Monday, May 26, 2008

What is accurate?

these are just some of the differences between two DVDs of the same movie

I have always wanted to write about the use of color in One Hundred and One Dalmatians (1961). To make sure that I am analyzing the most accurate version I superficially compared the newly restored DVD (2008, PAL, RC2) to my old DVD from 2000 (the Regional Code 2 version of the 1999 NTSC transfer). Having compared earlier Platinum Editions with their older standard counterparts in the past I expected a lot of issues regarding color accuracy. Although the 2008 version not only looks cleaner but clearly superior, a few issues I’ve always had with the Disney approach of restoring pictures to “never-before-seen” beauty immediately came to my mind again. Since I didn’t want to incorporate these general thoughts about color accuracy and updates to historical works into my analysis, I will deal with it in a separate essay.

The Platinum DVD release has already triggered a lot of analysis of animation and storytelling by the likes of Hans Perk, Mark Mayerson and Michael Sporn who are far more qualified than I. As I certainly cannot provide any revealing inside information, my interest lies in analyzing what ended up on the screen and how that affects me as a spectator.

It certainly helps to have some background information so I hope to find out more about Dave Bossert’s and Walt Peregoy’s work.

I would be grateful if someone could provide me with historical facts that may clarify reasons for some of the incomprehensible differences between the two DVD versions. Please feel free to leave me some comments on this.

So before digging deeper into color schemes I will point out the main differences between the two releases in a few posts.

Note: I don’t consider the inferior DVD to be marginal because this was the only version of the film most people have seen within the last eight years.

Sunday, May 25, 2008

Final post on Horton

Part IV: cartoon poses, specific expressions and motion blur
As I wrote at the end of my first Horton post, the thing that intrigued me most about this otherwise uneven movie was the decision to treat the characters in a real cartoony way.

There are two informative AWN-articles on the technical challenges the makers were facing:

Cartoon poses
While the story modernization (adapting it for the global warming area) might not be that successful – after all there still is a vulture with a russian accent – the rigging development obviously pays off. The animals of Nool – especially Horton – move like classic cartoon characters and the animation of the Whos even hints at pre-classic rubber hose style which is very much in keeping with Dr. Seuss’ curvy drawing style. Of course some key elements of real rubber hose animation – the absence of joints and the possibility to elongate any limb to whatever length a gag demands – don’t work with rigging-heavy on-model-animation. It still is a long way for 3D to something like Clampett’s Baby Bottleneck (1946).

But all of the motion arcs are a lot more exaggerated and rounder than in average CG features. This simplicity leads to strong and organic animation. As a biped Horton even shows a pear shaped body like any 40s character. There are great poses that recall Chuck Jones’ pose to pose films. Fortunately it didn’t feel that zippy to me like in Dreamworks’ Madagascar (2005) for example.

One of the highlights to me is the handling of Horton’s ears in acting poses. They not only reflect the general idea/line-of-action of a pose (as in the middle picture on the suspension bridge), they actually serve as an acting device like gesturing hands.
They also take on the form of various hats throughout the movie.

Of course there are lots of stock expressions involved. Particularly the most basic “thinking gesture” (one eyebrow up, the other down) is massively overused although this could be one of the not-so-healthy Jim Carrey influences (Yes, he does it even in his live-action films).So I was quite positively surprised to also experience some fantastic cartoon acting. I remember the Councilman to have some very specific expressions. Of course clarity and readability are key factors of facial expressions but this does not mean that the audience is only able to understand the most basic emotions such as “happy”, “sad” or “angry”. While specific expressions tend to distort a face in unsettling way, the animators usually only seem to get away with complex facial expressions of minor characters (the shark in Finding Nemo (2003) for example). It came as a surprise to me then, that a character as important to identification as the Mayor of Whoville actually gets some fine acting scenes. Most of his scenes are pretty standard but some of them, like the ones down below really stand out. They are pulled off quite expertly.

Look at these faces. Aren’t they wacky?

Motion blur
One thing that really stroke me was the amount of motion blur applied to some of the wilder animations. What essentially simulates the way a fast moving object is recorded by a standard camera is used here in yet another way. In lesser animation it would be welcome as a device to disguise ugly inbetweens in blurry frames. But here it accidentally diminishes some funny poses.The apparently visible motion blur during zippy motions like Morton’s running around also help to highlight the organic arcs. The way Morton starts and stops reminds me of Chuck Jones’ Rikki-Tikki Tavi (1975).

By the way, look at the rubbery treatment of the arms that still look bony when hitting poses.
I assume that motion blur also serves as an unobtrusive way of bringing back the quality of dry-brush smears and “elongated inbetweens” (a Richard Williams term). When I look at these pictures I can’t help but think of what Michael Barrier calls Rod Scribner’s “Lichty style” (Hollywood Cartoons, page 436). At present motion blur might be the only way of achieving a looser off-model style of 3D-animation. So there are two solutions: It is only a matter of time until someone discovers a way of integrating real Lichty style poses into 3D. Or: this is finally recognised to be one of the crucial arguments to favour hand-drawn animation for certain stories…

Friday, May 23, 2008

Still more on Horton

Part III: A little more on color choices and stylization

From the establishing shot (with the directors’ credit overlayed) the already implicated cartoon-ness of the setting is largely expanded. As the sun is now illuminating a blue sky, high key lighting takes over. Mood lighting as done so masterfully in this beginning is reserved for special plot points such as the introduction of Vlad or the field of dandelions.

Both of these examples apply a very limited color palette as you can see in the screenshots. Both of them are expressively lit (Horton’s ears’ secondary action overshadowing his face for example). While Vlad’s cave is almost monochrome making dramatic use of cold muted blues and hard lighting, the magenta of the immense clover field is unified by salmon colored magic hour sunlight. Thus the sky is a balancing neutral beige that corresponds to the stone Horton is standing on. The tone in tone composition certainly helps to emphasize the vastness of the clover field. The areas of highest contrast as expected are Horton’s face, the falling flower and Horton’s backlit body.

Compositions in the open jungle in full daylight are not as impressive anymore. Still all the plants and rocks are shaped in organically curved forms and there is a limited range of colors such as green for the vegetation and blue and pale red for special plants and fruit. Unfortunately set ups like the one below (there are similar ones in the actual movie) tend to have a generic feel to it and the jungle vistas sometimes look more cluttered than well composed.

Horton himself gets introduced in much the same way as the movie starts (unfortunately I can’t provide any visual reference). First we see large palms and trees swaying (not unlike the blowballs in the opening sequence). Hence we anticipate some large force (supported by the slow heavy thumping on the soundtrack). As we see only fragmented parts of Horton’s elephant body, we assume him to be some kind of pompous Colonel Hathi type elephant. Even when he starts bending down a tree this matches our expectation. But then we suddenly see his face and all the supposed heaviness is lifted for good. We absolutely know that we are in a light hearted movie when Horton dives into the water in a silly, cartoony way.

Talking about color, it is certainly noteworthy that Horton himself is completely grey as you would expect an elephant to be, whereas the other animals (the kangaroo is violet for example) all seem to have fallen into paint buckets or – in the case of the Wickershams – seem to be made of Playdoh. Their colors are a little oversaturated to my taste but since they come in all sizes and already compete for attention with the colourful jungle, they cancel each other out. So Horton is the one that stands out from the crowd by not having any color at all (besides he’s the biggest and talkiest of them all). Apart from that it makes sense to have Horton look much as we expect him to look, because he is the protagonist whose viewpoint we share and thus we shouldn’t be distracted by his appearance.

Have a look at this b/w-version of a high angle on Horton teaching his supposed pupils. Do you immediately see the little buggers (two of which are actually called Tommy and Katie, btw, am I the only one who sees a pop reference in this?)?

Now look at the color version from above again. Their saturated primaries and secondaries make them stand out although their light value doesn’t contrast with the red earth.

In the next post I will finally deal with the cartoon poses and the actual animation.

Thursday, May 15, 2008

More on Horton

Part II: from gently curved lines to a fully fleshed out 3D world

Note: Since there is no DVD out yet, all the pictures in this review are screenshots from Quicktime trailers and promotional excerpts. I believe that it is crucial to show the pictures I am writing about. By the way, there is an interesting article by Kristin Thompson about fair use and frame enlargements.

Previous adaptations
As you may know, Horton’s adventures have been translated into animation more than once. Bob Clampett’s 1942 Looney Tune “Horton hatches the egg” is certainly the most artistically successful attempt, even though it is more classic Clampett fare than anything else (down to the usual “Now I’ve seen everything”-suicide-gag). But Horton himself is sweet and gentle (all pink for a change) and Dr. Seuss’ simplicity is preserved in lush watercolour backgrounds. The other advantage: it is as short as it should be.

Much has also been written lately about Chuck Jones’ 1970 TV special. Ted Geisel himself helped making changes to the original (and thus made it virtually critic-proof) and expand it to 26 deliberate minutes. After all, simplified TV animation wasn’t the worst thing that could happen to the faithful elephant.

In the 1990s Dr. Seuss’ original illustrations were converted to cut-out shorts narrated by stars such as Dustin Hoffman or Billy Chrystal (

All of these versions followed a different approach in adapting Geisel’s illustration style making use of the graphic possibilities of two-dimensional media. One of the challenges of 3D is the necessity of defining everything.

Filling in the blanks

It has often been argued that reminding the audience of their watching a movie destroys all identification and connection to the characters. While this might be a desired distancing effect in a Godard or Monty Python film, post modern cinema of the 80s and 90s has proven that formal experiments and obvious external interventions don’t necessarily break the emotional connection an audience has to a character (e.g. the films of David Lynch, the Coen Brothers or Woody Allen’s Purple Rose of Cairo (1985) among many others).

In classic feature animation (unlike shorts) the goal has always been to make the audience forget they were watching drawings. Of course I mean “forget” in the sense of “willing suspension of disbelief”.

So usually – as in Snow White (1937) or the Jungle Book (1967) – when we enter into a story through a book, this device serves only as a bracket in the beginning and the very end while the whole plot is set completely within the story’s world.

One notable exception though was the first Winnie the Pooh featurette (1966) and for that matter the first three later combined in “The many adventures of Winnie the Pooh” (1977). I bring all this up because the Pooh books to me are in many ways similar to the Hortons.

Although the original Pooh illustrations have been completely Disney-fied, there is no doubt that the three-dimensionally animated characters are living in a book AND come across as believable characters all the same. Instead of breaking the spell the narrator becomes a commenting character himself talking to the animals.

Apart from all the narrative aspects this way of presenting the story has a lot of visual potential in not filling all the blanks.

The loose quality of the illustrations is kept in the backgrounds that nevertheless offer enough information to suggest perspective and volume. Of course it might be a fortunate coincidence that Disney’s 1960s style with Xeroxed pencil lines works so perfectly well with this approach. But the crucial decision to me is defining only as little as possible graphically.

What makes illustrated children’s books (and the best fantasy films) so fascinating, after all, is the possibility of immersing oneself into a fantasy world that leaves room for imagination and para-social interaction with the characters therein.

While not having to define everything is one of the great advantages of hand drawn animation this becomes a bit of a challenge in 3D where you practically can’t have blank space. As I mentioned in an earlier post, part of the storytelling issues with the new “Horton” come from over-defining the characters psychologically. Nothing is left to the imagination.

While this may be true for some of the visuals, there are lots of good examples to prove that strong composition and design can make up for the obvious disadvantages of 3D compared to painting or drawing. As in Chuck Jones’ previous adaptation the key notions here are simplification and stylization.

The opening credits: an example in clarity
Through white mist the (virtual) camera dives into a close up of a dewdrop falling from a leaf onto some kind of a chestnut that rolls successively through a field of dandelions and a cave where it hits a flower with a tiny speck that floats into the jungle of Nool, where it eventually passes Horton. Although it has become a common cliché to start cartoon movies or commercials in the style of a mock-lyrical nature travelogue (dating back to Tex Avery’s early Warner cartoons), for once this works on both visual and narrative levels.

To guess the impact each part of this small series of reactions will have, it is important for the audience to exactly understand continuity and screen direction (from left to right, first down, later up again). While this is most basic film making it gets neglected alarmingly often in modern action movies.

This opening sequence serves as an ideal example to reflect on how simplicity and stylization can sometimes lead to stronger compositions than elaborate design and how cartoon forms and photorealism work together effortlessly. After all, the quality of a shot is judged by its effectiveness of getting the plot across.

In order to focus on contrast and values I’ve decided to desaturate the pictures. The colourful versions are presented in the next post.

As the Blue Sky logo fades to white we enter a very tiny world to which we have to adjust our perception gradually (a few dewdrops and a leaf come into focus [image 01]), much the same as Horton has to listen really hard to communicate with the Whos. We now know that the story will deal with small events that appear to be big. Then the dewdrops conjoin to one single large drop that is heavy enough to make the leaf bend down [image 02], analog to the Whos’ need of working together to be heard by the inhabitants of Nool. Even if I’m wrong and this has never been the intention of the makers, it nevertheless can be interpreted that way.

As the picture comes into focus our eyes are immediately drawn to the moving center of the frame to the area of highest contrast (little white stain [image 01]). We follow the drop as it merges with the others and ultimately falls off-screen. The shape of the leave guides the eye to the spot where there’s also the highest contrast [image 02].

The falling drop is still high contrast and in focus where as all in the far back is out of focus. It’s shape changing suggests real life physics.

The harmonically curved twig and leaves resemble Dr. Seuss’ ink lines and lead the eye to the visually heavy ball that is clearly readable in front of the blurred BG in spite of the overall dark nature of this picture. The brightest part is the splashing dewdrop the weight of which sets the spiky ball in motion.

As no one would argue that any tree in real life would look as stylized as this one, it still organic and believable. Since nothing moves and the silhouette reads clearly we can’t miss the bouncing movement of the chestnut on the upper edge of the trunk that serves as a visual down-hill slope to the field of dandelions. By the way the absence of squash and stretch underlines the bouncing ball’s solidness.

The heavy dark ball contrasts in value, speed and weight with the soft white blowballs.

The large forms without further subdivisions imply that we are very close to small objects. Of course there are details to make it believable but they don’t change the appearance of the forms and shapes.

We only notice the elements necessary for the storytelling. First of all we know that the ball will come from the left. Aside from that the moving leaf on the left attracts our attention immediately. The crack follows the established path. This area is also bathed in a pool of light so that nothing in the surroundings is competing for our attention. The crack itself and the ball on it are of high contrast while still maintaining a dusky light quality. The back lit up-shot emphasizes the ball’s spikes and makes it look threatening to crash on us any minute now.

Although lighting in the cave comes across as theatrical the light sources still feel organic. Focus adjustment in combination with a clear path of light makes this background instantly readable. Anticipation is generated by tracking back to shift the center of attention to the big and bright flower. All the elements around the path of light enwrap the scene warmly, with mushrooms curved in Seuss-like lines. The yellow flower in the spotlight is composed of simply shaped petals and a cozy cushion for a centrepiece like a child would imagine it. Since the flower is all the same value the speck is seen even though the picture is this small.

As is reinforced in this close up with blurred BG the speck is clearly in danger. The light comes from left above, the unseen threat (as Mark Kennedy would call it) is expected to come from the left as well.

When the relatively giant ball finally hits the flower in slow motion, this is not just for comic effect. That way the shot stands out from the rest of the sequence, so the moment of setting the Whos’ world in motion gets special attention. It is finally made clear that the journey of the speck is caused by nature and simple coincidence.

Whereas the ball falls into oblivion the bright small speck becomes the focus of interest. When it floats out into the fully sunlit jungle of Nool the strongly directed lighting makes way for flatter high key lighting. Before I will comment on the conclusion of this sequence (unfortunately the trailer doesn’t provide imagery for this part, let’s have a look at the chosen colors in this segment.

Possibly this doesn’t cover exactly all the shots in the sequence since the trailer only provides it in an intercut way.

The balance of stylization and organic naturalness of the images is clearly enhanced by the color choices. As opposed to a lot of contemporary CG features there is no garishness and no misstep in the color scheme of this sequence. The whole sequence seems unified by superficially realistic hues. Whereas it is quite tempting to unify a color palette by tinting all the colors with a certain overtone (in warm and cozy environments usually a yellowy brown that leads to colors that in the real world are associated with excrements if anything), here it is the delicate quality of early morning sunlight that affects the earth tones only as much as necessary. Interestingly the warm browns of the twig with the spiky ball (5th picture in mosaic 1) stand out over the muted greens of the blurred background. In the monochrome version there was far less contrast perceived. Most of the colors are rather muted (earthly greys, wood, the dandelions, the ball itself), even the naturally blue sky doesn’t draw our attention because it is softened by pale yellow morning light. It only serves as a contrasting device when the white seeds are in the air. But within the composition this is still the area of least contrast so as not to distract us. Before the ball falls into the dandelions its practically neutral color is more saturated so as to further stand out from the pale beige and blue (in the shadow area) blowballs. The inside of the crack and the initial part of the cave shot look almost monochrome as they are supposed to be in the darkness. In the theater the subtle contrast of warm (the almost neutral mushrooms) and cool colors within the dark cave setting is far more distinguishable than in those quicktime screengrabs.

Now the purest colors that stand out are the comparatively warm green of the leaf that fills the screen as well as the unicolored yellow flower. While the green leaf is set against a darker unfocused area of green, blue and black and thus appears to be tone in tone, the yellow flower is not only surrounded by muted dark colors but also by two cooler areas of dark blue and bluish green. Since the flower is illuminated by a ray of light, its brightness works well despite the overall darkness.

The above mentioned pools and paths of light that anticipate the path of the rolling ball are always pale and warm and contrast beautifully with the cool dark colors of their surroundings as is evident in the first picture of the second mosaic: the dark blades of grass on the left leans towards cool blue, the green of the blade/leaf through which the ball enters is more on the pale/beige side. The only exception is the ray of light that enters the cave at the crack where the ball falls down into the visually cooler area. But as stated above, this serves to increase the contrast after the flower comes in.

Obviously red in its pure form is totally absent from this part of the sequence. So are its neighbors magenta, purple and even orange. Even though the palette is clearly stylized and limited, it doesn’t come across as unnatural or distracting.

By the way, red is saved for the letters of the main title “Horton hears a who” which follows while the speck is floating over a waterfall. As for the magenta: of course this becomes a highly prominent color because of the clover Horton puts the speck on. It appears to be exclusively tied to that flower which makes it identifiable by its color alone, even when it is only a mere spot on the screen.

The soft, slightly warm light in combination with camera angles that are very close to the objects, this sequence communicates a feeling for the sheltered environment Whoville has been enjoying so far. As the speck floats into the open the camera is already far back which reminds us how small the speck must appear to an animal. Until now we have seen the diegetic world in a highly fragmented way that leaves plenty of room for the imagination.

Thursday, May 8, 2008

how can it be so hard to draw Joe Pesci

I can’t believe I’m having so much trouble with drawing Joe Pesci (the young actor of 1980), since his features are really quite distinct. I’ve always thought it would be easy, but wanting to do a caricatured cartoon version of him I failed totally and decided to just go with something vaguely resembling this great Scorsese-actor and try some facial expressions straight out of a movie.

The inspirational screenshots were from Raging Bull (Martin Scorsese, 1980), one of the films that get better every time you watch them. It’s the shot-reverse shot scene in the second half, where Jake (De Niro) gets upset about Joey kissing Jake’s wife on the mouth.

Isn’t the guy in the middle looking more like Edward G. Robinson?

Right now it seems I even managed to tone it down… what John K would call an “underture”. I tried to fiddle with proportions and started out with a smaller cranium but on every subsequent drawing it grew back into normal size…

I’ll be coming back to Bull and Pesci. There are lots of fascinating DeNiro scenes as well and besides these two actors were in a couple of great films together. I hope to come to the point where I’m actually able to invent original expressions only possible in animation. But until then I work on copying and (hopefully) exaggerating expressions I see in movies and newspapers.