Dogstar has its day

On the voyage from Old Earth to New Earth in 2347, a freak accident causes the Dogstar, a giant space ark containing all of the world’s dogs, to become lost to mankind. Borrowing Dad’s spaceship, the Clark kids go on a quest to find their missing dog Hobart and they won’t rest until they find him, and every other dog from Old Earth. That, in essence is the premise of the Australian children’s animated comedy series Dogstar, which was named Best Animation at the Asian Television Awards (ATA) in Singapore. Produced by Media World Pictures of Melbourne and Perth, the series of 26 half-hours took two years to complete, and a second season is in pre-production, with the support of Screen Australia, Film Victoria, ScreenWest, the Nine Network, Disney Channel and international distributor Daro Film Distribution. “Dogstar was a project that we had been developing for several years but we had little success at the time so the idea sat in the bottom draw until 2005 when the concept was pulled out again and virtually started again,” according to producer and creative head, Colin South. “The original idea came from a series of meetings we had with designers and writers led by Doug MacLeod. Once the characters and the basic storyline were written up on the page we then sought to find a unique look for the show. The design of the series is very much inspired by design director Scott Vanden Bosch’s unique illustrative style.” “The animation style is also unique. While we tried to emulate a classic 2D look that is actually created in Flash. Our animators come from a traditional 2D background, so their experience and skill working in Flash really shows. The limitations presented by using Flash are largely overcome by strong design and the character staging,” South added. Created in widescreen high definition (HD SR) and featuring 5.1 sound (Dolby SR), Dogstar was produced entirely in Australia from concept through to post production. Production was largely conducted in Melbourne and Perth with crew in four main studios communicating via an online asset management database called the Nelnet. Design and animation work for Dogstar was done on PCs while post and post FX work was done on Macs. The designers used Photoshop and Flash MX. The animators used Flash MX. The post team used Final Cut Pro for editing, After Effects for the majority of Post FX work, Particle Illusion (Wondertouch) for some FX, Photoshop for the occasional clean up or paint task and iView Multimedia for library management and batch renaming and processing of materials. iView’s batching function was especially useful for pre-production set-up tasks when preparing the picture and sound files for the Nelnet. All audio was recorded, edited and mixed in Pro Tools. The dedicated online database known as the Nelnet was leased from Canadian company, Nelvana Ltd for the storage, management and transfer of all design and animation assets. FTP sites were also used extensively for the transfer of picture and sound files. The series had input in scripts, designs, and picture edits from the UK (BBC), Germany (children’s broadcaster ZDF) and Australia (Nine Network) during the production process. Edited pictures were compressed in the Perth post studio and uploaded to a dedicated ftp site enabling parties to quickly download, view and provide feedback. Design Director, Scott Vanden Bosch’s lavish designs were brought to life by director, Aaron Davies and enhanced via the FX and editing expertise of Post Production Supervisor, Merlin Cornish. Working in high-definition in Flash (1920 x 1080 pixels) meant that the designers could add a huge amount more detail and texture than is typical in animated series. And with Post FX planned in from the beginning to overcome the two dimensional cut-out style limitations of the animation program, the team were able to conceive the show in far greater depth than most Flash-created programmes. At the design stage each Flash symbol (equivalent to a cut-out piece of a character) was created with further layers within it, allowing sophisticated tweaking of movement, shadows, lip-sync and cyclic animations. Many ‘robot’ characters and background elements were created with automatically playing detailed animation loops. This extra animated detail added colour and interest without adding repetitive work to the animation teams. Master designs for characters, props, locations and FX were created at the very beginning for use throughout the series. Then, upon delivery of each script, episode-specific designs were drawn up. These were detailed but not fully finished as they were created specifically for use by the storyboard artists and it was the boarding process that defined what design elements were needed. Once the storyboard was finalised, it was carefully analysed so that the Design team knew exactly what they had to produce for each episode in terms of background angles and detail, characters and character expressions, and props and FX design. Storyboarding was far more complex than any other Flash series the company has worked on. Because of the intricate backgrounds, number of characters and elaborate interaction between characters and backgrounds, the storyboarding had to double as the layout stage. Storyboards were intensely revised to best suit the needs of the stylistic and complex designs. Upon completion of a storyboard, each board panel was scanned and an “animatic” – a picture edit composed of storyboard stills – was edited to the pre-recorded voice track. The programme was edited to length so that each scene was already timed to the director’s wishes. Each animatic went through a minimum of three edits to fine-tune the content and pacing before the animation went into production. Each animatic was then broken down into individual elements of image and sound. The image file for any particular scene contained 1 – 24 storyboard panels and the accompanying voice tracks. On occasion, guide audio for key sound FX was also included so that animation could be carefully timed to synchronise with the FX. The movie and sound clips were incorporated into individual Flash files by the Scene Planning team. This team also added the appropriate designs to the scene files then uploaded these scenes to the Nelnet ready for download by the animators. The planning process meant that the animators were neatly presented with dialogue, designs and a set of timed and edited storyboard panels for each individual scene. Each animated character was broken down into many more parts than is usual with most Flash animation projects. Although this meant animation production was more involved and time-consuming, the complex animation that was produced gave a richer, more intense and full feeling to the animation. Characters in Dogstar typically have separate hands and arms, separate feet and legs, and separate eyebrows and eyes. A wide range of eye expressions were created, as well as many extra mouth expressions beyond the standard lip-sync. Animators were encouraged to push Flash beyond its normal limitations. Typically, Flash animators use the cut-out elements supplied to them by the design team but on Dogstar animators were encouraged to do a lot more “free” animating, to draw more symbols and to draw more animation within symbols. Thus in an arm move, instead of letting Flash simply in between one drawing of an arm from one position to another, the animators were encouraged to draw several more naturalistic in-betweens, thus resulting in smoother motion. Once this set of drawings had been created, it could be reused in other scenes to ensure that the animation was consistent across the series. The Nelnet online database enabled efficient communication and data transfer between crews working in four different locations. Media World Pictures had first used the database created by Toronto-based animation studio, Nelvana Ltd when co-producing the animated sitcom John Callahan’s QUADS! with them between 2000 and 2003. Media World negotiated to use the system again and thus all the series’ designs and animation assets for Dogstar were housed in Toronto. Entering the Nelnet site via password access, crew members could search for designs, download and upload scene elements, and communicate with colleagues through a comments panel available for each design or scene. The Nelnet set-up enabled team leaders and production management to get a detailed snapshot of production progress department by department. This was a substantial leap forward compared to the laborious tracking of hundreds of paper folders across multiple locations that traditional 2D character animation production management demanded. To address the limitations of Flash animation and create the most visually appealing programme possible, visual FX formed a key component of each episode. FX work was layered in at the storyboard, design, animation and post stages of production. Over one third of the scenes were given some sort of FX treatment in post production with the crew focussing on adding visual depth and on giving any element that would occur naturally in the environment a more organic appearance (e.g. light, smoke, water and flames). Early in production, Post Supervisor Merlin Cornish made custom filters using presets in After Effects. Then, as episodes were first assembled using the completed animation, Cornish would note the key or ‘hero’ FX for each episode and commence work on the trickier treatments whilst the picture was going through its editing stages. This way, by the time the main FX block in the schedule was reached, the more detailed FX were well advanced, if not already approved. The FX were mostly achieved by taking apart the Flash scenes and applying filters to individual Flash layers using After Effects. In order that different effects could be applied to each layer, each Flash animation scene was constructed in a hierarchical manner so that the Post FX team could easily separate elements such as background, mid-ground and foreground, as well as each animated character. As the series progressed, the Design department managed to incorporate some of the team’s FX plans within the designs rather than leaving them for the post FX stage, thus allowing Post to direct more time to creating other FX. To achieve the goal of adding visual depth, focus pulls and background or foreground blurs were regularly applied. Where focus pulls were done between Flash layers in a scene, minor scale adjustments were made to the layers to create a ‘lens’ stretch feel. The resulting effect subtly suggests that the scene was shot by a camera and subconsciously creates the impression of greater depth in the viewer’s mind. Where time permitted, glow FX were added to make the most of incidental lighting design. The lighting layer would be given a transfer mode so that the light would spill over into other layers and thus appear to be in the location rather than a single cut-out lighting element. ASIAIMAGE

Sponsors & Media

Subscribe our Bi-Weekly Digest

Receive up-to-date news by
signing up to our bi-weekly digest.

Have a tip or pitch to share?
Send us a glimpse.
We love contributions.
Digest Media