My previous post on Project Structure was before I had completed my first episode. So, what has survived as I move on to episode 2 and what did I get wrong? (As I write, episode 1 is going through final music still, but everything else is done. Almost there!)
My Project
It is first useful to be clear what my project is like as some of the tips are not relevant if you do your project differently. For example, Episode 1 (less than 10 minutes long) has 80 scenes in Character Animator, 30 background images, 100 separate voice recording files, and 7 puppets for 3 characters (I have kept front/rear/side puppets separate for now for each character). Having so many scenes increases the organizational overhead.
In comparison, Okay Samurai in his February 2018 video introduces a production he is working on that has an establishment scene 30 seconds long, and other scenes over a minute long. He talks about interleaving the talk track of multiple characters using Adobe Audition – I do it all in Character Animator because I typically do a separate camera angle per line. It gives a different feel to the video. So the style you are going to follow can affect your workflow.
I suspect if you want to reduce effort, longer scenes are better.
Oh, another interesting difference was Okay Samurai said he took around 4 hours per puppet. He is a designer and expert on creating puppets however – I took much longer than that to learn the product and get my characters right! But I am faster now than when I started, as subsequent characters typically follow identical structure as previous ones for consistency. He also mentioned spending days on some background images. That is where I managed to cheat. I almost gave up the project because of background images, until I started using Unity 3D and assets from their asset store. I now just take screen grabs within the 3D world. That is still working well for me.
Another thing to keep in mind is what is your episode length. Writing for YouTube people want fast paced content. That is, they prefer content that could be stretched to 30 minutes crammed into 10 minutes. That means I try to move the story along fairly quickly. I suspect my first series (currently planned at 13 episodes of around 10 mins each – 2 hours all up) might be the same as 6 to 10 hours of longer form content. Each episode is relatively standalone (important if people find you on YouTube), but the purpose of my series is to show progression of the characters as they develop. That means I had to map out a timeline for the growth of each character and intertwine that into the story as they progress. I want to get a bit more depth into the story line of the series. Either that or I am just over complicating things! 😉
Script Structure
So how about the script structure I was using. I still follow the same structure as I described in the previous post, except I now drop the assets section. That section was very useful for the first episode as I had to create my puppets (and learn how to use Character Animator at the same time) and generally get organized. But episode 2 was pretty much the same, so I dropped the section. I still use all the rest however as described.
I noticed Okay Samurai in comparison had scripts that were just character name and dialog. I include additional directions for camera angles, the emotions I am trying to get across, etc. because of the nature of my episodes. For me the character emotions matter as they character is trying to progressively grow. The voice actor however only sees the current episode – they need the extra detail at times to understand the right emotion to use.
I also include reminders for myself as I go back and try to assemble the overall script. I need this as this project is a hobby, not a full-time job! I forget the details between sittings.
The hierarchical numbering I continue to find a life saver to help me keep it all organized. 13 episodes x 80 Character Animator scenes (what I call a camera shot) in one episode is around 1,000 scenes to keep under control. I still include who is in a shot per shot – it makes assembling the rough cuts faster.
Overall Workflow
My overall workflow is shaping up as follows:
- I flesh out a series overview document, listing highlights from the 13 episodes. I use this to get sequencing and balance per episode right. This document evolves until I start working on an episode, at which time that section of the overview document is locked down from change.
- I then flesh an episode out to the next level of depth starting from the episode outline in the series overview document. I work out a part outline (links back to the series overview), then scenes per part, and camera angles per scene. I tend to have one to three scenes per part, then lots of shots per scene. It is quite common to have a part with only one scene.
- I build up a Character Animator project per episode, making a copy of all the puppets, backgrounds, sound files etc in each project. I may end up producing an episode every month or two, so this allows my puppets etc to evolve slowly between episodes as I learn better ways to do things.
- I try to put a first rough cut of an episode together – create all the scenes in Character Animator including the background image and puppets for that scene. No movement or details yet. I do a voice over myself quickly to help work out the timing, just recording directly in Character Animator with background noise etc.
- I then review/edit the rough cut to work out if the length feels right. I create scenes in Character Animator built from other scenes to the part level, so I can do a quick preview of the overall episode without rendering externally. (My laptop struggles to deal with a full episode in a single gianormous scene).
- Next I work on any special effect sections – these make the episodes feel more interesting. Masking is a simple one. Some episodes need more complex effects. They improve the visual interest, but can be quite time consuming as they are often custom per episode.
- I then get the voice actors in to record their lines now the script is stable. This reduces the need to go back to them afterwards to re-record lines. I found that recording a second time can sounded different (I am using friends at home, not professional voice actors with a stuido). So getting it all done in a single recording session worked better.
- I then flesh out the puppet acting for each scene to line up with the voice track (e.g. tilt heads to line up with what they are saying). I do some of this while waiting for the voice actors to come back to me.
- I then render each scene to a H.264 movie clip and link them together with Premier Pro from which I create a complete episode video. I give this video to a friend who offered to help with background music. He then provides me music files.
- I do a final layering of background sound effects and music in Premier Pro as well (so sound effects and music can gracefully span scenes), and generate the final video.
Conclusions
Is this the best approach? Maybe, maybe not! But it is working for me given the style of cartoon I am after. I would recommend you think about the feel of the cartoon you are after – such as the number of scenes you plan to use. I think things would be faster if you have longer scenes.
Also my cartoon artwork is more Japanese Anime styled (semi realistic) rather than pure cartoon. That also creates some additional effort especially since I am not really an artist – I really struggle to get even the side view and front view of characters to look similar, let alone additional face angles. I am still at the stage where I can see that a drawing I just did looks wrong, but don’t know how to fix it. Its amazing how different even slight changes in a jaw line look. I think that will only be overcome with more practice.
But an important thing also to remember is cartoons are cartoons. If you want 100% realism, use real actors! The challenge therefore is to understand the medium and style you are using, and work out how to convey emotions best using the tools at hand. I cannot turn the head of my characters, but I can control their eyes and eyebrows, as well as head tilt. So what emotions can you control with just those controls?
Extra Ordinary, the Series 