The Register

Making Muvizu: DIY animation is here

Even though it's in the alpha stage, MeshMellow's Muvizu is one of the most impressive new technologies I've seen in the past few years. But don't take my word for it. You're going to be annoyed by Muvizu output pretty soon.

What the software - pronounced "Movie Zoo" - does is simplify 3D animation to the point that a child can do it. Or a drunk or stoned student - which is part of the target market. In doing so, Muvizu leapfrogs basic Flash creation tools, which are currently the tool of choice for stoner virals - 'Hey Mr. Taliban', say. But it also takes elements from professional and hardcore enthusiasts such as machinima software or ToonBoom, simplifies the features, and gears the software around creating a narrative very quickly.

Barry Sheridan, the lead developer, explains that the software is based around direction, not animation. "It differs from machinima in that you can jump straight into narrative; you can get to work on the movie immediately, and tweak it bit by bit."

From File, New… a character appears - blinking, and ready to animate. Slap some clothes on him, add a speech soundtrack - the eyes and mouth synchronise to lipsync automatically - and drop in a set, and you have a very quick and dirty animation. Because Muvizu uses non-destructive editing, it lends itself to an iterative process.

The harder stuff - such as adding cameras, sets and moving the characters around - has been the focus of some clever thinking. For example, you just point the mouse to where you want the character to walk; in walk mode, the mouse cursor drops a trail of dots. Changing speed and gait is easy. Adding in sets is easy, with particular attention to lighting effects. For someone wanting to do the next Tally- My-Bananas, I noticed that you can map an image on to your character's face. MeshMellow is taking a community approach, where small groups can collaborate. One person may specialise in set design, another in writing jokes, another in doing voices. It looks like a jobs board. You need to join to get the software. At this stage, pre-beta, there are limitations. You're limited to four cameras. The character rigs mean Muvizu animations will have the same body shape - for now at least. The team says rigs will "probably" be contributed by the community - resulting more interesting characters (animals and aliens), but the "physics needs to be there - we need to know that a character's finger can touch the nose." Since it's based on the Unreal engine, Muvizu accepts open standard Collada digital assets such as 3D geometry.

This month sees a one-click "upload to YouTube" option, and for the Beta, expected in January, new moods, textures and fighting - a much-requested feature. I wondered why, with so much innovation here, and with a product that clearly offers something quite new and disruptive, Muvizu was showing it off - albeit in a low key way. MeshMellow says various design elements - such as getting the character to walk - are protected by patents.

"We've filed patents around some UI techniques. What MeshMellow has is a rare combination of skills; we want people to start using the software. This is not software that benefits from being hidden, it only gets good by being used by ordinary people," an executive told us.

MeshMellow has sprung out of the DA Group, the company behind the original Ananova. Developers there already had experience in animation, 3D coding and web skills, but rather than create yet another Sadville VR world, the potential was end-users creating their own animations. "A Garageband for animation", if you will. It's backed by private investors.