Real Old School McDonald’s Apple Pies

Ok, those of you who are generation-X or older certainly recall the glorious, lava-like goodness of the old, original McDonald’s apple pie. Deep fried and crispy with a super sweet, magma hot filling- these things were awesome. When I was a kid we didn’t eat at McD’s often, but when we did 9about once a year) I always got me the lava-pie. It was my right- nay, DUTY.

Lame-O New “baked” Apple Pies. Yuck!

But somewhere around the late 80’s or early 90’s McDonalds caved in to the food-nannies and ditched the lava-pies. Maybe they were afraid of the lawsuits from someone burning the roof of their mouth with the apple filling (duh! That was half the fun!!). Ostensibly they caved in to the cries about how unhealthy the fried pies were, opting instead for something ‘healthier’. Puh-leese! It’s a friggin’ McDonald’s Apple Pie, not a brussel sprout casserole. We knew what we were eating, we knew it shortened our life span by weeks -but we enjoyed it! It was a treat that made the living of that life span something to look forward to. What’s the point of living longer if your food tastes like crap? But no. In the good ol’ US of A the old school McD’s apple pie died an ignomious death, killed by the do-gooders who think they know what’s better for us than we do. Of course they rolled out some new apple pie- a nasty dry, cakey, styrofoamy “baked” applie pie with less than half the sugar and waaaaay too much cinnamon. Like we wouldn’t notice the difference. Eating my first one I nearly cried. Horrid. The switch happened in my early adult years and marked the passing of my childhood. Truly I was a man now and I could never go back, the magic of my youth sent away like a summer sunset.
But lo and behold, what do we have here in Brazil?

Awesome amazing old school lava-pies in Brazil!

You guessed it. The food mamby pamby nannies here aren’t vocal enough to shout down the classic fried lava pie. I had a bite of one today. Oh my… it was like getting zoomed back in a time machine to my youth. Perfect. Absolutely, gloriously sweet and crispy. And of course you know I burned the roof of my mouth.

I miss many things from home, but here in my new home I have something even better. The ability to trundle back in time to enjoy a treat from my childhood. How cool is that?

Chicken Little Cartoony Animation- Oh yeah!

This is fun stuff.

QT 6 version here.

I was watching CL again recently when I saw this “throw away” scene and just had to stop the DVD and go over it again. Check out these nice, clean, easy to read poses…

click each to enlarge

#1, 2 and 4 are nice, but not out of the ordinary. But pose #3 is great. This doesn’t happen much in CG- the floating anti gravity character. And look, he’s not standing straight up. Very cool. It’s an awesome old school cartoon thing to do, but has been rare so far in CG work. Even in this film it’s not a common thing to see.

Even so, the real genius of this scene isn’t the poses as much as the breakdowns and inbetweens. Check these transitions. Starting from the left and going to the right…

click each to enlarge

That’s it- 3 inbetween frames for the whole move. And it works because the animator is thinking in terms of moving shapes and color. Now check out this transition…

click each to enlarge

Niiiiiiice! See those 4th and 5th frames? That rig is so busted you can’t even tell what this character is. But that’s the beauty of it- you don’t need to. The shapes in motion play together perfectly. It’s all artistic, it’s all about shapes in movement. It’s just great old school, classic genius. It’s SO great to see this in CG. So whoever animated this scene, great job! You are now the author of one of my favorite CG animated scenes ever. That and $3.50 can get you a coffee at Starbucks. heh.

The real pity? One of the best animated scenes in the whole film is a throw away piece of work stuck inside a TV set that another character is watching. I can’t wait until we see this style boldly attempted across a whole film. This is what I think is needed for the future of CG hand keyed animation. You can’t get this kind of motion (and emotion) any other way but to animate it. And you can’t animate this by faking your way through it. You gotta be an honest-to-goodness-you-know-what-the-heck-you’re-doing animator. Sounds like a great challenge, no?

My Crystal Ball: Motion Editing

(image courtesy of Justin Barrett)

In CG there has always been a kind of short tether to the pursuit of utmost realism. The very foundations of the medium were formulated in recreating a versimilitude to what you can see out your window. Or on your desk. Or under your bed. For 20 years the driving force behind most technological improvements in CG has been the mandate to recreate reality in ever more convincing ways. Each year there were technologies developed to help CG stuff look “better”- ie: more realistic/convincing. Once the technology arrived for making something look “real” it was a short step to create simple to use tools to tweak the data to represent stylizations of realism. It’s not a heretical statement to say, based on a cursory observation of many CG feature films, that “real” and “stylized real” are at the top of the CG foodchain. Pixar coined a term to describe this a few years back- hyper real. This has been the hallmark of their (very profitable) house style since the very beginning.
I think that the following premise is accurate: If there has been an obstacle to achieving realism in CG then any meaningful success over that obstacle has been via a technicological accomplishment. Once that technology was achieved then the stylization of that easily obtained realism was more of an act of editing than raw creation. If you need an example think of the modern “texture artist” in a CG studio. How much are they creating original art versus editing digital photographs to be pasted onto objects? This isn’t an argument of “artistic vs non artistic”. It still requires that the editor have an artistic eye. I’m just trying to define the tasks and skills involved in an objective manner.
So looking at the current trend of character animation in CG, where the top of the foodchain is a kind of “live action lite” realistic-ish motion (but slightly stylized), why do we suppose the end game will be any different? Even the ’stylized’ motion in ‘cartoon’ CG films is still grounded far more in the physics of realism than fancy. And in an industry trend that some thought would never happen 8 or 9 years ago when motion capture was in its clumsy, ugly infancy, we’re seeing more and more animators whose primary task is manipulating motion capture data to match the director’s vision. Increasingly the bulk of this work is artistically editing or adding onto existing data, not necessarily creating new data. And they’re not sticking with doing this for just human or human-like characters anymore, either. The recent hit film Happy Feet mocapped humans while animators and technologists edited the data to work for penguins. Right now that editing task is still a bit primitive, so the need for animators to sometimes heavily edit or even create new data still exists in significant quantity. But the technology for editing this data is being worked on by very bright minds. As with all technology it improves and becomes cheaper. The day will come when they will succeed and you will have the ability to quickly and cheaply do all kinds of ‘artistic’ things with motion data that right now requires the hands of an animator to accomplish. Knowing what we know about the history of CG I say that the technology for capturing and digitally manipulating (ie: stylizing) realistic motion will only improve and the need for ‘animators’ to stylize the motion will decrease. Instead you will have people using cheap, fast and friendly technology to quickly alter motion data resulting in stylized digital motion based on the original. For example: What we can do with photographs now in Photoshop will be just as easy to do with motion in not too many years. The difference in how you spend your days is simple. Are you the one creating the motion/performance, or editing it?

Now mind you, I’m not saying this is the end of the world and that it’s the work of the devil or whatever. There are clearly very talented and artistic editors of all kinds in the world today and we all enjoy their work. But editing is different task and skill than raw creation. So in future years we might need to think about what it means to be called an “animator”. In the old tradition “animator” is a term that was hard earned and not lightly esteemed. It conferred a high level of accomplishment in the art of creating performances with nothing but your talent, skill and work- and some basic materials at hand. It was clear that if you used film footage as the basis of your performance (rotoscope) that you were an inferior animator- if one at all. But the definition of the term “animator” is getting kinda blurred. If the future technologies play out as I suspect they will, can we still claim to be “animators” if we don’t actually animate? If the bulk of the day is spent editing and stylizing motion data to craft a performance then perhaps a different title is more appropriate. I think we may need to create the position of “motion designer” to better describe what happens on data manipulation projects. It’s a different skillset. To me the difference is like that between a jazz musician and a dance DJ. One creates music from scratch using only the instrument, their skill and talent. The other makes new combinations of music by mixing and matching existing music data created by others. Both can sound good in their own way, but you would be hard pressed to find anyone willing to call a DJ a masterful musician in the classical sense.

Now here’s the heretical part that will surely get me in hot water: How long until you can make most (not all of course, but a majority) of the character performances in a film like The Incredibles with such advanced motion editing technology? If you watch the performances of the characters in a film like Monster House (created in large part by editing captured real motion) you might be tempted to think it won’t be long. And when that day comes do you think producers or directors will still hire animators to slowly and expensively ‘hand key’ it when they can hire them to just edit captured motion (directed in realtime with real actors and immediate retakes) with cheap, fast and friendly motion editing technology? Again, you might be tempted to think that they won’t. In the future this might add even more importance to the singular question: Why animate this?

Nouns, Inc.

Animator/director/stick figure genius Don Hertzfeld has a great way of describing the current CG conventional wisdom…

CG models and perfect life drawings leave me cold. All that a realistic, representational drawing of a bicycle tells me is, “bicycle.” There’s so much more mood and psychology to bring to the film if your artwork communicates more than just nouns. It’s why photorealism in animation is usually so boring and pointless. It’s all nouns. Ninety percent of CG animation is all nouns. I can’t feel anything going on behind the image.… If you’re going to strip animation of all its subjective power and just show me what things look like in real life you might as well be shooting live action.

You should read the whole article, though. I’m always inspired by Don’s success as an independent short film maker.

Sadly I don’t have much to show for it so far (for a variety of reasons- I won’t bore you with the details) but I’ve been wanting to do fewer “noun” projects and more “adjective” and “adverb” projects. Especially adverbs. I like adverbs. I have a few adverbs cooking in the lab right now, actually. I should probably start sharing their progress soon.

Principles of Lipsync Animation: Now in Chinese

A big word of thanks goes out to Li Leon for his great translation of my Principles of LipSync Animation article into chinese. (I think it’s a great translation. I can’t tell for certain since I don’t know a thing about the chinese language) So anyhow, now over 1 billion people can read about how to animate lipsync more effectively. Wow, talk about making an impact- over 1/6th of the world’s population can now read this article. My mom would be proud. Or not. Anyhow, go, read, enjoy- in a chinese sort of way!

Click tracks and bar sheets

Ask most CG animators today what a bar sheet is or what a click track is and they’ll look at you with that slight head tilt furrowed brow look that dogs use. “Click what? A bar whos-it?”. i admit that I used to be the same way.

If you don’t know what a bar sheet is (or if you had read about them but you don’t know how to use them) or what a click track means, then you need to go get you some learnin’! Where to go to school about this stuff? Well, you need to run- not walk- over to Hans Perk’s blog. Hans has a gold mine of old bar sheets from some classic Disney short films. For a while now he’s been posting the bar sheets and explaining about how to read them, what they mean and what they were used for. Lately he’s also taken to re-creating click tracks and superimposing the results onto some Disney short films from the old days. You definitely need to check it out. (Through The Mirror and the classic The Pointer are now both posted with superimposed click tracks).

His conversation kinda assumes you have some basic exposure to bar sheets, though. I’d wager that most CG animators don’t have a flippin’ clue what he’s talking about. We’ve never used them or even seen them- or perhaps even heard of them. After the jump I’ll give all of us See-Gee kids a little primer on what this is all about….
What’s a bar sheet? Well in the land before scrubbable audio in CG application timelines and realtime story reel editing with Avids or Final Cut, animation directors needed to pre-time their short films with the music composers for the films. And like all things pre-digital this was done on paper. The director for the animated film would have a sense of the pacing of the action in a film figured out as beats per frames. For instance, a bar sheet might mark a section of the film as “2×12″. The 2 is the number of major beats for the measure, the 12 is how many frames for each beat. Thus a beat every 12 frames is 2 beats per second (24fps remember). This comes out to 120 beats per minute- a system of timing that translates into music language and can be marked out on a metronome. Bar sheets were the master document regarding the tempo for a film- what it was, how it changed, where it changed. It would have rows for describing the animation, another for sound FX another for dialog and another for the music. The bar sheet would mark out seconds of film and the composer would even write down and number every bar/measure for the music. Timing and tempo changes were all done on the bar sheet, cuts were determined on the bar sheet- the whole blinking cartoon was figured out on the bar sheet. And they did this all before they animated a single scene. How freakin’ genius is that?! This would give a timing structure to the animators as well as the musical composer. The animators would try to have major pose hits or actions occur “on beat” knowing that the music- whatever it may be- will also be sharing the same beat and timing. Scene cuts would happen on the beat or just before it to let the first accent of a new scene fall on a beat. The connection and harmony between the visual action on screen and the beats of the music played for the film would be immediate and very strong. A click track is just an audio marker system that the composer would use to listen to the beats for a film as they watched the story reel. This way they could hear the timing as they watched the story reel and get a sense for what the music needed to do there. When you watch a golden age animated cartoon with a click track superimposed you can actually SEE the animation hitting the beats as you hear it in the click track. It’s a huge, huge educational experience. The sense of connection and structure to the film, the link between the animation, the editing and the music is so strong- it’s almost darn near perfect. I’ve worked in CG animation for 13+ years in every imaginable form- commercials, short films, direct to video, TV, feature films, game cinematics- you name it. I can tell you from experience that hardly anybody in professional CG animation today even thinks about this stuff. Maybe somewhere somebody does, but I’ve never seen it or heard of it and I’ve been exposed to a lot of different systems and projects.

Today timing & tempo is not generally applied in a universal fashion for a film. The realtime editing paradigm allows endless noodling of cuts and individual animators are given no guiding sense of musical timing for actions (unless it’s a specific “song & dance” sequence). But a musical tempo for all other scenes in the film? Doesn’t happen. Today music for animated films is usually treated as another post process, done after the final animation and layered on top like icing on a cake. The sense of connection isn’t as strong though and what you often get is this mish-mash of music and animation that rarely play off of each other.

Anyhow, go to Hans’ blog and go to school. You’ll probably need to go over the stuff a few times before you get your head around it. But when you do get your brain wrapped around what’s going on and you realize just how incredible it is you’ll sit back in your chair and realize just how little we really know about animation on this side of the digital divide. I for one love the rediscovery of this stuff and I can’t wait to apply these things to my next short film project.

The Manufactured Image: Ideas for Change (part 2)

cont'd from part 1...

- Get lead modelers to create second pass rough Cg sets, but do this while being informed by the camera and pencil layout data. So this is another place for technology to come into play- develop a system of exporting rough camera and pencil/color layout info for import into any asset at a later time. Get this working so that camera data can begin to be stored and shared as early as layout or concept development for a set. If you’re smart you can save your story artist exploratory camera data as well. The sooner you can capture camera data for use downstream the better.

- Following pencil layout (concurrent with color tonal layout) put your modeling team to work on individual asset creation. Approve the individual assets in context based on looking at them from the various camera angles exported from previous steps. Hold off on materials and textures, though. We’ll get there in a moment.

- Master lighting for sets could be a two stage process. Stage one would be rough master lighting- ie: before textures and materials are applied to the objects in the modeled set. Yes, light the set without any textures at all. I used to do this a long time ago when I was doing Cg illustration because I found it easier to see the macro level art in the image without all the details cluttering things up. The idea is to shift away from a micro focus (textures, materials on individual assets) and onto a macro focus. Paint with light to build the environment. As with modeling, do this as informed by published rough camera data and in scene color key projection maps created by the colorists. Now you use broad strokes of color and tone to build your compositions. Materials and textures still wait.

* A word on the camera data to this point: It’s all rough. Yes, it will get tightened down & changed a bit later. Yes a fair amount of work will be thrown out and re-done. This is no different than the current system. But a rough idea of how something will look is better than no idea at all.

- After rough master lighting, NOW you can make informed material and texture callouts for the detail tasks. The individual texture artist now has not just concept drawings, but camera specific tonal, color and lighting cues to work from. They can better accomplish their tasks while working in concert with the larger macro artistic focus instead of getting lost down the rabbit hole of pretty details in isolation. Again, approvals from real camera views, not isolated turn table renders.

- Give supervisors provisional approval power to get assets out of the isolated setting and put into the set in context. Basically never allow an art director or director to look at a set dressing asset all by itself in a grey background render. This is a recipe for noodling. Final approvals should be in context only (as much as humanly possible). For approvals the art director and director can see how these things all play together and in the dailies theater by having the editor toggle from pencil layout projection pass to colorist color/tone projection pass to no materials rough master lighting pass to detailed asset pass -all in moments. All notes for revision to individual assets are given in light of the scene/camera view context. Look at the set with sub assets from as many camera views as you can for a set. (remembering that we already have rough master lighting here). I guarantee the results will be better. Producers will love me because this one thing alone will cut the revision budget for materials and textures by a third- at least.

- Second pass master lighting can be done to tighten the integration of details and macro art elements after all the assets for a set are textured & approved in the above step. All along this process the camera data is updated and made stronger via layout, animation, etc. As camera data gets significantly adjusted new rough pencil layout & colorist projections can be made if necessary. However somewhere along this stage of the game (or slightly before) the pencil stuff fades away and is no longer updated because the CG imagery is starting to take stronger shape.

- Using “same as” camera information, pre-render your background plates for scenes that do not require character/set interaction. Build a library of rendered backgrounds for scenes. Then, to better the compositon, use colorists to punch up the images digitally. Push areas darker, pull saturation more in other areas. You can even digitally paint out things that make the composition klunky. What would take hours to do in test renders and lighting adjustments can be done in minutes in a digital paint program. This is an example of mid-pipeline compositing that could work.

- To ease the crunch on scene lighting, get scene lighting happpening sooner. Don’t wait for all the textures and details to be done to begin scene specific lighting. Make scene lighting a two stage process like master lighting- first stage sans texture details, second pass after details are done. Do first pass rough scene lighting for each scene based on the rough master set lighting. Scene lighting can work in parallel with animation, immediately following Cg layout. Most studios have a technological system to allow scene lighting to occur in parallel to animation. Rather than building a tsunami of inventory for lighters, work down the task load in phases. Rough scene lighting (sans materials and textures) will again give scene lighters the ability to macro focus on the artistry of their scenes and use the effective “big brush” of light color to build better compositions.- Second pass scene lighting occurs after all assets for a scene are detailed and textured. But since the broadstroke lighting for the scene has been handled in rough scene lighting second pass scene lighting can go faster to tighten down the details and how things integrate, making life more humane for lighters at the end of the show.

- Create tools that allow scene lighters to macro adjust base colors for assets in a specific scene. Got a scene that needs everything to be a cool green to the left, but all the assets are textured with warmer colors? Or is your character feeling too much like the background and your lighting setup is only getting you so far in helping them stand out? Allow lighters the ability to marquee select a section of the screen (or specific objects) and hook up a master HSV modifier into all the materials & textures for those objects- including characters and props. To go one better make it have depth capabilities as well (ie: stronger adjustment power further back in z space, etc.). Go even one better-er, give it basic masking functions like feather, radial or linear gradients. Make a similar tool to macro control specular highlights or bump map depth or texture detail contrast & brightness. Basically allow the lighters to have more tools to “paint” the scene than just lights and give them access to make gross adjustments to object materials on a macro level. It’s not about them sending things back for fixes, nor is it about opening up every texture or material detail for a lighter to re-interpret. It’s about giving them simple, easy to use macro tools to quickly make compositional improvements right at their workstations without creating an avalanche of re-do orders. Make your lighters your army of colorists. Make the scene specific results publishable so they can be propagated across all same-as and similar-as scenes. Production managers can thank me later.

-Why not use late stage compositing to fix problems? You can, but compositing by nature tends to need many layers and elements in final-ish render form to be of best use. A good compositor can rescue almost anything. Problem is by the end of the pipeline they rarely have the time (or disk space for all the elements). Do rough compositing along the way? Now you’re fighting for hardware assets like network bandwidth and storage capacity as you render rough layers for many, many stages of the show. The data storage and tracking tasks become unwieldy. Compositing is used extensively for lower volume projects like commericals or VFX jobs. A commercial typically will have fewer than two dozen scenes total, so compositing makes economic sense. And a big VFX job is 250+ scenes. A monster VFX job is anything over 500 scenes. But a typical 95 minute Cg animated feature film is well north of 1,200 scenes. With each 2k film image taking up around 7mb of space on disk, every comp element significantly ups the needed size of your server space. The economies of scale work in reverse here. In other words, there is a reason why for years Pixar rendered everything “in frame”. Massive use of anything more than the most basic of scene compositing on a feature film scale is murder on your infrastructure budget.

Anyhow, those are some ideas. How many will work? I don’t know. I do know a number of them will work and are very do-able. These certainly aren’t the only ideas or solutions possible. They’re just the ones I have in my head or have tried. Several years back when I was installed as CG Supervisor for Big Idea’s second feature film project (which never got made, sadly- the studio went bankrupt before we could get very far into production) we tried many different things to get the films looking better and make the pipeline more artist friendly. Other ideas have come to me in the years since. I’ve listed most of them here in the hope that other bright people will use them as jumping off points for even better ideas. Others will no doubt find it easy to poke holes in the ideas as well. All the same, I’m just some dude who’s willing to stick my neck out and say the current system isn’t all that hot and here’s how I think it can be better. That’s a conceit that assumes a lot of responsibility, and I hope I’ve handled it well. And if I’m wrong, well… nobody can fault me for lack of trying.

Thanks to Mark Mayerson, Amid Amidi and all the others who have taken the time to link back to the post series over the last few weeks. Those of you new to my humble little corner, thanks for coming! I hope you found something useful in all of this rambling. Hopefully we can see the artistry in feature CG films pushed in ways that challenge technological paradigms. I’m eager to see things yet unseen.

The Manufactured Image: Ideas for Change (part 1)

Sorry for the delay in getting this last installment issued in this series of posts. I never intended for this topic to grow like it did, but I kept finding thoughts and ideas that needed to be expressed.

Anyhow, this is where I stop saying “It’s broken” and start saying “Maybe this could help.” But before I get into specifics, I think it’s important to clarify my thoughts on pipelines. Many assume that because CG films are made on computers that CG production pipelines are technical in nature. In fact they are not. They are rather creatures of task organization, asset delivery and revision/approval control. In other words, pipelines don’t get improved so much by technology or new PERL scripts or programs. These things are tools used in applying a fix, but the fix does not come from the tool. Pipelines are fixed on paper first, in how you think about and organize your task and approval systems. It’s more about the bigger picture and less about what tool you develop to shuttle and track data from point A to point Z. And how you think about your primary goal in the system determines much about how you build that system. So how can we bring more of an integrated sense of artistry into the CG film production system (aside from having good directors and art directors)? By adjusting the arrangement of asset creation tasks to better accomplish the simple goal of making better art in our films. More after the jump….

I’m a big believer in applying the right kind of solution to the right kind of problem. If your problem is isolated task management, a lack of macro artistic excellence and a parts inspector approval paradigm, then a data tracking software system isn’t gonna be enough. Nor will an asset publishing regimen meet the need. While these things are valuable, the truth is that software solutions are not functional remedies for managerial shortcomings. That calls for someone with skills that range more in line with macro problem solving, multi-faceted organizational skills and, inevitably, a negotiator’s knack for brokering deals between various combating task groups (ie: departments). Absolutely technology is a necessary component of any solution, but in itself no technology is THE solution. Just as a wrench in not the solution to what ails my car, but rather it’s the application of that wrench that is the true secret to success. So my suggestions for improvement will carry this idealogy- it’s not the technology that needs fixing, it’s how we think about the challenges.

Now, here are some ideas, in no particular order, about how to adjust the current Cg large budget paradigm to allow for better integrated artistry in the images on screen.

- Thoroughly workbook every scene from both a technical AND artistic point of view. Vital.

- Very early on have your director, production designer/art director & DP think through the working script to define key scenes that they can already ’see’ in their minds. Rough sketch those ideas to create the beginning of a scene catalog. Create rough color keys for these scenes. It will change later, but the idea is to get key artistic people thinking whole imagery over assets as early and often as possible.

- Hire experienced DP’s from live action to tighten up the use of cameras and scene blocking. CG is lazy and uses too many unique camera set ups simply because they can do so with no added cost. For instance: One film I worked on had a sequence that was 67 shots long. If it was a live action shoot it would have been cut down into 4 blocking scenarios (for 4 different takes- usually more than enough for a director to get his coverage for editing) and a total of 17 unique cameras. By live action standards even this amount of condensing would be considered a ’sloppy’ blocking solution to the scene. With 17 cameras coverign 67 shots that’s something a line producer will try and get reduced to keep the budget in line. But the sequence in the CG film used 46 unique cameras for those 67 shots- all arranged in a way that couldn’t be described as any kind of cohesive blocking set up. This makes the task of creating artistic images that much more difficult because the workload is too high. Fewer cameras = greater ability to focus on each composition. Fewer blocking scenarios = greater ability to build a set to best match the views used. Not to mention it makes a film feel more professional and tight.

- Get your rough CG sets built as soon as possible and with absolute minimal detail. At Big Idea we taught our concept artists (yeah, the blue pencil guys) how to build rough sets using Maya. Very simple stuff- cubes, planes, cylinders. They developed the basic structure of their sets as they drew them. No details, really. Save the details for the drawings. Just get the set existing so key creative people can start to see it and work in it. This is very early in pre-production we’re talking. But we blurred the line between pencil and Cg and as a result we had rough (and I mean rough) Cg sets to start working from as early as storyboarding.

- Board from the rough sets. Create tools so story artists can take quick snapshots of camera angles they like. If things change, things change. The sets are cheap enough at this level. But boarding and story blocking from the sets allowed the story artists to fly around and find where the action could work, get interesting staging angles- all informed by the Cg asset, not in a vaccuum. (and save the camera positions for later use in rough blocking). And if the set didn’t work for the story we changed the rough set- it was easy at this stage. And cheap. We also had the DP look them over along with the head of story. The basic idea here is to get some kind of representation of the Cg set in front of artists eyes (especially the artists responsible for macro artistic decision making) as soon as possible.

- Use colorists for more than thumbnail color scripts years before production begins. Hire and use colorists in actual production. How? Read on.

- After approved rough CG layout (right after story reel is approved for a scene) have colorists create tonal and color keys for every unique camera set up. If your DP is good this won’t cost too much because the camera set ups will be smarter. (see above). The colorist can do this ‘offline’ by painting over a screen grab of the rough Cg layout and putting the image into the scene packet for later handoffs. Or better yet they can do it “in scene” by making use of camera based projections. Basically project their tonal and color paintings right onto the rough geo for the rough CG set from the camera view. And here is the perfect place where you would use technology to help the larger solution. A programmer can develop an easy to use tool for the colorist to accomplish this task. Now when you look at the openGL render you can see the tonal and color right in the camera. Get it in there and let that ride right thru the rest of the pipeline by attaching this info to the scene camera so it gets referenced into every variation of the scene no matter the task or time. This way all downstream artists (both asset and footage) can see the color and tonal keys for the scene and make informed artistic decisions about their individual tasks.

- Expand on the above idea by having a skilled traditional layout artist (yes, a pencil guy) draw up a very quick rough layout sketch for the scene to play out in. Do this before the colorist gets it (ie: before you do what I just described in the previous item). Here they can start to develop the details of the set in inexpensive pencil form. They would take their cues from the approved concept sketches for the set as well as the rough Cg sets they see in camera. With art director and director feedback they can tighten down many of the specifics of the scene layout. And camera adjustments can be made here as well to accomodate different scenes working together. This helps the set evolve from the camera/scene point of view rather than the ‘bird’s eye” concept art overview. Anyhow, project this pencil data onto the scene in CG. Then when the colorist gets the scene they have details of form to work with. Informed compositional data flowing down the pipeline? Revolutionary! heh.
Overall use camera projections more. It’s an old technology but it works great.

more in part 2...

The Manufactured Image: Color, Tone & Texture

As we have seen in plenty of examples the deft use of color and detail are vital to the creation of an integrated artistic image. Handle them poorly and you are playing from a handicap from the start.

The task and asset tree in hand drawn animation is a much simpler system (generally) than CG. The entire image is thought out in workbook. Workbook is where you go from an approved storyboard for a scene and translate that into all the things needed to actually produce that scene for the film. This usually includes a rough color version of the scene. Along with this art comes call outs for specific assets and tasks. Workbook is basically the bible for the scene. It defines everything from color to layout to animation to FX to whatever (the art side) and every asset is identified, given a name and a task (the production management side). So from the workbook stage you would have the callouts for these basic assets:

Background line (traditional or CG. hand drawn made by one artist, CG by modelers & layout staff)
Animation line (anything that moves. Multiple artists create the line based on character, task, etc.)
Foreground line (traditional or CG)
Background color (painting or cel)
Animation color (cel)
Foreground Color (painting, more commonly cel)

More after the jump…

Assets like animation line are usually handled by a larger number of artists, but the end deliverable (from the standpoint of color/detail/etc/) is unified and simplistic- namely line drawings. The background, foreground and animation line finals are put together into a basic line composite. This is often what you see in ‘behind the scenes’ features on DVDs. Here’s a small example from Beauty & the Beast.

(click to enlarge)

Increasingly the backgrounds are being made in CG with models and 3d cameras. This has it’s own structure and task heirarchy, but the asset focus again is on line only. The Cg models are used to create line drawings without color usually. With cel shading being more common now this may change, but not many folks are making hand drawn big budget animation anymore. Thus the evolution of the process of integrating CG assets into traditional pipelines has kinda frozen for the time being.

Anyhow, at this line composite stage what you usually have is all lines, no colors. A background painting may already be in progress to define the background color. The look of this would be determined by the workbook color key. Anyhow, now a colorist goes back to the workbook and figures out the colors for the scene- specifically the cel colors for moving items (animation) and foreground elements. They’ll do this with an eye toward the rough color key for the scene that was made for the workbook. Once the colorist has defined the final colors for the whole image then new color tasks are assigned. The background is painted by one artist. The animation line is colored by another and the foreground by another. Animation line may be made up of several different line assets (per character, etc.) and thus the coloring tasks may fall along individual assets. (one artist colors Gaston’s line, another colors Belle’s line, etc.) But the color callouts for the whole scene all come back to one place- the colorist’s color call outs based on the workbook. The whole thing never strays very far away from the workbook and it is usually worked in context. The pool of tasks is fairly shallow and perhaps most importantly- individual asset color, detail and texture are defined generally once the entire animated scene is assembled in line form. Thus color is finalized with almost all of the information about staging already defined. Color, detail and texture can now be created to serve the overall staging of the scene. It’s all right there with no guesswork involved. In fact for most of the assets there is really no other way to do it until after the line has been assembled. So it only makes sense that the result of this system would be more cohesive (allowing that the artists involved are good at their jobs). So color, detail and texture (including “lighting” even thought lighting is just another form of color control) are all handled within a short section of the overall asset & task pipeline. First they are defined as a whole in workbook and then, after line assets have been assembled, colored as a whole to match the defined layout and animation. This is of course a somewhat simplified view of the system, but it is generally on target for most typical scenes.

CG has a much more wide spread asset and task structure. The whole concept of workbooking in CG films is still not well defined. Some do it, others don’t although by now most have some form of workbook in their systems. Some who do workbook don’t bother much with color and use workbook only as a way of translating 2d story images into CG camera data. Assuming a Cg film finds it’s way around the initial workbook challenge and they manage to define color, staging, assets, etc., the tasks only get more widely flung.

Currently most CG sets are made up of many sub assets. To meet scheduling demands it is difficult to allow one artist to build an entire set. It often would take too long. So the set is broken down into the parts that make it up. A chair, a desk, a sofa, a room, decorations, etc. By breaking up the task load you can have many assets in development in parallel. So now instead of having a single layout artist create the line drawing for the living room, you have 10 modelers creating models of pieces in the room. Then you have 10 material/shading artists creating the materials. And then 10 texture painting artists creating the textures. Each sub asset comes with it’s own concept art for shape, form, material, texture and color. And each stage for each sub asset is approved or revised in isolation. It’s very rare to review the textures of one asset (the table) while seeing it next to another asset (the sofa). And even more rare to review and approve them all as a whole in the context of the entire set. And it pretty much never happens within the context of any camera data at all. So you see the set colors, materials, textures and details all in pieces- and then never through an actual camera from the show. They’re designed as a whole, but approved and reviewed in isolation. It’s like looking at a jumbled picture of of someone’s body parts instead of a picture of the person. It’s all there, but it’s all broken up.
So after all the individual assets are modeled, have materials and textures applied, and after they’re all assembled into one file you have your background set. But it’s loaded with all kinds of different color and textures. All based off the original set design, but still done in pieces. It doesn’t have any lights added yet, though. So now it goes to a lighter. This person will create the master lighting recipe for the set- a lighting solution that looks good from most angles within the room. If they’re lucky they have some rough camera data to work from based on all the scenes that use this set. Most aren’t lucky. Once it arrives at the master lighting artist’s desk the materials and textures are pretty much set. So they go from design drawings for the set (outside the context of any one scene) to try and make it look good overall. Again- none of this is happening in the context of any one scene at all. The whole color, texture and detail development part of the pipeline is happening ‘offline’. As I have said before, this is no small distinction. This is only the background set. The characters are undergoing their own similar chain of development, as well as all the props and moving set pieces, etc.

The first time you see the animated characters, the background set, the props and all the pieces put together is usually in scene lighting- towards the end of the pipeline. By the time you get here you find that most of the color system is set- practically in stone. Contrast this to the traditional paradigm where at the same stage of assembly very little of the color and detail system is set. For typical Cg productions at this stage the character’s colors and materials have been approved months ago. The colors and textures of each piece that makes up the background set were approved long ago. The master lighting model for the scene has been approved a long time ago. The basic cameras and staging were approved ages ago (not counting minor tweaks in post animation layout). Currently is it only common to have animation working in parallel with scene lighting. So the very first artist to see everything all together is looking at months and months of approved colors, materials, textures, layout staging and base lighting. They’re also looking at most of the schedule and budget already having been spent. They cannot send anything back except in the most dire of occasions. And even then they need to sell it up the chain of command. Send backs at this stage are expensive. Never mind the time element involved because by the time a scene gets to scene lighting most of the schedule is gone and a production is often crunching. The end result is this- the first artist to see and have the whole scene assembled before them has the least amount of authority, time and budget to make it work as a whole.

So compare the two systems. One has fewer artists involved in creating color choices and most of the color choices occur pretty much in concert and at approximately the same time. The other has a wide array of artists involved where color & detail decisions are isolated, incremental and separated by significant gaps of time. All things being equal- great individual artists, equal budgets, etc. which one do you think will yield more cohesive and integrated results artistically?

Once again I stress that these are simplistic generalities, but they are not wholly inaccurate ones. What can be done to fix this? Well, there are no simple answers. You cannot ignore the technological, production and scheduling realities that drive so much of the CG feature film economy. You can’t make the Cg system just be the same as the hand drawn one- it’s not that easy and it won’t work. But you can rethink how you do some things, shift some things around and try things in different ways. And you can use your strengths (ie: technological solutions) to address your weaknesses. In the final installment of this series I propose some ideas that have been knocking around in my head- some experimental, some not so much. Some won’t work, others might. We’ll see.

The Manufactured Image: Tasks & Assets

Lots of fun discussion in the comments section of previous posts. That’s good to see. And thanks to Amid at Cartoon Brew for sending folks this way for the party.

So why do I think that traditionally animated film pipelines have some kind of inherent advantage over CG pipelines when it comes to making integrated artistic imagery? Am I just some disgruntled old 2d animator tipping at windmills? No, I’ve been a CG animator from the start. I love CG. It’s my home turf. So why do I think typical Cg pipelines are dealing from a short deck? For me it comes down to how you look at your tasks & assets. What do I mean by tasks & assets?

Assets are the things you see on screen. Characters, props, furniture, vehicles, special effects, rooms, environments, etc. With no assets you have nothing to look at. In order to have each of these assets on screen you must assign a task in creating it. So in Toy Story, in the scene where Buzz Lightyear is looking at the moon delivering his ‘empreor zurg’ speech, Buzz is an asset. So is the moon, the gas truck, the petrol station and the sky. And in CG you need Buzz’s performance, which is a separate asset than Buzz himself. In the Iron Giant the scene where the squirrel runs up Dean’s pants, Dean’s animation is an asset, as is the squirrel’s animation, the newspaper, coffee cup, coffee, diner and background characters with all of their props. So an asset is merely a thing that needs to be created to show up on screen. How you make these things (ie: the technological nature of your tasks) determines how you build your pipeline to create your assets. How you create your assets in large part determines how they can be used.

In a typical traditionally animated film the focus of the asset production is scene based. Assets are created specifically for a given scene. Occasionally you can re-use a background or a bit of animation in another scene, but the overwhelming majority of assets are created specifically for a particular scene. The defining goals for the asset are scene specific. From the moment the scene is conceived until it is shot onto film every asset in that scene was crafted specifically for that scene and no other. This is not a minor thing. The character of Hogarth does not exist outside of any scene in TIG. There is no Hogarth in a file folder. No Hogarth on a shelf. Hogarth does not exist outside of the task of drawing & coloring him for a specific scene. In reality there are hundreds of Hogarth assets in The Iron Giant -one for every scene he appears in. This is a productivity downside due to the loss of the economy of scale in re-usability, etc. But there is an upside as well- the ability to completely customize the asset for the specific scene without having any limitations due to other scene needs. The Hogarth in scene 200 is created (drawn and colored) from scratch. The scene 200 Hogarth animation asset is built only with scene 200’s needs in mind. There is zero separation between bringing the asset into existence and fulfilling it’s core purpose for the scene. This is a major, major advantage for creating integrated artistic imagery.

The same applies to backgrounds. The nature of a hand drawn background requires that it be camera specific. Only in the rarest of instances can you use a single background asset for every scene in a sequence of film. Each scene pretty much requires that you draw and paint a new background asset to accomodate the new camera view. Again, loss of economies of scale, but more freedom to custom tailor the background to best meet the artistic needs of the scene. By and large the creation of a background asset for a traditionally animated film is driven by the needs of the one scene for which is it built- again with a zero degree separation between the asset’s creation and it’s scene specific goals. Now when you assemble the assets they seem like they were made for each other simply because they were made for each other.

CG in contrast has a bigger hurdle. In Toy Story you could not animate Buzz Lightyear under the truck until you had a Buzz Lightyear puppet to move about. You could not render that Buzz lightyear until it had materials and textures and lights (all seperate tasks done by different people at vastly different points in time). So the Buzz puppet is created not for any one scene, but for every scene that Buzz will appear in. Ditto his colors and textures. Only his lights would seem scene specific, but even they are not. They are environment specific (asset driven, not scene driven). All scene lighting for all characters and props in the petrol station environment are derived not from the specific needs of the scene as much as from the petrol station asset’s master lighting recipe. The master lighting recipe for the petrol station occurs outside of the needs for any one scene but is built to be a baseline for all scenes in a sequence. There is only one petrol station asset and it is used in all and it rules all. The petrol station is created before you even know what every camera angle will be (and thus what it’s bgrd needs to be) for a sequence. In Cg there is by nature a gap between the task of building an asset and it’s application for the needs of a specific scene. This again is no small matter.

Cg assets are built to meet the overall needs of many scenes, not one. So the upside is the economy of scale. The downside is one size fits all. As noted, exceptions exist, but generalities are the topic here. No one general asset can be great for all shots. It will be better for some scenes, worse for others, depending upon a wide array of variables. Conversely specific assets for specific scenes are great for only one scene and will be of practically no use for any others. The difference is in the value of customization vs. generalization. A suit off the rack at JC Penny will be OK for most fellows of a certain size and fine for your garden variety family wedding. But a custom tailored suit built just for you (which would be an ill fitting annoyance to anybody else) is just what the doctor ordered for life’s special occasions.
The difference should be immediately apparent to those keeping score.

But this doesn’t explain everything. As noted earlier, the giant in TIG is CG. Increasingly many assets in traditionally animated films are CG assets. So while the scene specific focus of asset creation goes a long way toward showing the inherent advantages for creating integrated artistic imagery, it doesn’t explain how re-usable general CG assets have been successfully integrated into predominantly hand drawn animated films. For this I think we need to examine one specific area of the process: The tone, color and texture process.

Next post: Where is the decision made about an asset’s color and detail? What affect does this have on creating integrated artistic imagery in animated film?

I like Mark’s take…

There’s been a fun thread of conversation on various sites and blogs and such about the benefits, dangers or value of pitching to TV networks. I won’t rehash it. For the latest on the conversation you can stop in at the inimitable Cartoon Brew. On his blog Mark Mayerson chimes in with his experiences with his TV show Monster by Mistake, explaining how that experience soured him on the value of even bothering with TV networks, etc. He closes his post with this great statement..

There’s no story and no character that hasn’t been done before. There’s only your point of view to differentiate your work from everybody else’s. If that point of view is stopped or twisted before it reaches the audience, the essence of your work has been destroyed. My experience tells me that it’s unlikely to survive in the TV industry and if there are alternatives that will protect your point of view, you should seriously consider them.

Amen and amen. It’s no special thing, but I too have had brushes with TV networks and development executives. In 2001 I was invited to pitch a TV show to Disney Channel based off of the characters for my short film “Lunch”. So I took a few days off of work and flew to New York City, hiked my ideas and myself up to the 5th floor of some office building on Park Ave. and then proceeded to try and pitch my idea. Shortly after giving the basic idea with artwork I found that I was basically wasting my time and had blown two vacation days on a goose hunt (but hey, it was educational so it’s not all bad). The basic gist of the meeting was this: they really didn’t want my idea in the first place. I mean, yeah, they saw Lunch, liked it and invited me to come talk to them. They made an appointment, welcomed me and let me present my idea, but they really didn’t care or listen to the idea. The formulas were set. The notes on artistic choices were set. Here’s a short summary of one such exchange that occurred in that hour.

DevExec: You’ll need to put clothes on the blue guy, Burbank (Disney HQ) won’t like this.

Me: Why? He has a turtleneck already. See?
DevExec: Yes but somebody might think he’s naked.

Me: It’s not like his junk is dangling there for all to see. Besides, he’s blue and this is a cartoon, right?
DevExec: Oh I think it’s fine, it’s just that Burbank won’t go for it.

And there is the rub- this isn’t about my idea or even what the audience likes or wants. It’s what “Burbank” will go for. They already had their formulas for what they wanted to do. They just wanted my ability to make something marginally interesting to look at. But from the very first words in response to my pitch it was nothing but change this, change that, drop this, do that. Did they want my idea or not? Like I said, I soon concluded that no, they didn’t. It was something else they wanted- a tableau to propagate their own ideas on what made for entertaining cartoons for kids. Hey, they pay the freight, it’s their network so they’re free to do whatever they like. Just as I was free to walk out of their office without agreeing to go deeper into the process where the ultimate prize is I get to be their monkey. Life’s too short for that kind of hassle. Since then I’ve been approached to pitch other shows or ideas. On one occasion I did follow through, but for the most part I politely decline. I’d rather make something that I enjoy than something that is merely a vehicle for some corporate media-employee to get a promotion. So I weigh in on the side of the “pitching to TV networks is a big waste of time and effort” in this debate.

The Manufactured Image: Something different (part 2)

cont'd from part 1...

As if the clear silhouette weren’t enough all of the vanishing points here draw you right in. Even with the crowd all a mish-mash of colors it still works.

Dark character on light bgrd, then light character on dark character. It seems so simple.

This one is particularly genius. The screen right waitress and mom are almost at the same exact place in the world. Same depth from camera, literally within 3 feet of each other in the same room and presumably under the same fluorescent light. But check out the shading on the other waitress. They made her dark to go over the light background, then put the lighter mom on the dark background, letting that high contrast pull you right where you need to go. Even with all the details, they’re pushed back using color harmony to pull the characters. And then you add the vanishing lines of the background push your eye screen left. Superb. If you lit this the ‘right’ way physically it would be a mess. But this thing totally violates the rule of consistent spatial lighting that throttles CG like a death grip - and it sings! Man that’s awesome.

Another amazing one. The entire background, all of the elements in it, are of one color theme. Characters, walls, windows, blinds, decorations- all of it a variant of that sea green hue. It pushes it all back, lets all the detail there fill in the space and keep the scene interesting without overpowering it. Anybody who has built, textured and lit an environment in CG knows that this is not typical at all. No way would anyone texture a complete portion of an asset (the diner set) to be ruled by one hue for one camera angle. And this is especially true for characters.

But not all is roses. It wasn’t as easy, but I did find some klunkers in the IG mix.

This one could have been something if they hadn’t shaded the background characters and trucks the same tone as Mansfield in the foreground.

But that’s all the klunky ones I could find online.

Some folks in recent comments have said that I’m being a bit too nitpicky about the layouts and compositional choices in my Accidents example. The argument is that the motion of animation is what carries the scene so it’s not fair to use image composition as a metric for artistic success. Yet TIG has motion- lots of it. It’s a high action film. Yet the artists who made that film didn’t shrug their shoulders at the value of integrated composition of motion and background.
So why does this stuff work better? Is it because it’s all hand drawn? Actually, it’s not. The giant and many mechnical elements in this film are CG. So something else is at work here. I do think it’s the pipeline. What about this production system made it easier to think in terms of complete imagery? All animated feature films need some kind of production structure of economy or else they’d never get done on time or on budget. Animated feature films need some kind of factory based pipeline. Pipelines are not all bad and we do not need to overthrow them in some kind of artistic bolshevik revolution. But some pipelines are better than others at doing certain things. The traditionally animated film pipeline (if one is to believe the visual product) seem to be geared towards allowing people to create more integrated artistic imagery. I believe that there are some things about the way the tasks are done- their order and the way they’re thought about- that hold back scenes artistically in typical manufacture/assemble Cg systems. No, we don’t need to throw the whole thing out. But like so many areas of Cg we could benefit from seeing how it was done before there ever was a thing called CG. We need to investigate and see how much of what we currently do is merely a result of the inertia of ignorance.
Next post- What are the key differences between Cg pipes and traditional pipes?

The Manufactured Image: Something different

Just for giggles I searched online for images of The Iron Giant. I was amazed at how consistently good they were. Image after image proved to be so well crafted I was really taken aback. Check these out….

Look at how the tones of the background characters is pushed back. The staging of the characters lead your eye right to the middle, especially by putting the taller characters bracketing Dean and then using those taller characters as mini backgrounds for the shorter Hogarth & mom.

Here we have tonal harmony without going monochromatic. The trees are pretty much just black silhouettes. The interior of the CG car is plain to allow the facial details to rule the scene.

Clear staging and simple color schemes with no textures? What a bargain.

The layout and background colors give a stage for the actors.

Crazy color schemes can work if used right.

I love the harmony in this one. Vanishing lines draw your eye, colors work well together, tone and contrast force a priority on the imagery and staging is superb.

Who says every scene in a forest has to be lost in a sea of leaves? The sweep of the treeline frames the character brilliantly. Clouds are pushed back into the sky color to give your eyes room to breathe.

Clear staging and breathing space can offset high details. Then use non-sensical lighting to frame the action.

more in part 2....