Animation 2020 - happening or not?

Actions are the same concept as Takes. There is no reason to have an Armature object. But like I was saying if Blender and the Blender way is it needs an object to apply an Action to then there is the answer for now.

I don’t have any issues sending to Unreal at all. It is seamless. What I see people struggling with is round tripping. That is the main thing. When you want to export out of a game engine and into Blender for example with rigs and animations. I don’t do this.

But in Unity you are forced to apply certain workarounds to get it to not apply a 100 scale to the root object and rotation of 90 degrees. This is not that big of a deal for the animation workflows, I use. But it gets complicated for the programmers I work with who really need clean transforms.

I figured out how to solve that. But it really took a while.

For me personally, in my use, I don’t have any huge issues with Blender in a game pipe. I figured out all of the gotcha’s already. But new people could get discouraged easily. Add to that, the fact that there is a lot of misinformation out there about the process. It is amazing how many people make tutorials about using Blender with a Game engine and I could swear they don’t know the first thing about Blender. And cause themselves all kinds of issues because of this. And others most especially who they try to advise. This is frustrating to watch, and engage in - which has happened. Funny a guy comes along to the conversation (me) been using Blender for over a decade, and maybe just maybe I don’t know what I am talking about? …lol

I always have things to learn but…

On the other hand there are some real issues people have had to solve with an addon. The root motion thing I mentioned is the biggest.

1 Like

It is mostly solvable. But it is round-tripping that is the main issue. Coming into Blender is worse than going out.

I swear I have it sorted for both Unreal and for Unity. It is not an issue at all once you know what you are doing and what things, not to do. That is the big thing. A lot of people “hear things” and assume it must “be the way”.

How bad is this? I am watching a live stream from the developers at Epic. They have developed some cool tools for Blender and use it in the animation pipe there. Very cool.

But they gloss over this one point in Blender. And I know they are dead wrong. It is interesting to watch.

And I know there are some people here who will swear by some of these things too. Want to start a heated argument…lol ?

I got better things to do. Anyway I am going to put all of this into some tutorials and that will be that.

It works if you don’t mind bones sticking out all over. Of course editing that rig would be problematic. The one on the right would be the preferred method in my opinion.

I think it’s mainly animation studios that support blender (tangent animation, cube creative, ubisoft animation studio…) BF produce short films, and I think animation is where Blender is aimed at for now.
Is that a good or a bad choice , it’s quite debatable as all these bad choice and bad designs lead to the software being more and more adopted and supported by companies.
BF must set their priorities because they can’t please everyone or support every workflow (even if they try in the long run). The opensource dynamics is, that BF has no clients to please , it’s up to people/studios to add feature they need or invest their own time developing blender instead of paying licences. BF is here to maintain the software, pay developers so big refactors can be done (like eevee integration, 2.8* ect…).

Of course there is donations, but it’s paying for blender future developments, without having much words to say on said development. And it’s the same with commercial softwares. If at some point SideFX want to get rid of everything that is nodal in Houdini it’s up to them. As a customer you can only decide to not use it or pay for it anymore.

It works if you don’t mind bones sticking out all over. Of course editing that rig would be problematic. The one on the right would be the preferred method in my opinion.

I do mind bones sticking all over. How do you imagine putting together complex controls onto something like that? And no, the one on the right is not the preferred method, because it reorients the original rig, which is unacceptable.

Why not Maya Indie by the way?

" SORRY, MAYA INDIE IS NOT AVAILABLE IN YOUR REGION." is what I see.

For sure. I agree. That is what I was mentioning about editing the rig, that would not work. That said, I would not have any reason to be importing to Blender with fbx for this kind of stuff. The rig on the right, looks and plays right. Beyond that, I would have to investigate more.

In general though I would not at all recommend a workflow that includes trying to bring in fbx rigs that you have to work with in a meaningful way in Blender. At least not in a direct way. There are other ways I would recommend using a round trip workflow with Blender.

This is an issue for some though, that I fully acknowledge. But I am good with it as it is for my needs. But I have seen people figure out workflows and write plugins and other apps for round tripping with Blender.

But Blender vanilla is not the way to go as far as I know. That is for round tripping anyway.

At any rate, there are a lot more issues, more fundamental issues with animating in Blender that need addressing as well.

You don’t get my point. I’m talking about Blender Foundation omitting an industry, where Blender lives for real. And by for real I mean some people using Blender in companies like Techland (I wouldn’t be surprised, if there were Blender users in CDPR too). There is demand for Blender in games, and there is also demand for real time engines in animation. Meanwhile BF focuses on the opinions of a couple of obscure animation studios nobody even heard of.

I’m not the Blender Foundation, but I guess in their opinion it’s not that they are omitting an industry, it’s the industry omitting blender’s development. It’s up to these big game’s studio to pay devs to improve code source and provide patches to BF.
They don’t focuses on opinion of other (animation) studios, they just accept code they provide and try to work hand in hand. Also these studios provide funding to blender’s development, but that doesn’t mean these studios have something to decide on which features should be prioritized over another.

A commercial software needs more users to grow, for open source software it’s more about contributors. Blender could be broadly used in medical industry, that wouldn’t mean they will push development in that direction to attract more users in that field, they IMO don’t care that much even if of course they try to please as much users as they can.

I agree that it’s quite a complicate issue to solve, because if nothing is done game industry won’t use blender therefore they won’t make any development to push blender forward.
It was the same situation with animation industry a few years ago and now it’s changing, maybe that will be the same with game industry.

Once again, many cool features blender’s got comes from outside contributions (UDIM, GreasePencil, Mantaflow, Cryptomatte…) BF don’t push their own development toward that and concentrate more on big picture stuff.

Maya/Max indie are being rolled out worlwide this year. Also, you say Houdini, but it actualky works out more expensive than Maya(unless you’re using free indie of course)

Have you looked into the ARP addon? UE4/Unity export retarget presets, mocap, automatic control rig and sticky out bone hiding, same for the auto facial rig. It only costs 30 quid.

1 Like

I watched one of their old streams about a year ago, regarding the Blender-UE workflow. They did talk a bit about the pains regarding proprietary FBX and import/export. It sounded like some Blender heads at Epic wanted to affect some changes, but I don’t know if anything is becoming of it.

They created a live link addon. Not released yet.

Well, I don’t know what they did on their side with Unreal. But I can say this much. Blender export to Unreal Engine with basic defaults is real smooth. And works surprisingly well with both Rigify and Blendrig 5.

What I did get from this recent stream is that they emphasized that the tool they are working on (live link mentioned above) does nothing more than tap into the vanilla .fbx from Blender. And the tool is pretty amazing.

The main issue with Blender workflow I can see is the Root Bone issue. This will come into play if you are planning to use this feature for characters in a game. And usually you do. Because you need the root to travel with the character so that the bounds goes with the character and does not stay in place. When you import from Blender it creates an arbitrary root bone. And apparently this does not work. But there is an addon which can remove it on export. I can’t remember the name of it unfortunately.

But for me, since I use Unreal mainly for rendering right now, I don’t have to worry about that. To keep the character in the view during rendering, I just scale up the bounds.

So I’m not sure the root bone issue actually is an issue any more? I just did a quick test with a single bone rig on a Suzanne, and when imported into Unreal the skeleton only had that bone and the leaf bone, it didn’t have a separate ‘Armature’ bone at the root. I also did an animation with that single bone and it was able to extract root motion from it properly.

Maya/Max indie are being rolled out worlwide this year. Also, you say Houdini, but it actualky works out more expensive than Maya(unless you’re using free indie of course)

Unfortunately, this doesn’t seem to be the case, look at ValentinN84’s second post in this thread.

Have you looked into the ARP addon? UE4/Unity export retarget presets, mocap, automatic control rig and sticky out bone hiding, same for the auto facial rig. It only costs 30 quid.

Yes, I’ve seen it, however I can’t find any information about being able to wrap controls around an existing skeleton with it.

Not 100% sure about Maya(should be the same though), but Max is definitely rolling out worldwide in the future. The when is the question mark.

During the Q&As following last week’s livestream to mark the release of [3ds Max 2021], Chip Weatherman, Autodesk’s chief product owner for 3ds Max, said that Autodesk “will be putting it [3ds Max Indie] out worldwide at some point”.

You can hear the relevant section at 01:58:40 in the recording of the livestream.

Yeah, “at some point” I can probably be quite efficient with Houdini already :smiley: For $250/year it looks like it’s worth a try.

Btw. do you know something about ARP being able to drive an existing bone structure? Does it have to create its own to work?

Yeah, Adesk have made one marketing/corporate blunder after another with this. They could have regained much needed respect from the community by just releasing Indie worldwide first time and been transparent about it.

I believe it is optimised for the in-built auto-rig setup, but I’m not sure if it can retarget custom rigs. I haven’t tested it rigorously enough yet.

Well I don’t want to derail this thread into a support thread. Just making a point. And no it isn’t fixed now. It is still an issue.

If you want to “monkey” around with this, lets do it in a proper support thread.

And I would suggest before continuing that you try something more than a few bones and the monkey. I will explain why in the other thread and give more examples if you like.

Lots of free rigs and such to play with, if you are looking to go the fast and easy way to test.

For now, I am good either way, and will be doing a tutorial on it at some point anyway within a month or so.

Okay, playing around a bit more, and per this thread: it’s an issue if the armature isn’t named ‘Armature.’

If it IS, Unreal ignores the bone that Blender adds on import.

Right. And so the only way to solve this for anything other than child’s play is to use the addon here:

It does seem to get rid of the root bone for any kind of rig. But leaving the armature name as Armature also causes other issues. No point in going into here.