This is where I'm at as of yesterday. What is this, 28 months in or something? Work over the last month has been highly effective, and I'm finally getting something close to the original visual look I imagined.
What you see here is probably hard to decipher. Looks cool but what does it mean?
This video was kind of a survey, or test, not of the visual side, but of the creative matrix of the Save Point pipeline (basically a think tank of specialist robot brains networked and bridged out to viz programs).
Here's what it demonstrates. I let the engine run untethered for 12 hours, and then just recorded the output at around one thought per frame. So what you're really seeing here is the current scope of how the pipeline can visualize scene X that it is given to direct. This video isn't long enough or fast enough to capture a big section of the existing scope, even right now at the beginning, but you can start to get an idea of what's possible when I can draw 3000 unique scenes in a single work day, and have them be pretty consistent in composition, lighting and color.
Right now, I've still got 1000 ways that I want to improve it, so that process is constant.
The interesting thing about this particular video is that you can pause on any frame and there's a decent quality picture that's on brand. I cut it down to around 1500, but that's a lot of screenshots for one day. Try it out, roll the dice and just pause on any random frame.
Upvote
4
