our water filter ministry project here like spring thunderstorms over Oklahoma. As such I’ve been busy with that and I’m a bit behind on other stuff. What stuff? Well, the APT Session 2 Student Reel, for instance. It’s almost done, I’m just making a new bumper for it. Hopefully it’ll be done by the weekend. We’ll see!

Meanwhile there are still two student slots available for this summer’s APT Session 3. I’ve been so busy that I haven’t really had a chance to promote session 3, but even so it’s almost full. If you want to grab one of the last two spots be sure to follow the link to learn more.

Rats!

I know I’m a bit behind on this, but here’s something I thoroughly enjoyed…

remy.jpg

This 9 minute sneak peek of Pixar’s upcoming Ratatouille was a lot of fun to watch. Good story telling, good filmmaking in this little clip. I’m surprised that Disney felt the need to put out 10% of their film online free before release. Somebody in marketing must be really nervous about this thing flopping. I dunno, for all the people who say that a rat in a restaurant is gross and offputting, I think it’s a great premise that kids will automatically take a shine to. Only grown ups think rats are gross. And when you make your rat look like a Jim Henson inspired muppet you just can’t help but think he looks cute.
The acting didn’t feel rotoscopy - which is a pleasure to see. I thought the animators did a marvelous job making Remy feel like a rat while still maintaining his unique character and personality. I think the secondary human characters had great design, but the primary ones kinda suffered from the “Alice In Wonderland” disease of toning down the design to make them more generic and “accessible”. As usual the details are handled well and - again as usual- Pixar does a very good job of putting the details in the proper visual heirarchy. So far this sneak peek has me keen to go see the film. I think this has a lot of promise to be a nice come back for Pixar after the soporific Cars.

Now, for marketing that does the absolute opposite…


What the…?

Thursday, May 10, 2007

Vinyl’s Attack = Fun!

As a huge fan of 1950’s era black and white space horror B-flicks this little 95 second bit of fun just made me smile.
Vinyl’s Attack! (click image to play)

VinylsAttack.jpg

I saw this over on 3dTotal.com. Congrats to Niko for making this great little homage to one of the best film genres ever to have a crowd munch popcorn in front of it. Check out the other short subjects on Niko’s MySpace page- the Moovie spot is pretty cool, too. I dunno who Niko is, but it appears he is French, he has a neat eye for stuff and he spent some time at Supinfocom (gee, no surprise there).

Sunday, May 06, 2007

Facial control GUI stuff

Enough wheels have squeaked, so here comes some grease! Actually this is something I’ve been meaning to do for a few months already, ever since I first showed an R&D character rig I was messing with back in January. Seeing as the next few months of my life are going to be pure craziness on a stick I figured that now’s as good a time as any to get it over with.
And so with that enthusiastic opening, may I present to you the How Do I Make That Nifty Colored Ball Facial Control Rig Thingie (or HDIMTNCBFCRT for short.)
Because the idea of making a written tutorial gave me hives I just did a screen grab walk thru of the concept. Thus it is a 29 minute movie. Yeah, you heard me. 29 minutes. Hopefully it’s a worthwhile half hour watch. The good thing is it’s the second attempt. I had the first one almost all done when the computer crashed while I was recording and I lost the movie- and the Maya file. Oops. Forgot to save. So you guys get to benefit from that rehearsal run.
FaceDemoClick.jpg
For those who like to poke about under the hood, I’ve made a ZIP package that has a before & after Maya file for you to peek at. It was made in Maya 7.
Anyhoo, enjoy the obscure control rig stuff. And once again my apologies to the non-technical types who come here for animation stuff. I’ll get back to that pretty soon. But first I gotta put a ribbon on that APT Session #2 student highlight reel. Whee!

Friday, May 04, 2007

Eye Rigging R&D 2: Cluster/Joint arrays for lid deformation

OK, more geekery with apologies for the non-CG folks who stop by to read. Some folks asked for a further explanation of the cluster array I described in my previous post about eye rigs. I mentioned that I use a fanned out array of clusters to bend and shape the lid to plus the emotion of the puppet. Even though I used clusters you could do the same thing with joints, and to illustrate the concept joints are actually easier to use. So follow along and see if this makes sense. (click the images to see & read things better)
array_01.jpg
array_02.jpg
array_03.jpg
I hope that makes sense and fills in some blanks. Rigging is extremely difficult to be brief with because it is so technical. I hope to do more of these little quick hit rigging R&D posts as time goes by. The hard part is finding the time- as always. heh.
Now I gotta get back to other stuff I’m falling behind on.