I’m wondering how the voices in cartoons get aligned with the visuals.
I would assume that the visuals are done first and the voices added later. Do the voice actors watch it a bunch of times until they can anticipate exactly when to say their lines? (That sounds tough.) Or do they just say their lines to their own time and then some tech person aligns it to the visuals? Or something else?
as someone working in the animation industry, usually the voice lines are recorded first, and the animation is done to match the voiceover.
Really?
That means that the voice actors would have to time their delivery to action that they have never seen. I mean, there are pauses of various lengths, in which action takes place that does not involve anything spoken. You’re saying that the voice actors have to guess how long that action will be and time their words to match up?
Why would things be done like that?
I would think that they go by story boards that lay out the broad details of the scene they are voicing. The story boards would contain a timeline and a cue for when they are to speak. These can be as simple as pictures on a wall or more detailed views of running shots in real time on a computer, but not quite animated scenes.
It’s not like they’re doing the whole show/movie in one take. If there’s a long break, then they just skip to the next scene.
Exactly. WB cartoons were recorded first, then animated with one actor (Mel Blanc) often doing all of the voice work. Each line (often recorded several different ways) is a separate clip that can be assembled by the animators however they like as far as pauses and beats are concerned.
Because it’s easier for an animator working on one frame at a time to sync up with a voice, than it is for an actor working in real time to sync up with a moving image.
So, how does it work for anime or other translated works, where the animation is already done?
Animated shows are meticulously storyboarded before production begins. By the time the voice actors come in to record their lines, there will be a video rough cut of the keyframes (animatics) that everyone can refer to. After the lines are recorded, they are synced up to the animatic and passed down to the animators. From there, the animators proceed to create the actual animation based on the animatic.
For songs like in Disney movies, the storyboard artists will get an early version of the song that they use to plot out the action to match up with the music.
example below:
Voices always comes first, otherwise how would the animators sync up mouth movement to the voices? It also helps the animators create a believable performance, as they can let the emotions in the voice guide the acting choices they take when posing the characters.
For an example of recording voices after the action is done, look at any sort of dubbing (anime, foreign movies). The voice artists will try their best to match the mouth movements of the original footage, but it’s never going to work perfectly.
Usually, badly. Watched Speed Racer recently?
No, not since the 80’s, but I seen have more recent animes that are not nearly so jarring. Unless you are talking about the movie, in which case, no, I did not see that either (only the Cinema Sins treatment of it).
I listened to an interview with Liam O’Brien a while back, and he not only did voice acting, but directed voice acting. They do a lot of work to try to match up the animation to the voices.
Modern digital editing helps. Play around with some voice files in Audacity and you’ll see what I mean. You can stretch things out, compress things, or make silences longer or shorter. The translator can also spend more time and effort coming up with phrases that more closely match the existing animation. There’s a lot more money in translating anime to English these days than there was a few decades ago, to they can invest more in the production of the translated version.
Indeed. Watch a recent anime in English with English subtitles on. The subtitles will be a direct translation of the original Japanese but the overdub will be quite different.
I imagine a lot of that is necessary when the actors are big names and aren’t trained dubbing actors like those used for live action.
If you ever had a chance to see the 1982 release of Fantasia, the one where Irwin Kostal re-recorded the music in digital format, instead of using the music that had been recorded first, then animated to, you know that the results are terrible when the sound doesn’t match up well with the visuals.
I think they only showed that version once, then yanked it from distribution. None of the versions available on VHS or DVD used this take. (You can, however, get a CD recording of the music. I suppose you could try to synch it up with the video, if you wanted.)
https://www.filmscoremonthly.com/board/posts.cfm?threadID=103731&forumID=1&archive=0
They should have hired Pink Floyd. They can do that kind of sync work by sheer chance.
I’m sure someone’s done a “Dark Side of Fantasia” by now.
A quick google search finds a couple of videos on youtube, but they’ve had the sound removed (copyright, I assume).
Damn you, Roger Waters!
Very often, the voice actors in the cast aren’t even in the same room when they record. They read their lines and the sound engineers put it all together.
I know that William Shatner did some of his voicework for the animated Star Trek in hotel rooms when he was traveling.
To see how it’s not done, look at Robin Williams in Mrs. Doubtfire. He’s supposedly recording as he sees the animation (by Chuck Jones BTW), but it’s pretty clear that it’s unsustainable, especially when he goes off script.