Animation Status Update (Q2 2021)

Animation
,
1

The road to 2021.2

As highlighted in the Q4 update , we are continuing with our commitment to Quality of Life, and advancing our investment in DOTS. This means there are no new changes to our roadmap since this announcement, we just wanted to keep you in the loop, as we know it has been a few months.

There are a few other related topics to let you know about though; new roadmap pages, Feature Sets and new Starter Assets

Major focus on Quality of Life
While there is always the desire to continuously bring new features and functionality to the whole of Unity, we wanted to make sure it is not at the expense of everyday animation workflows. We won’t have an exact number of issues/improvements/fixes addressed, until around September. As Unity grows, we believe it is crucial to stay on top of those workflows, which we often use multiple times a day.

Advancing DOTS
We have been working with productions to ensure our new high-performance, multithreaded Data-Oriented Technology Stack is delivering on our vision of a scalable, streamable, networking ready and highly customizable platform. This is foundational to our goal of future proofing Unity for many years to come. However, given the breadth of the work involved, we still don’t have any new details on the timing of when DOTS might be ready.
We do have a dedicated DOTS Animation forum to help you find and discuss these important topics.

New Roadmap Pages
We recently introduced a new landing page, to give you a high level overview of our roadmap for the Unity platform. While it is there to keep you up to date, and give you insight into our plans, there is also the facility to send us feedback. All feedback will now go directly to the appropriate Product Manager, so rather than going into a large pool of requests, there is instant visibility on the request for each PM. These then get grouped, linked, and prioritized. The more information we can gather from you. the clearer the insights we gain.

Dedicated Animation Roadmap Page
We will add any new announcements around 3D characters and Animation here, and from this page, you can also provide us with feedback on any area of importance for you.

Feature Sets
Feature Sets are new groups of packages designed to work well together. You can see Feature Sets in the Package Manager , starting with 2021.2a19. There is a Characters and Animation feature set which includes the following packages:

Starter Assets
A few days ago Emma Vig shared a blog post about the new Starter Assets. At the core of the new Starter Assets is a set of two lightweight character base controllers, adapted for third-person and first-person control.
First Person Character Controller
Third Person Character Controller

Animation Rigging Package
Just a reminder that the Animation Rigging is now a verified package for 2020 LTS
For those who might be new to Animation Rigging, there is also a dedicated forum for the subject

Kinematica
Kinematica is still on hold
You continue to use Kinematica as you do today. However, we will not be releasing new features or improvements for the product this year.

3 Likes
Is DOTS being abandoned?
2

May not directly related but what are there any progress on curve evaluation in DOTS ?

Do we have any api better than Kinematica’s curve struct? Can you move it up to unity Mathematics or something?

3

Hi.
We currently don’t have visibility on these at the moment. As I mentioned we are still working on fleshing out the roadmap, so we are still some ways off from being able to give you any concrete answers, sorry about that.

4

How about legacy Animation, is that still supported?

2 Likes
5

Hi

As the name suggests, legacy is our older system, so we are not actively working on it. However we would always take a look at any crashes.

6

It is also far superior to the animator for any animation that isn’t for a character (simpler to use and much lower overhead), hence why we have started replacing a lot of animators with the legacy animation.

Judging from your forum join date, but maybe that’s inconsequential, you’re probably haven’t been with Unity for a long time, so here’s some history: Unity’s messaging in the past has been mixed and confusing. The manual has been telling people to not use the Animation component, while in Unite talks and generally when people complain about Animator being slow for simple cases, Legacy Animation suddenly becomes “an important part of your toolset”.

So I’ll ask the question more directly: are there any short term or medium term plans to completely remove, deprecate, or further break the legacy animation system? My fear is we will complete our transition away from the Animator and Unity will announce that the legacy animation component is now deprecated.

2 Likes
7

There are no short or medium term plans to deprecate legacy. Long term, yes, once we have a suitable replacement.

2 Likes
8

Do you have any concrete plans on matching it’s performance in MonoBehaviour land?

Last time I checked, the Animator with an AnimatorController spends twice as long playing the same animation as the Animation component, and an Animator with the simplest Playable graph you can build is still slower than the Animation by quite a bit.

I know that Unity’s all “wee DOTS” these days, but it’s worrying that Playables still isn’t capable of matching the raw performance of something that’s almost as old as Unity.

2 Likes
9

We know that performance and simplicity are a key part of why legacy is still used, so any replacement will have to address these factors (including MonoBehaviour). As that is some way off, we still don’t yet have concrete details on how that might look.

1 Like
10

Is Playables still being worked on or is it all just dots these days?
If that’s all a really a long way off, it’d be great to get some performance improvements and love for playables in the meantime.

1 Like
11

I absolutely agree the legacy Animation component is an important tool that can’t be replaced by the Animator. Would hope for this feature to remain there; also: The docs say you shouldn’t use Animation for new projects, which is also confusing if there are no plans to remove it, yet it’s just much faster to work with and more performant.

1 Like
12

As mentioned, we are focusing on advancing DOTS, and making it a powerful alternative to many of the older parts of Unity. This means Playables will continue to be supported, however we expect to really concentrate only on fixing critical issues.

1 Like
13

Wasn’t playables introduced few years ago? It is already being phased out?

14

Yeah holy hell this is the wrong move.

DOTS just isn’t there. From the outside, it very much looks like it will never be. You’d be delivering 100X the amount of value to your customers if you’d instead make Playables perform, and clean up the crufty API. It is possible to create a better performing system in MonoBehaviour land - you already have it, it was made like a decade ago, and it’s called Animation.

And, look, you got to rewrite everything from scratch for Playables - at least presumably you could. If you weren’t able to create something performant there, why do you believe that you’ll be able to create something performant for DOTS?

3 Likes
15

Ah no, it is not being phased out, and it will continue to be supported.
We wish we could advance everything at once, however right now, our focus is mainly on support, and accelerating the DOTS animation vision, to help future-proof Unity. But yeah, we know this is taking longer than expected.

16

These are fair points, especially as you say, we still have not delivered on DOTS
I am not a DOTS expert, but it is a very different architecture, and everything we are working on in these stages is showing massive improvements, not just in performance, and not just for animation.
Perhaps not what you want to hear, but we truly believe there will be huge benefits once we start rolling it out.
Annnnnd for your follow up question, no I don’t have a timeframe, but hopefully more sooner than later.

17

If you stop improving the current feature set for pretty much half a decade, it’s easy to show massive improvements.

Also as a paying customer, it’s frustrating to know that you’ve all been working on future tools I do not care about, instead of the tools I am actually using and desperately need work.

Because the fact is, there isn’t a single animation tool that Unity has produced that isn’t half baked at best and that every time people complain about it, instead of getting improvements, we are always being told to wait for the next big thing.(only for the next big thing to release, and then instantly get abandoned or “supported”)

The above are not really directed at you, it’s my frustrations with Unity as a whole, so if you want to keep anything from my post it’s this:

When you say “we are putting this current thing on maintenance while we work on this future thing”, what I hear is “we are never going to stick with a feature long enough to actually make it good”.

7 Likes
18

You may not think it. but your candor is appreciated. While it might pain us, we hear you (and others) loud and clear.

5 Likes
19

Hi, sorry to post again on his one but it’s bugging me. I’m an indie dev who’s main focus is character animation so new tools are always interesting to me but I feel waiting on hold for DOTS is a bit too much.
It looks like dots is not going to be in the next LTS, which means 2023 at the earliest for an LTS release and then it’s gonna need some more battle testing for sure.
So we’re being asked to wait for potentially years without any animation updates, and then at the end of that wait we get to start again from scratch in this new system that we just have to trust is going to be all that we need?
That seems like a big ask!
There must be someone at Unity who could continue some work on playables at least. There are still a lot of us and our projects that will run on the current systems for years now.

1 Like
20

This might as well be the definition of Unity in the dictionary.

HEAR THIS:
Product management 101: Make your products reliable before moving all of your developers on to a new project. When you stop that project, keep some people on it to maintain its state and have them listen to the community to make updates to that project so that whoever’s using these products is not left to the wolves.

So put at least a few people on the Animator system. Don’t just let it die because it’s THE ONLY OPTION for users RIGHT NOW. Make it have enough features that it’s in a working state.
What kind of planning is that? Is this engine in alpha?
THE ANIMATOR IS A BASIC FEATURE THAT A LOT OF YOUR USERS ARE USING AND THAT THEY HAVE COMMITED TO.

@UnityChinny Your features take multiple years to come out. HOW DOES IT MAKE SENSE THEN TO DEPRECATE WHATS THERE IN THE MEANWHILE.
=> Ask your superiors to add people on the animator team.
=> ASK. THEM. FOR. MORE. PEOPLE. ON THE ANIMATOR TEAM.

OUR OPTIONS AS USERS ARE:
1- Wait for a new system that comes in a long time.
2- Stop using Animator, because Unity decided to stop using it. <= Great way to anger your user base.
3- Keep working with an half-assed product that has very little API access to it. <= Solution: Have more people on this. THIS SHOULD NOT BE AN OPTION.
This small team then creates a thread on the forum and asks what people want for the Animator. Clearly state that it’s only going to be a small team and that eventually something new product will replace it. But that you’re not going to let the people that are using the Animator down by just abandoning them.

As it stands, those are the 3 very terrible options you’re providing your users. Very poorly though out by whoever’s planning the releases.

Say to whoever’s planning this: EVERY TIME YOU WANT A NEW PRODUCT, YOU LEAVE SOME PEOPLE ON THE PREVIOUS SYSTEM AT LEAST UNTIL THE NEW THING IS PRODUCTION READY. DO NOT SUDDENLY ABANDON THE PREVIOUS PROJECT.
(Whether you think the previous product is “stable” or “has enough features” does not matter: Until the new product is production ready, the previous product should still have developers on it.)

THIS HAPPENS NON STOP WITH UNITY.
Please don’t come back here to say you understand frustrations, come here with the solution.

To “quickly” work on the new feature is not the solution.

Please understand this is absolutely not personal. Despite the somewhat abrasive formulation (yes I am frustrated). I do think that what I’m asking for is reasonable.