Reflections on the Process: Chapter 3

Whew! This is going up late.
This was a difficult chapter. It was pulling teeth every stage of the way, from getting the illustrations done, to alignment, but somehow, I managed to get it all done on schedule.

** Why was chapter 3 hard?

I can't account for why /illustrating/ it was so difficult--except maybe that of the early drawings, some were so /nice/ that my standards for myself shot up, and I struggled to meet them. (I did meet them, ultimately.) I think by far the bigger reason the chapter felt like such an uphill battle was the crowd scenes, which took a couple hours to illustrate, nearly as much to clean, and were a massive alignment time sink.

** What worked

• Following the same basic process/order of operations as chapter 2 worked fine
• Making quick thumbnails with no repeating images rather than a real storyboard allowed me to a) do visual brainstorming & think visually first, while b) minimizing the amount of time spent on storyboards that won't be followed closely
• Improvising JavaScript animation effects when I needed them gave the story an added level of visual expression (I do need to come up with a place/process for abstracting and saving these for reuse--but in its own time).
• Having a firm deadline helped me pace and push myself
  

** What didn't work

• Trying to draw all the illustrations on back to back days, with no breaks, just made me burn out. I need to figure out how to pace myself so that I can have illustrations ready in a timely manner, /and/ feel OK while I draw them. I like drawing. If this is gonna work long term, I have to figure out how to stop turning /Roots/ illustrations into a stress factory.
• This chapter was waaayyy more visually ambitious than I accounted for in my planning. I got it done it two weeks, yeah, but it nearly killed me. I'm glad I had the deadline, and I'm even glad I stuck to it, but for future estimation purposes, I did not give myself nearly enough time. *There needs to be a step, early in chapter planning, where I go, "what's unusual about this chapter? Is there anything I haven't really done before? And then if there's any part that's likely to take a lot of learning, custom code, engine updates, etc, I multiply my time estimate by like 150%, and decide how I want to schedule the chapter release/my life based on that.*
  

** What I learned

• At least at this point in my learning/engine development, substantially animating /anything/, even if the illustration burden is low, is going to massively inflate the time spent on a chapter. That doesn't mean I want to avoid it--I want to look for chances to increase it, actually, but it does mean that any time I decide to include animations, I have to plan for extra time.
• The engine I've written is pretty easy to hack extra JS animations onto! I don't know if that's because it's /well/ designed, or because it's a really shallow stack on top of Vanilla JS, or because I have an insane idea of what "easy" and "nice" look like--but whatever the case, it's great news for me!
• You /have/ to get all the white edges off the outlines of background characters. Even crowds. Especially crowds. Getting it right may be tedious as hell, but it is needed.
  

** Developments in the Engine

• Frame based animations can now auto-generate however many cross-fade tweens you want

I think that's the only big one.

I also have the first interlude up, as of yesterday, so I'll be posting another update today. Stay tuned.