A tad overdue? yes.
Here’s the previous posts to catch you up:
Part 1 – Thinking + Planning
Part 2 – PreProduction
Part 3 – Layout + Camera
And now… on to Animation! (strap in – it’s a long one)
Animation is a funny thing. There are practically an infinite amount of ways to do it. However, the end product is what’s really important, and everyone has their own way of getting there. I never see any two animators work exactly the same way. In fact, its usually more like a constant search for a betterworkflow rather than a singular consistent method for just about everyone I’ve talked to.
Even during this one project I went through several different workflows. From Pose-to-Pose to Straight-Ahead and everything in between.
Where to begin?
For this particular project I decided to start with what I felt was the most important part of the piece, the reaction shot. It’s the first time you see the Rocker up close, and its the point at which you connect with him. Not only do you get a chance to see what he’s feeling, but you are fully aware of WHY he is feeling that emotion. This is why I felt this was the best place to start. Whether or not the shot actually turned out to be that successful is another matter – but now you know why I started there. You can see theplayblast by clicking here:
I spent probably the better part of 2 or 3 hours just on the one pose. To me, it was THE POSE of the whole animation, so I was willing to spend the time. I posed everything. Right down to the face, the fingers, and all sorts of levels of minutia. Once I had a solid pose that I was really happy with, I found a place for it to go. You can call it a moving hold if you choose, or a new pose, depending on how you define the terms, but what happens is that he slowly reacts.
To be honest, I didn’t know exactly what I was going for on the move, I animated it straight ahead and just did what I felt came naturally. What I ended up with is that I dropped the guitar a bit, and raised his head and neck up. I was happy with that general direction, so from there I just started layering, paying very special attention to Opposing Actions. It’s a very important concept to grasp, and something I find myself constantly struggling to remember when animating. The head was moving up, so moved the jaw down, thus opening his mouth a bit more. And since the head was moving up, I also dropped the shoulders down away from it. As the guitar drops the fingers tighten and move upward. These sorts of opposing actions happen throughout.
The next thing I did was start to block the last shot. Using stepped curves, I blocked in the dialogue for “a huge section”, including facial and mouth shapes. I think I got down to having a pose about every 4 frames, and then for the next “-of my music” line, I had a pose every 8 to 10 frames. It was not a quick process, and took me about 2 nights worth of work to get there.
*sorry there’s no sound on this one. He’s saying “-huge SECTION of my music”, but theres only mouth shapes for ‘huge section’
At this point I was struggling quite a bit, and felt I needed a change of focus to keep my motivation and creativity – so I stopped working on shot 3 and moved on to shot 1. I used the basic timing of the director from the Rough Layout Movie. I chose three major poses to create: when the director entered frame, when he took his first step onto the stairs, and when he takes his first step onto the stage.
From there I changed the curves to splined and started working in the curve editor. Yes – after only 3 poses. I have defined where I’m going and at what frame, but a WHOLE LOT of how I get there is left to be determined. I feel that I get the spontaneity and naturalism of straight-ahead, but the grounding and direction of pose-to-pose. This is actually my favorite way to work (as of now). It’s almost like a ‘glorified straight-ahead’ since I know my next pose, but it might be 15-25 frames away. However, it’s also almost entirely ‘Layered’ animation as well, because of the actual process I use (described below). So call it what you will. Glorified layered-ahead?
My Generalized Workflow (as bad as it is…)
For the case of the director’s entrance, here’s a little insight into how I worked. After those 3 poses I was in Splined Curves, and I started cleaning up by going into the Root and focusing on the up/down movement to make it look like believable steps for his entire way up to the stage. I work almost entirely in the curve editor. Which doesn’t mean I’m not watching what’s happening in the CAMERA VIEW (it’s very important to see how things look from the view that your audience will see. Who cares if it looks great from the side or the back if people will only see it from the front?). When working on the Root, I often times turn off the visibility on the arms so they don’t distract. Next I added the twist and side-to-side of the hips, and then the feet. The hip stuff was all in the curve editor, and the feet were done mostly by pulling in the perspective view with auto-key on.
I didn’t worry about the toe roll, heel roll, feet contact, or anything like that because no one would ever see it – his feet are off screen! BUT, his feet need to be somewhat working correctly to make the visible parts of the legs look believable. Once he takes his steps up the stairs, you CAN see his feet, so at that point I payed attention to contacts.
After the lower half of the body was working the way I wanted, I moved on to the torso and did the twisting, side to side, and cushioning of the weight transfers. Then I moved up to the neck and head. The reason I usually start with the root and work my way up the back to the head is that everything comes out of the Root, and if you have mistakes there, you’ll just be counter-animating them in the torso or head. Or worse yet, something will be very off in the head movement and you spend hours trying to figure out where its coming from, just to find out its a hitch in your Root, or maybe the back.
*you can see how I ignored the stuff that would be off camera (outside the green box). And the guitar player is still the layout version.
After this I went back to shot 3 with the dialogue. I tried to do some traditional pose-to-pose blocking and breakdowns again for the untouched part of the line, but quickly aborted and went back into the first line to clean it up. Usually NOT what you want to do. So now I have the first line cleaned up and close to final, and the whole rest of it isn’t even started. Not a critique-friendlyworkflow, since you can never really show it to anyone for feedback. However, this is the way I continued to finish the project.
I took the next couple of lines section by section, and did only the body (not the face or the lipsync). I did it in much the same way I did the Director – where I would define the next pose I wanted to hit, but just straight-ahead figure out how to get there. I had a pretty good idea of the little accents I wanted to hit along the way. Like head shakes/nods or shoulder shrugs, and I just roughed them in as I went. (once again, in spline and working from the root outward).
Then I would go in and add the face and lipsync (which I’ll talk about in a second), and then started anew with the body on the next section of dialogue:
Now when doing the facial stuff, I tend to start with the eyes and brows. I think they can tell the audience a lot about what the character is feeling. I try my best to keep the brows working as one and feel connected, like Carlos Baena describes in his tips and tricks. Keep in mind also that brows do more than just up and down, then also move in and out. The inner sections move in and out quite a bit, actually. As for the eyelids, I try to keep them really alive. With every eye dart, our lids will follow. Notice how this woman’s eyelids are posed for her casually looking down and up. When you’re character is looking down don’t keep the upper eyelid up where it usually is! Also, Irecommend you check out Kevin Koch’s blog posts on eyes (movements and blinks).
I also remember reading somewhere that the upper lids display our level of alertness (drunk/sleepy-thru-frightened?), while the lower lids display the intensity of our emotion. Which I’m pretty sure is why I squinted the lower eyes up for the more intensely delivered words.
I didn’t start with mouth shapes, I started with the jaw. I think this is something really difficult for students to start with (certainly was for me) but at this point in my life (and it could change tomorrow) I like to start with the jaw. I try to get the ups and downs of the chin working in a very readable and entertaining way. But lets backtrack and illustrate the bad way first.
If you do OPEN-CLOSE-OPEN-CLOSE for the jaw on every syllable, it starts to strobe and become very boring
This poorly drawn image is supposed to represent the amount the jaw is open. The dark blue line is supposed to represent the path of the chin.
Basically, if the line is low the jaw is really open, and if the line is touching the lighter one then the jaw is closed.
This is how I approached lipsync in school. But not every sound needs its own opening, and not every opening needs to be all the way open:
*Note: drawing this out isn’t something I usually do, the image is just to try to explain it better
Notice how the jaw doesn’t get wide every single time it opens. And “I can’t” is almost one jaw move (and could probably be simplified into one). By doing it this way, you are practically creating a visual rhythm with the jaw.
After this stage, then I will go in and start to add the lip shapes. The same principle applies here – you don’t have to hit every sound there is. I didn’t separate out the “I” from the “ca” in terms of the lips (but I did click the tongue off the roof of the mouth for the ‘K’ sound in it). And its probably pretty far from perfect – but I think it reads correctly.
One of my favorite parts is how he starts to form the shape for ‘rewrite’ long before he says it. Proof you don’t have to suddenly snap into the correct shape. You can (and usually should) anticipate what sounds your going to make. And in this particular case, I wanted to play it like he was struggling to say the word – like he couldn’t even fathom that he’s being asked to rewrite it – which is why I have the extra long anticipation. His lips form whats coming next before his brain will let him say it.
Seems to be a bit of a touchy subject in the realm of animation. In my opinion, as long as you are using it to gain understanding in order to animate then its fine. If you are using it to copy, or rotoscope poses, or steal timing, then it’s just a crutch. Besides, your character is a better ‘actor’ than you are – because he’s really feeling the emotion. He isn’t even acting. Use the video to increase your understanding of what you’re studying. Find the essence of it (whether it’s a pose, the timing, the… anything) and push and pull and exaggerate it.
I used video reference for only one part of this project – the eye darts in the silence of “i can’t… rewrite what’s perfect.” And the reason I filmed the eye darts is because I realized I was lost. I tried it without researching it, and I had the darts going rather sporadically. Up-left, then center, then lower-right, then upper-left, etc. It was ugly. I had just haphazardly thrown in some different eye poses and expected it to look like he was searching his thoughts. It didn’t. So I put the camera on myself and tried to ask myself a question and really search for the answer (what did I do last Tuesday night? or something). Upon studying the video I realized that my eye darts actually followed a pattern. They jumped around on a path that would create a circle (if they kept going), it didn’t appear to be random.
I simply used the video to inform my creative juices and let myself exaggerate from there and create this character’s performance based on what I learned.
In fact, once I realized it wasn’t random and had discovered the generalized pattern, I deleted the video so I wouldn’t be tempted to copy the timing or the spacing. Which is why, unfortunately, I can’t show you the video now. 🙁
Well… looks like that’s it. I apologize once again for taking so long. I’m not going to close with any sort of ‘final thoughts’ except to say that animation is always a changing process for me, and the next project will probably be handled totally different than this one. I would also like to reiterate that I’m no genius, and I don’t claim to have any sort of authority on how to animate. I merely wanted to share how I went about making this piece, and a couple of things I try to remember while working. I just hope that it helps somebody… even if it’s because they learned how NOT to animate by reading this. And even if no one gets anything from it, at least I know that I learned something by analyzing myself.
Now that I’ve come to the end of the workflow, I suppose it only seems fitting to once again show the final product:
Anyone have any thoughts on anything I’ve mentioned?
Thanks for reading!