Today we have the pleasure to chat with James Neale, manager of Character Mill and Corporate Member of the Blender Network, to talk about their latest project Gorilla Band.
Charactermill is a branding we use to market our character-based work. All our mascots and avatars, posters, illustrations, animations based around character work, including our project Kajimba and Jellibots fall under this umbrella. This work forms the bulk of our workload for the guys here on a week-week comparison but we also provide visualisation services to other types of clients. We brand this work under The Visualisation Company (completely different style of work, so we decided to separate the folios). We also service gaming and gambling clients under the brand Red Cartel, but myself and my wife aren't involved in the day to day running of this production as that style of work is very long form and our guys can operate with minimal input.
Our physical studio is really just a central point for workstations, files and data. At any time there are 6 or 7 people onsite and again the same number of remote freelancers, which I manage personally (Dropbox and cloud rendering make things so easy in many phases of 3D production).
Thanks! We're definitely excited as it is our first full Unity app. Gorilla Band is actually our 5th e-book. Our publisher is called Wasabi. One of the guys who owns Wasabi is a writer and the other is a musician, so they wrote all the storylines and musics and we (CharacterMill) built the entire project into an book app using Blender and Unity. In house we did all the art assets, 2d and 3d, menu design, animation, all touch interactions, UI, UX, etc. All our Unity "programming" was done using Playmaker and so the most technical aspects such as analytics, in-app purchasing and social media integration were outsourced.
The project itself took 6 months but much of that was waiting on feedback. Overall, the work was accomplished by a total of 3 modelers/textures, 2 animators (myself and Jeremy Davidson) and 1 Playmaker/Unity guy.
On the workflow, there was an awful lot of reworking, mostly owing to the fact that we had no known base upon which to build, and the sheer limitless capability of making better and different graphics in a 3D engine so quickly. This meant that any time our work was reviewed, we would get newer and better inputs on visuals/storyline or approach to screen interaction. This created major changes almost every time we recieved feedback, many of which resulted in our original submission anyway, so the whole process was not without conflict, frustration and added cost.
In the end the app works really well (we think), but it definitely wasn't an easy project to build even though it isn't a complex piece at all. We all appreciate the lessons learned on both sides and hopefully this will translate to an easier and simpler Unity project next time. (nervous laughter)
No custom tools. We just used Blender and Unity in their default state. Blender is already a fantastic tool and it connects almost seamlessly to Unity (save a blend file into the Unity project tree, and instantly it appears inside Unity, available as a 3d asset without any export process at all).
Strictly speaking, the creation of 3D realtime assets is quite simple, but the Unity/Playmaker route proved to be the big tricky issue since there was only 1 guy onsite who could decipher the miles of noodle-trees that he needed to build in order to solve each of the constant changes/adjustments from the publishers.
Playmaker is incredibly powerful, but without proper optimization/documentation, it can be very difficult to troubleshoot issues, especially when the project has grown organically, where much work consited of patches and bugfixes on top of the original code. Most of our Playmaker "coding" work was done tactically to meet the needs of the request at hand, so as more and more feedback-zombies arrived, the bug-chase/fix got harder and harder. Thankfully, the build process from Blender-Unity-Mac-iPad was straightforward and we could test it easily on the devices. The executable was under 45meg for the whole book, including all music, models and motions, making distribution via Testflight very quick indeed.
Needless to say: finally publishing the App was very rewarding.
Blender still suffers from the mentality from industry that it's for hobbyists only, yet time after time, the Blender community produces such excellent work, increasingly so. I already run introductory information sessions for professional 3D artists about Blender, and so far we are getting a great response. It's only a matter of time before it will become the norm to know and use Blender daily.
Nothing of any note that I can talk about, sorry. Sometimes work we do doesn't reach the public domain for several years, which makes it really hard to build a folio when nothing we've done in the last few years is allowed to be shown. Very frustrating when competing against artists who update folios daily!
This Gorilla Band app is thankfully not one of those projects :) That said, Blender is and will remain our primary production tool in our studio and we'll definitely still be using Unity to deliver realtime projects, along with our usual prerendered projects from Cycles/BI.
While I think Blender is almost the perfect tool for small studios, it's also important to focus on creating good artwork and leave the software to speak for itself. Personally I'm a huge Blender fan but every day I can see the damage that rabid Blender fans do on the internet.
It's just frustrating to have these over-eager fanboys who just reinforce the hobbyist-toolset mentality amongst pro 3D artists.
Our Blender info sessions are being well received so I feel we are slowly turning the tide for artists around Sydney helping them realise how powerful it can be for an artist when used properly.
So, really for me, it's onwards with art-making and troll-dodging!!!
Thanks for the chance to chat.