Team Fortress 2 - Update #2

Good evening folks. Welcome to the weekend. Here's another update on Random Team Fortress 2 Bullshittery (part 1).

It's the end of day 10 of an expected 35. And I should be be around the 5:00 mark for the text and animation keyframing. I've reached 5:15, so I'm slightly ahead of schedule. The plan now is to complete 5 to 10 minutes next week (week 3). Then dive into the Source Filmmaker work the week after (week 4).

So one of the things I'm noticing immediately is how each game type changes the "DNA" of a bullshittery. Producing vastly different comedic moments. And Team Fortress 2 is no exception.

But it's doing it in a way that I really didn't expect.

Whilst I was able to predict that most of the jokes would be fast paced in nature, I didn't predict how few opportunities it would give me for over the head tracking. Therefore much of the text is having to be static on screen.

The reason for that is that you spend surprisingly little time looking at your team mates in TF2. And the individual responsibilities of each class differ so greatly that there's often great distance between each person.

Someone might say something funny as an engineer. But if I'm playing a Medic, there's virtually no chance that I'll be looking in his direction at the time to let me overhead track it. I might not even be in the same area of the map. And it seems to be a true for most classes.

My first instinct was to merely move the video out of sync with the audio. To find a place 10 or 15 seconds before where I was looking at that person. Or just putting the text above someone else's head and let the audience assume it's the speaker. Both tricks I frequently use in order to make the text easy on the eye.

But Team Fortress 2 has some very distinctive audio that immediately makes tampering very obvious. It's not like DayZ where one bit of grass underfoot is the same as any other. And the moment the weapons start firing out of sync it's immediately noticeable that something is wrong.

And the cosmetics make everyone extremely distinctive. Meaning you're able to properly identify consistently dressed people (Messy almost always wears a Santa hat, Quebec has a fez, etc). Meaning that bit of trickery isn't going to work.

Instead what I've been doing is using wiggles, quick slides or scaling changes to make the static text look a bit more interesting. Or tracking annotations that aren't the spoken words, but are much more interesting to watch. Like this gif below, for example. 

Additionally, on watching the video this week I found myself a bit dissatisfied with the middle. And felt I got the balance a bit wrong and it was all a bit too quick.  So last night I did some late night work recutting it and now I'm much happier with the result. It should flow much better now.

So now the plan is to keep keyframing the other half of the video. Getting set to animate the Source Filmmaker segments immediately afterwards.

The editing continues :)