DRI/Bitmoji: Singheads – Character Animator

My primary role at Dream Reality Interactive was to be the character animator. I also took other responsibilities such as 3D modelling, texturing etc. However for this project I was mainly the Rigger/Animator.

Singheads is a free to play karaoke game. Using face tracking technology – create an avatar, play your favourite songs and watch your avatar come to life.

What you see here was the 3rd iteration of the Signheads character avatar. This was just before the partnership with Bitmoji/Snapchat, which ended up becoming the primary assets in the experience. We did however spend a lot of time creating a complex customisation system using our own rigs, characters, and props which enabled us traction throughout the project to attractnew partnerships, publishers and investors.

The animation side

The base animation mainly consisted of basic poses. The reason being is that the game itself was built off of a face tracking system that would be manipulated by the positions on the users face. Another layer of facial customisations are present underneath the face tracking layer.
Control Rig: This rig was made with a combination of joints and controllers. This is a fully customisable and easily transferable rig, with blendshapes used specifically for the customisations.
Control Rig: I also learnt a lot about the Node editor and how to create relationships between the joints and the controller attributes. This also led to me learning about interesting effects such as squash and stretching.
Facial Features: Most of the customisation was done through Blendshapes, (eyes were customised through joint rotations). The blendshapes all work well with the animations which allowed the player the ability to fully customise their avatar. You could even combine various Blendshapes together to create alternative looks.
Facial Features  – A lot of the facial features that we used were separate models that were rigged to the same as the face.
Facial Features – They all worked well with the animations, without clipping but in some poses we had to fix – we did this by adding corrective Blendshapes that would activate when certain poses were made. The beard is a good example of this, we had different blendshapes for around the mouth or for different jaw shapes.

Leave a Reply

Your email address will not be published. Required fields are marked *