-
Making the write-on effect smarter and more efficient.
I already bit the bullet and my machine is slowly trundling through this render, but in order to learn for next time, I’m wondering if any of you have found a mechanic to leverage the Write-On effect without kiling the machine.
Backstory:
I’m using write on to trace the path of moving nulls, about 30 nulls with about 3 minutes of movement, keyframed in 10 frame increments.
This kills the machine. Write on is extremely slow to process each following frame (which is perfectly understandable, really, given the amount of data it’s crunching), but it also seems to be doing it stupidly.
What I mean by that is, it seems to re-calculate the entire write-on for each layer at each keyframe, rather than using the previously calculated frames intelligently and only calculating the change incurred at this current frame. So at frame 20, it’s calculating 0-20. At frame 21, it’s calculating 0-21 rather than using the work it just did at frame 20 + 1 new frame.
It’s not scalable, working this way, especially in very large comps. (Think an airspace tracing over a major airport, 1000s of lines with a variety of movement patterns!)
Stoke or path animation is not possible here because it would mean manually/painstakingly analzing %age transition to movement that is not linear. (Null might move 20 pixels, then wait 200 frames, then move 50).
Any other alternatives to the write on effect (that use an identical mechanic, brush position and time) or any advice on how to make write-on be smarter with its render management?
For now, I did my best to leverage stroke distance and minimal keyframes to make this as renderable as possible, but it’s still just a silly amount of processing per frame.
Thanks!