Technical Report – Last Days: ante diem, episode 1
Last Days: ante diem was a project that was developing slowly until we had an opportunity to pitch it the guys at Celtx. When they responded by pick it up as part of their first crop of seeds it basically jumped from the middle of the priority list to the top. We had a great company that was excited to help fund and promote the project so we instantly went into gear working out all of the details.
The script was written as a small piece of history that would tell the story about how a virus was created that was part of a much larger, much more elaborate and much more expensive project called ‘Last Days’. It was also being developed as a motion comic. Since it was our first experience with animation we recruited an experienced animator who we used as a resource for how to create the final product. I’ll get more into that later, but the gist is that we only had our live action experience to guide us through this process so we basically started with these steps in mind.
- Record all of the Dialogue and lay it out onto a timeline
- Create rough storyboards to go with the dialogue track
- Have illustrators draw out those storyboards into individual elements (background, characters, props, etc.)
- Place all of those elements into compositions that can be manipulated into moving scenes
- Do all of the finishing (music, SFX, color, etc.) and make everyone happy with a cool scary short
Creating the Dialogue Track
We started with getting Chris Mollica (Albert), Murisa Harba (Elana), Jeff Elam (The American) and Chad down to Big Joe’s Sound where Tim Burgoon recorded the entire script with the actors isolated into separate booths. Chad then took the dialogue tracks, picked the best performance and laid them onto the timeline based on the tempo that he wanted. This worked out great, but everntually required a bit of shifting simply because more time was needed in certain scenes to make the action fit or the transition not seem too rushed.
The next step was to illustrate each scene shot by shot. John Sandel was our Scenic Illustrator and Travis Stanberry was our Character Illustrator. Chad gave them a list of tasks to draw out specific elements of the scenes, which were then scanned and clipped out of the PDF. Chad created photoshop projects and placed all of the elements for each scene as separate layers, shaded them and then forwarded the projects and assets to me. The photoshop process is what we did in the beginning, but eventually realized that it would be much faster if he just used Photoshop to clip and shade the images, then composed the shots in After Effects. Once they were composed I started the process of animating them in After Effects.
We originally decided to create all of our AE compositions as 1280×720 at 24fps. By the end of the project this created a couple of obstacles that we would have to resolve, but the reasoning was basically that since we were doing animation for the web that doing it at 24fps would make animation smooth and we felt that since this was going to be posted on youtube that it would be overkill to go full HD. When we were eventually given the opportunity to screen it at a festival we regretted the resolution decision, but it ended up looking great anyway. I will talk a bit more about the problem that 24fps posed when I get to audio.
So the animation process was basically to take each composition and place all of the elements in a 3D environment. For some of the more complex shots, such as The American and Albert walking through the hallway (Scene 2) I took all of the wall images rotated them 180 degrees on the Y axis, took the floor and ceiling images and rotated them 180 degrees on the X axis and lined them up on the same plane and placed them next to each other on the Z axis, essentially creating a long box. I also added a few other props such as the lights all the way down the corridor that I created. This made it so I was able to start the scene with the characters starting at one end of the corridor and move down to the other end. I duplicated this scene so that I could add a camera from each angle whenever I need to have a reaction shot.
I timed out each scene based on the dialogue from the audio track, which was basically just a list of timecode points where I would need to create a specific movement or points where I could animate the character a bit based on what they were saying. This part was pretty time consuming, but I found that if I copied only the audio for the scene that I was animating and placed it on it’s own timeline then the timecode would be pretty easy to track. Once I had the timecode for each important key frame I basically made precomps in AE for each character image that had multiple parts (movable eyebrows, eyes, lips, etc.) and adjusted those elements to reflect what it was that the character was saying or expressing. I also used the puppet pins occasionally to supplement some of the movement. At points where I needed to switch between the image for each character I lined the two different images up and placed key frames 3 frames appart where the opacity of the first image went from 100% to 0% and vice-versa.
Then I would place a camera into the composition frame up a well composed shot (that’s debatable) and depending on whether the characters were moving (as they are in Scene 2) I would place key frames on the camera position timeline to either lead, follow or simply create some movement to mimic a dolly or crane shot. I would also place keyframes on the cameras focus distance so that a specific part of the image would be in focus. There are several moments that you will see a rack focus between characters.
The final step would be to render the composition, which I rendered as ProRes, and import the image into Final Cut Pro 7 where I had the dialogue timeline.
I am fairly new to After Effects, but haven’t found a way to get realtime audio playback that would make things any faster. I did the final edit in Final Cut Pro 7 because it is the software that I am most familiar with, but the next episode will be done in Premiere Pro because it can round trip between AE. In most cases the first render would either not match up perfectly or I found that I wanted to do more to make it better. The quickest way that I could think of to replace the MOV files that I had imported in FCP was that each time I rendered an updated version I would simply rename the new render the same as the previous render, delete the old one and place the new renamed file into the original source folder. Final Cut would just assume that it was the same one and I wouldn’t have to reimport and/or shift the position. I call that duct-tape-round-tripping.
Once we had all of the animated video that was rendered from AE, everything was placed on the timeline and everything was synched to the audio, I exported OMF’s and a low res reference video which Kendall used to do the sound mix in Soundtrack Pro. Since I didn’t do that work I won’t go into too much detail, but basically she effected the dialogue to match it’s environment, added ambience and sound effects and Chad layed out the music that was created for this project by Kyle Puccia.
This is when the 24fps became a problem. When Kendall exported the final audio track everything was out of synch. The first part of the audio was fine, but as you went down the timeline it would get more and more out of synch. We brainstormed why this was happening and realized that when most NLE software puts out a 24fps it is actually 23.98 and since our project was exactly 24fps the export from Soundtrack Pro was out of synch. As in most situations in film production there are several ways to tackle the same problem problem. We tried a few that failed, but eventually the one that put everything back in synch was to adjust the speed of the audio track in FCP to 99%. Which is pretty lucky because we were only a couple hours away from our delivery deadline for the screening.
I never intended to be an animator and this is a lot more like animatics, but in the end I found that I learned a lot very quickly and the project has exponentially increased my understanding of After Effects. We have at least two more episodes and I have a lot more ideas on how to improve the viewer experience. I am glad that the story alone is strong enough to keep the interest of the audience, but I hope to eventually make the animation just as entertaining. When we are finally able to produce the next episode I hope you check it out. I know the story and it is pretty intriguing.