Sometime back when I was in a Night Club and totally out of my comfort zone, I got an idea for a game. I can’t dance to save my life and I took this concept literally. Imagine you had to play a sniper whose job is to kill the guys in the Night Club who can’t dance. This was the seed of the idea for Night Club Sniper.
Around that time, Yorrick Mendes decided to join forces with Yellow Monkey. I pitched the idea to him as a warm up project/prototype and he liked it. We began prototyping it on Unity. We got some cool free “Clubby” tracks from royalty free music sites and we setup a basic prototype. Inside a night club you could see from a first person view, a bunch of dudes dancing. Each song would be broken up into “sections”. Each section would have a set number of legal steps and exactly one dude would be completely off timing with his dance. The key difference between this game and a Whack-a-Mole is that here everyone would be doing a different legal move. You would also have a bunch of laser lights which would look exactly like a sniper rifle laser guide (which was your only weapon). The player has to shoot the guy who was off timing, If you shoot correctly a new person would take his place and someone else would do the incorrect move. This would enable you to get combos. We started off this test with an animated cube that Krishna worked on.
We liked the result but were not sure if it tested the skill-set of spotting off timing dances with humans. We decided we needed human dance animations, and since we didn’t have any animators who had enough time to do this, we went the Kinect way. We setup a poor mans mo-cap rig using a Kinect and all of us and some of our friends choreographed and danced (kudos Kapil Samant) to various sections to populate our dance step library.
After battling lag and quite a few rejected animations we had what we needed.
NCS Prototype (you will need the Unity Web Player Plugin to play, we suggest full screen play!)
We playtested that prototype and realized that we needed very tight animations, something which was difficult to achieve with Kinect mo-capping. We also needed a huge library of unique dance steps (upwards of 100 unique steps a song on an average) and while the core gameplay was challenging and interesting, we weren’t sure if it was strong enough to carry the game by itself and justify the high content requirements. Finally we have decided to shelf this project for the time being as we are not sure after all this effort if it will be worthwhile, but it was an interesting project to try out and we had fun. On the bright side, the shelving discussion also spawned the idea for our next prototype which all of us are really excited about! :D
PS: We spent close to 3 weeks perfecting the laser projections to give this crazy BIT.TRIPesque confusion and the link had a build without lights. Great! Anyway we fixed it now. Enjoy!