Random Gooseshittery - Update #2 (Patreon)

Good afternoon all. It's been a week since the last post. Here's another update on progress.

I've finished cutting together the bullshittery. It exists in non-text form with lots of crude placeholder graphics. And its ready to go into full animation keyframing. A slow process that will see about 5 minutes per week.

The total video duration is 36 minutes. So 7 weeks. Yikes.

To summarise the prior post, the reason for taking longer with the initial cutting was my concern that I would spend ages keyframing only to trim things out later. And I wanted to make sure the whole video is absolutely what is going to be in the final cut. That it's worth the animation time.

For this update though, let me tell you about some of the things I've had to thinking about during the cutting. A look into my process.

"Dead air":

One of my priorities as an editor is to have things play back smoothly. It's likely something you've noticed in the prior bullshitteries. My goal isn't to show things accurately - but to make people laugh. I must never lose sight of that objective.

There's an interview with John Cleese (sadly no longer available on Youtube) that's lived rent-free in my head for years. Describing how the 1975 comedy "Fawlty Towers" had guest actors not wait for laughter of the live audience to finish before jumping straight into the next gag.

John Cleese explained that the audience watching on television, in their ones and twos, would not be affected by the contagious nature of laughter as the people in the studio were. Therefore they had to compensate by speeding things up to try and get the same comedic energy.

A major way that I choose to do that is the removal of "conversational dead air". The breaks, the pauses, the "umms" and the "errrs". The normal gaps that define human conversation, but that rapidly become the kiss-of-death for any comedic compilation or montage. Acting as the energy breaks on the whole experience.

Lets take the following example scene. Here I playfully respond to somebody making a mocking comment (dropbox link with audio):

"Indeed, well I am just a goose, Froggy. My cranial capacity is eliminate errrr limited! But not my capacity for evil. That is unlimited. For I am a goose!"

That's how it went in the livestream. But collectively that's 15 seconds long. And about 7 seconds of that is just silence. And this specific example includes a verbal stumble, followed by an "errrrrr". Both of which would be taking up animated text space, perhaps even their own paragraph? Making the viewer work harder if the text is moving rapidly.

So what do I do?

Well, if I chop it into pieces and remove the breaks, it becomes this. Still imperfect. But much smoother on the ears (dropbox link with audio):

Those who watch the VR livestreams may remember me caring so much about the Elgato capture card trying to mash all of the audio layers together. This is why! For having isolated audio layers lets an editor do this sort of thing without cutting background sounds.

But it looks terrible doesn't it? It's skipping all over the place. A common problem for any timeline with lots of heavy cutting. So this is where I start getting tricksie.

For I can simply preserve the microphone layer, delete the existing visual cuts and associated game audio. And stretch out one scene to make it flow continually. The voice now offset by at least 8 seconds. But the benefits should be obvious (dropbox link with audio):

It's not accurate to how it went down. But it's now drastically improved. Even before the text is there, which will in turn animate smoothly above the goose.

And this is the sort of sneakiness I regularly pull. I do this sort of thing in most, if not all bullshitteries. Tweaking the audio so that it's out of sync with the visuals if it makes it funnier. To make it smoother to view.

Maybe there's some audio that was said whilst I wasn't looking in the wrong direction? Maybe there's a joke that's falling to the cutting room floor but I can squeeze it into a transition between two other scenes to keep the energy up? Maybe the original visuals were so jumbled with sub notifications and alerts that I should swap it out with something less headachy.

Whatever makes the scene flow smoother. Whatever works best in post. Whatever produces the biggest chuckle.

Eyeline continuity:

A little harder to demonstrate before the text has been put in. But another consideration during the cutting phase is eyeline continuity.

As you can probably guess, this is where I need to try and make sure the action in the frame is in roughly the same place in the next shot. Below would be an example of bad eyeline continuity. For if you imagine some text above the goose, your eyes would be bouncing from left to right to follow what's being said.

Once or twice is fine, but imagine if every single cut was like that.

Therefore, during the cutting phase, using the visuals trimming technique above, I've also got to consider whether the prior scene has the subject in roughly the same position. And if not, whether time can be moved forward or back to make it so.

Here for example, the goose remains in the middle of the frame throughout. I intentionally shimmy the cuts to make it so.

And there are some other little tricks I can pull.

• I like to create a central invisible square called a Null Controller. Then I parent all subsequent objects onto it. Because that Null Controller is always in the exact same position and has the exact same scale values, I can quickly copy and paste values onto adjacent Null Controllers. So I need to shift all scenes/text 20% to the left? Maybe zoom out slightly? Whatever it takes to make continuity between scenes smooth. Even after all the keyframing.
• Alternatively, I can freeze the visuals for a couple of seconds if the camera is moving around too much just before a transition. Perhaps add a cheeky camera wiggle effect to hide the freeze. Whatever it takes for a smooth transition.
• In extreme examples, if there's too much camera wonk or too many sub alerts, I could go back into single player and stand in the precise same position. Refilming 5-6 seconds and cutting it in without anybody even noticing.

Essentially, I need to make sure that the viewers eyeline is not bouncing around the entire frame. That it's not exhausting to try and follow. And that thought process starts at the initial cutting.

So to summarise - my plan currently is to begin the animation work. Aiming to make around 5 minutes or more per week. I shall get started and let you know if there's any problems.

Thank you all for your support as always folks :)