A few questions on animation

DDanimate · 6 січ

Hello, I am self studying Spine and I've accumulated some questions (not strictly Spine specific, but more like animation in general). Sorry if this is not a good place to ask those, but I don't really know where else to post this. I tried somewhere else and got no replies.

Is it me or is editing the animation is and will always be rather hard? Especially when it's pretty much done and especially when you want to add or remove entire chunks of it?

And more often than not, there isn't an easy a-few-clicks trick to do so and instead, you need to carefully consider each part on the timeline and work with them manually?

Timings/duration. So what I usually see and do is setting a few poses, adding some details, then offset stuff. The parts are offset (either manually or in case with cyclical animation - through "offset" tool) but the duration is usually the same.

But what about different duration? When, how and why can it be applied? How common is it?

Timeline. The more there are keys, the harder it gets to navigate through. Knowing that, I think it can take me quite an effort to try to keep track of it, sometimes keeping things measured and organized.

Of course I look at what's going on on the screen and as of late I'm trying to rely more on the sense of what feels right as well as training the temporal and spatial feeling.

But I wonder what's the general approach and philosophy towards it? And what's the progression of skill in that regard?

Head, arms, body.

I'm doing some facial animation for the first time. Through articles, chatGPT, a few videos, some self recordings and studying some disney animated movies frame-by-frame I more or less stitched together the general idea of what goes after what: Brows/eyes, eyes/brows; eyelashes (sometimes they go first); mouth; head. Am I correct?

Then I tried to understand how that works together with body movement. Now I've read somewhere that body and face (and head by extension?) should be treated as separate entities, but that alone didn't really clarify much. The animated movies and self recordings show that body can move at the same time as head. And that arms can move at the same time as as body or even earlier(?). Now it's clear that if someone is in a more tight and collected state, the offset would be smaller, but seeing it being around zero is rather confusing. Especially opposed to the idea that the movement starts around the pelvis, then goes through torso, then through shoulder-arms/neck-head.

I feel like maybe for mechanical (getting up, walking) or ample power moves (e.g. jumping, throwing, punching) the movement starts from pelvis and for expressions/reaction/emotions it happens pretty much at the same time? If this one's too much to answer here, maybe you could suggest some good literature or tutorial on this

Spinebot · 6 січ

Editing animations, especially when they are nearly complete, can indeed be challenging. This is because animations are often complex, with many interdependent parts. When you want to add or remove chunks, it usually requires careful manual adjustments to ensure everything remains cohesive. There isn't always a simple solution, and animators often need to meticulously manage the timeline and keyframes.
Different durations can be applied to animations to create variations in timing and pacing. This is common in animation to convey different emotions or actions. For example, a slow, deliberate movement might have a longer duration, while a quick, sudden action would have a shorter one. The use of different durations depends on the context of the animation and the effect you want to achieve.
Managing a timeline with many keys can be difficult, but it's a skill that improves with experience. Animators often develop a sense of timing and spacing, which helps them organize and navigate complex timelines. The general approach is to keep the timeline organized and to rely on both visual feedback and an intuitive sense of timing to make adjustments.
Facial animation often involves animating different parts of the face in a specific order, such as brows, eyes, and mouth. However, the relationship between facial and body movements can vary. While body and face can be treated as separate entities, they often move together in a coordinated way. For mechanical or powerful movements, the motion might start from the pelvis, while for expressions or reactions, movements can occur simultaneously. Understanding these nuances comes with practice and study, and resources like "The Animator's Survival Kit" can provide valuable insights.

Nate · 6 січ

Is it me or is editing the animation is and will always be rather hard? Especially when it's pretty much done and especially when you want to add or remove entire chunks of it?

Definitely. A smart workflow is very important. You can easily waste 4x the amount of time or more by taking the wrong approach. Plan ahead of time to reduce rework. Block out your poses first, as it's much easier to make changes early on. Refine your between poses before you start fiddling with curves. Polish at the end.

It's common for people to approach it willy nilly. They pose straight ahead and polish curves as they go. Later they need lots of changes because they did no planning. It's very painful because they have lots of keys and changes ruin their curves, so they do the same work multiple times.

If all this sounds painful even in the best case, welcome to animation! Consider the traditional way to do this is to draw every frame by hand, and if you want to adjust anything, you have to redraw ALL the frames again. As you can imagine, planning for traditional animation is even more important since the reword pain is so much higher. We have it easy with our digital tools!

Check out the favor tool workflow:

It's a very efficient way to work for many animations.

But what about different duration? (offset)

Offsetting is usually for a looping animation. You define your animation duration before doing any offsetting.

The more there are keys, the harder it gets to navigate through.

Certainly, there can be many keys in a complex animation. You don't normally look through the dopesheet and necessarily understand what all those keys are doing. You would think in the context of a few bones and analyze only those keys. That is why the dopesheet and graph have the selection locking features. The dopesheet has a sync mode that shows the timelines the graph is showing.

Our resident animation expert Sinisa can field your more specific questions.

I think you are taking a great approach to being an animator! Many people don't have the patience to put in the time and effort to learn thoroughly. Have you seen our Animating with Spine videos? I highly suggest doing the exercise in each video, you'll get a lot more out of it than just watching.

Sinisa · 7 січ

Hey Danimate!
First, I want to congratulate you on choosing animation as your field of study. I think it is the best job in the world. I have been practicing it since 2002, and I still find new and exciting things to learn and excel at. Let me try to answer your questions as best I can without creating any confusion. Although animation can get extremely technical, it's still an art form, so sometimes some terms and ideas can be hard to understand at first.

Yes, editing animation can be very difficult and tedious. The best way to go with it is in my experience to be very organised. By that I mean, plan your animation ahead. Know your main, 'golden' keys, then your breakdowns. So if you need to edit your animation you know exactly which part of the animation you are working on. e.g. you are animating a character getting out of a chair, and your anticipation lasts from frame 1 to 8. you notice that it feels a bit fast, so you add frames in between those keys to slow it down. What helps to stay organized in Spine is to key the whole character, all the controls, so you dont think of the character as a bunch of individual body parts, but as a whole, as one drawing.

Timing and spacing are pillars of good animation. That is something it's good to train. Its good to start with simple things, like bouncing ball. With offsetting things you should be careful, sometimes moving the curve is not enough to make the parts of the character to look believable enough. My general workflow with most of the things is to make the main poses just right, considering all the elements of a good design, like the line of action, good and clear silhouette, where is the center of gravity, straight vs curves etc. then when I play the animation and I see that the timing and spacing feel right, I will add all the extra movement according to main action. Sometimes it is hard to see where exactly should some things fall so it's better to have the main things working and then add all the extras. Dont just offset stuff just because you know that some parts should move later in time, try to really pose them. This is where artistry and experience come in play, with time and training you will get better at adjusting the poses to feel better. I designed Animating with Spine videos to have exactly the exercise to make you a better animator. I did those exercises countless times and I recommend you go through them.

Timeline can be tricky if you think of your character as a collection of bones. Think of your animation as "poses" and look only at the first, main timeline, and think only as where are my poses in time. Then when you have your animation working more or less OK, you can go through individual bones and polish their curves. This is something that you will also get better, at because you will be able to identify just by looking, like OK my translate Y needs to start moving faster.

Well, this is a complex topic, in general, no, elements in our body are not separate entities, we work in unison, and one thing always affects the other. BUT, things move in a different times. My experience is again to think in poses, not individual body parts. Let's say you are animating a jump animation. So you have a starting pose, anticipation, jump, mid-air, fall, recovery, and some kind of neutral pose. With every pose you are making you should think of the whole body, the whole pose, and how it connects with the other. Once you have that dialed in, you can go through different parts of the body and poolish them, like e.g. you want the left arm to drag a bit more. So each individual part of the body has its own arc, timing, and weight but the whole thing needs to work as a whole, otherwise, the character will look and feel disconnected. So always think of a pose and unity first. As the face goes, the same thing. The face moves as a whole. So try to think in expressions, not in elements of the face. So let's say you have a face going from happy to a frown, you would start with making these main two poses, where all the elements are included, eyebrows, eyes, nose, mouth. Now, how you transition between those poses is where again your skill is important. Maybe the character was happy but then he saw someone he didn't like, so first he blinks, with a bit of relaxing the brow, slowly losing the smile, then the brows go fast into angry one and the mouth follows. It's hard to smile and not include the eyes or the eyebrows. So yeah, always try to think of the expression you want to achieve, and then how to add good breakdowns to look natural and organic.

I would definitely recommend you to always first and foremost plan your animation. Maybe some basic thumbnails, shoot references etc. If you are working with reference try to really analyze where are the main storytelling keys, and where are the breakdowns. I would also recommend you go through the Animating with Spine videos and try to do all the exercises. Some of them are tedious but stick with them. You have an analytical brain which is great. If you have some specific questions, dont hesitate to ask. Animation can be very topic-specific, and more often than not rules or principles do not apply so some creative thinking is necessary.

DDanimate · 8 січ

Nate
Sinisa
Thank you guys for your replies!
Regarding questions, I'd like to try to clarify the questions number 2 and 4.
The second question was more about...well, let's take simple example of the basic workflow: You set a few poses. For simplicity, let's exclude breakdowns, details and such. The obvious next move is to Shift or Offset the keys on the timeline. Virtually all durations will be the same. So shifting would be very easy, right? But what if durations (number of frames) on some are different? I guess a part of the question was somewhat...anatomical?... and was actually about "In what cases and how often can they be different? Is body parts moving with (obviously different speeds but because of different spacings but with) the same duration is the norm and different timings are more niche and specific cases?" and the other part is how to actually best handle the situation when some parts have different timings?

Thanks for linking the favor tool video. While watching it, I noticed you didn't do any Shifts. Is it because you tried to emulate it manually through graphs? Was it the application of the "Dont just offset stuff just because you know that some parts should move later in time, try to really pose them. " advice you mentioned in your reply (sorry, I didn't quite understand what was meant here, did you mean I should try making key poses as if they are already offset/shifted?)

Regarding question 4, the two situations where head and arms move at the same time as opposed to the pelivs/hips->torso->shoulders etc chain still confuses me. Not that I can't comprehend how the former (simulatneous movement) is possible, albeit after following the tutorials where movement starts from pelvis, there perception could get a little skewed and you kinda start viewing the latter as a basis; but it's not really clear when to choose what.

Thanks again

Sinisa · 3 лют

Danimate
Well, what you are referring to is a really good and tough question to answer. I dont prefer just moving stuff around because I animated for years drawing on paper, so I like to think of the character in Spine as a drawing, and all the poses as breakdowns and in-betweens, so if I need e.g. do drag the hand behind the body, I pretty much know how it should look drawn. Looking at Graph View curves just helps me to see the fluidity of motion for certain elements (so if there is some bumps, weird timing, or weird keys, I can spot it immediately). Maybe the best practice for you at the moment would be to find very good references and go through them frame by frame. Thats the best way to analyze and learn animation. I do this practice almost daily to keep my animation skills sharp, and I always get surprised by how some things are moving. And sometimes, for good animation, you really need to sculpt every frame to look good. Look at your character how it looks, not your bones, so if you need, translate or scale the bones to deform the drawing to get the poses you want.