Since its birth in the early 20th Century, surfing’s unique allure has seduced people all over the world. It’s incredibly difficult to learn, but its laid-back sub-culture and cutting-edge attitude – and the fact that you get to hang around at the beach for hours on end – has established it as one of the world’s most popular sports.

It is also an inspiration to the creative world. The sport has inspired fashion designers, musicians, and moviemakers, but surfing has spawned a design culture of its own, too. The surfboard is a classic example of the convergence of form and function. Surfboard designers, or ‘shapers’ as they are known, have developed the board, starting with the wooden long boards used by the first surfers on the beaches of Hawaii, California, and Australia in the early 1900s. The shortboard revolution of the 1960s saw surfboards become smaller and lighter, while today surfing is facing the fact that new materials have made the surfboard one of the most toxic items of sports equipment on earth.

While shapers have developed the efficiency of surfboard design, artists have seen the surfboard’s potential as a canvas. Influenced by surfing’s inherent coolness, surfers aren’t the type to be seen with a ugly board.

Facial animation and lip-syncing can be the difference between a successful animation and a ropey one. It requires a massive amount of effort and incredible skill to get it right.

Lip-syncing is the act of getting your character’s mouth to move along with an audio track of some spoken dialog in order to create a convincing performance. While simply moving the jaw up and down may suffice for quick character animation work, it’s not going to be enough if you intend your character to be engaging and believable,
or at least for it not to be distracting or weird-looking.

Just like in non-verbal facial animation, lip-synching seems to get exponentially more difficult the more accurate you try to be. The brain is hardwired to decode human facial movements to the nth degree of subtlety, so it can be very difficult to deceive your audience into accepting that what they are seeing is a natural and realistic performance from a digital character.

First, in order to make a character talk you must animate the face so that the physical representation of the mouth shapes match the words in the audio file. The basic unit of speech in this case is not a word but a phoneme. Phonemes are the fundamental sounds that make up words, such as the ‘th’ sound in ‘that’ or the ‘oo’ in ‘spooky’.

When lip-synching in 3D, you don’t need to create custom mouth shapes for each and every word. Instead, you create a library of phoneme shapes and cobble together the performance by keyframing those shapes on your mesh, making sure the timing matches their occurrence in the audio file. Despite the relatively small pool of phonemes, it’s still a complex and time-consuming business. You do have the opportunity however, of deciding on the size of the phoneme set that you will use depending on the quality of the final animation. For good basic results you can get away with around nine phoneme shapes, while for more accurate results you may need to use as many as 40 or more.

The phoneme shapes, called visemes, are generally created before you animate and as each phoneme occurs you set a keyframe that applies the viseme on the mesh. How this is done will vary from application to application. In Maya, for instance, you might create a series of Blend Shapes or bone poses for each viseme, while in LightWave you would probably use Morph maps. Either way, you deform the mesh to match the sound file.

Most 3D applications allow you to perform basic lipsynching using their standard tools, though being able to view and hear the audio waveform directly in the timeline is a great help if it’s supported. Without this ability you will be forced to plot out the timing location of the phonemes with a pencil and paper and using this as a reference.

However, there are a number of dedicated lip syncing tools available that help speed up the process and even automate some of it. In this round-up we have six of the best lip-sync tools and utilities on test and it’s interesting how differently each handles the process.

One of the most laborious aspects of lip-synching is ascertaining which phoneme is where in an audio clip. It’s no surprise then that some of the applications focus on this problem by analyzing the audio and extracting phoneme animation from it automatically. Some of the applications let you apply the animation to a model in 3D view, while others only export animation files that are then applied to control objects or for keyframing morph targets in your main 3D application. While their functionality and methods vary they, all should help speed up what is a time-consuming and laborious task.

One of the most technically accomplished effect sequences in the movie has been tagged ‘Gotham Sunrise’ by its makers, and shows the sprawling home of Batman in its entirety – and it presented a huge task to create for the team at Double Negative.

“The art department miniatures unit, led by James Hambidge, created a model showing the layout of Gotham City,” explains Double Negative visual-effects supervisor Paul Franklin.

“The concept was of a city built on series of islands and sprawling out across the neighbouring landmass like a gargantuan version of New York. Chris Nolan was keen for the final shot to look realistic, while simultaneously conveying the idea of a city out of control, bigger than the biggest city that exists anywhere,” Franklin
explains.

The initial plan of attack – to shoot it entirely in miniature – quickly ran into budget difficulties due to its sheer size, making a digital recreation the only alternative. Yet Nolan needed convincing.

“Our first task was to prove that a digital version could capture the same level of complexity that was present in the maquette, so we put together a construction kit that allowed us to lay out a city of 80,000 structures, light it, render it and shoot it to film.” It proved a hit with Nolan.

“The buildings in the final shot are sourced from Chicago originals, with slight modifications to make sure that we didn’t end up with multiple copies of the highly recognizable Sears Tower in Gotham,” says Franklin. The model library was assembled through a combination of Lidar, photogrammetry and online research.

We ended up with a database of over 2,000 structures, all of which were individually textured with photos taken by location scouts in Chicago, each showing their correct facades.”

In order to capture a more observed reality than the overly stylized Gotham City’s of previous outings, Double Negative mounted a series of stills shoots from the tops of tall buildings in Chicago, culminating in a shoot of the dawn rising over lake Michigan as seen from the roof of the Sears Tower, 1,400 feet above street level. The photos gathered from the shoot served as the primary reference for the lighting scheme in the final shot.

“Eventually, we ended up almost as victims of our own success as Chris felt that the shot looked so convincing that he was worried that audiences wouldn’t pick up on the city’s extreme scale,” says Franklin. “We spent the last few weeks on the shot adding buildings until it stretched all the way to the horizon. In total, there were over 500,000 detailed digital buildings in the final shot.”

Based on the Riverside Plaza in Chicago, the imposing Wayne Tower is a full CG model. Double Negative extensively photographed Riverside in HDR panoramas in May 2004, and acquired high-detail Lidar scans of the building. Using the data, Double Negative began to model the tower “in the same way that a model maker might build a spaceship from existing plastic kit parts,” says Paul Franklin, VFX supervisor at Double Negative.

“The Wayne Tower complex was our most detailed piece of digital architecture, and was designed to play for both day and night,” says Franklin. All of the windows in the building feature detailed, illuminated interiors. Each window is also individually dressed with drapes or blinds.

To supplement the huge amount of detail in the surface textures gathered from Riverside Plaza, Double Negative added procedural brick and stone block textures and displacements and added extra weathering for the late-stage version of the building.

“From the outset, VFX supervisor Janek Sirrs and co-VFX supervisor Dan Glass stressed that Chris Nolan expected us to raise the bar on the quality of all the VFX work,” says Double Negative’s Paul Franklin, visual-effects supervisor, who was charged with this unique VFX task. “Chris’ primary brief was ‘make it look real’, by which he meant that he didn’t want to be able to tell the difference between digital VFX and live action when projected side-by-side on film.

“Chris was also concerned that digital methodologies would be at odds with his approach to filmmaking, which is very much rooted in the established language and techniques of cinema – his worry was that the unrestricted freedom of digital camera moves and the inherent ‘weightlessness’ of CG would jolt the audience out of the finely textured reality that he intended to create for Batman’s universe,” adds Franklin.

All of which presented a challenge for Double Negative. It had been approached by Sirrs and Warner Bros in November 2003, and invited to bid on the project. Preproduction had already begun the previous June and, while the team had a considerable time for VFX creation, the CG planned meant Double Negative needed to offer Sirrs and Warner Bros the lure of coming up with a completely new toolset and methodology in order to meet the demands of Nolan. It won them the project.

“We didn’t feel that there was an existing toolset or pipeline out there that would fit this show, so we came up with a proposal for how we would structure our work around the needs of the production and Chris’ film making style,” says Franklin. “Our ideas dovetailed nicely with Janek’s thinking and we were chosen as primary vendor in December 2003.”

From the go-ahead, Double Negative installed a small previz team at Shepperton Studios, which was acting as the base for the production of the movie. The team worked closely with Sirrs and Glass, spending six months designing and developing potential VFX shots that were then offered up to Nolan. Yet, rather than create a previz-heavy storyboard, the previz work was used differently than in conventional films.

Batman Begins has swept in as one of the surprise blockbuster hits of this summer’s cinematic releases – and sees director Chris Nolan drag the caped crusader back to his roots. The result is a dark, edgy reinvention of the dark knight that avoids the camp commercialism of previous movies – and it’s all the better for it.

Yet reinventing Batman involved more than a new latex costume and retooled batmobile. A vital aspect that Nolan needed was a raft of topdrawer special effects; only, Nolan didn’t actually want them to appear as the centrepiece in the way that films such as War Of The Worlds showboated its VFX. For Nolan, Batman Begins was to feature effects so seamless that if the audience were to watch a reel alongside a CG-generated sequence, they wouldn’t be able to tell the difference. Instead of bombarding the audience with special effects, Nolan was determined to create a film that used them with the subtle touch of a master surgeon.