3D/2D rigged characters in Visionaire: giving up

Today I thought I had found an alternative solution to Spine-based animations. Simply stated, I don't want to spend $300 when I can do all the animations of 2d characters in Blender, for example.

Blender has this nice cut-out character animation addon which allows you to quickly build and animate 2d characters, similarly to Spine and Dragonbones, but with far better animation controls. And I am used to animating in Blender, so I my thinking was that I ought to be able to import those animations in Visionaire via the 3d model import option.

8 hours later, spending an entire day on figuring out a way to get this to work, I am giving up.

- The visionaire manual is of no help whatsoever. Visionaire users on the forum generally don't use the 3d options, because apparently almost no-one gets it to behave properly.

- According to the manual, Visionaire uses the Open Asset Import Library to import 3d files. It advises the Visionaire user to use the viewer to check if an object will load up correctly.

- The problem is, that library hasn't been updated in quite some time: since 2010, I believe.

- The library is supposed to load up *blend, Collada DAE files. But these formats have evolved somewhat, and no matter what I tried I couldn't get it to work. Either textures would be missing, parts would be missing, animations wouldn't work properly, and so on, and so forth. I also tried DirectX, but to no avail. Something is always wrong. And I tried many, many settings and combinations. I even downloaded various other tools to see if I could export as FBX, then convert to DAE, etcetera, and so on, and so forth.

Basically, it doesn't work. And no-one seems to have a good idea when it will work. And I am not stupid. I am pretty well versed in game development, including complex asset pipelines. But this is SO frustrating.

So I give up. I really want to import 2d animated cut-out characters in Visionaire, but I refuse to go with Spine (which, as far as I can tell from others' experiences isn't supported that well either anyway). Last year I tried Dragonbones, but there were problems with that import as well, and besides, we are not allowed to use the Spine export of Dragonbones without a Spine license. And Dragonbones itself has problems as well (the license, having to be online and logged in to export).

So that's it for me, I'm afraid. I'll keep an eye out for the next version, but for my next project I'm switching to something else.

So frustrating, because Visionaire is SO close to being exactly what I need. Damn. Damn. Damn.

Gutted.

Here's a basic summary of AkcayKaraazmak's process that he used to create & import 3D characters into Visionaire Studio for his game: "The Dark Inside Me". I have no idea if he still uses that process or figured out a better way - the only thing I can tell you is that the format should be .x.

Unfortunately I can't be any more help than that as I don't care for the 3D side of the engine & have never gone beyond importing a test model that someone provided me with.

As for DragonBones, I believe you only need the spine essential license to be able to export spine character models & use them commercially regardless of which features you used in DragonBones. I think the different license models just differ in what features are available in Spine itself - I'm not 100% on that, but you could probably find out with a bit of snooping around on google.

Thank you, and yes: I read that page early on. It's a very convoluted workflow, and the first part ("workflow") can be done much more efficiently in one application such as Blender. The use of Mari and zBrush are complete overkill for what I attempted to do today. And .X is not the only format that (semi) works: Collada DAE files are also accepted, but the library is too glitchy and hasn't seen any updates in quite some time, limiting its use.

As I see it, Visionaire's 3d import options are too limited, with too many workflow issues. And that is aside from the limited quality that is achievable by Visionaire's 3d built-in renderer, so...

I did succeed in partly importing things, and simple object did work, but even those resulted in a less than ideal final quality. Visionaire is great for 2d stuff, and if you are willing to spend more money on Spine, than I guess it's workable.

The level of control over a 3d object once it is loaded up in Visionaire is quite limited: camera control is very rudimentary, for one thing.

Another issue that popped up: items cannot use 3d files. Only Spine files.

Anyway, I took a break last year from Visionaire and adventure making, and was hoping at least some of these things would improve. What I am really interested in doing is something akin to Broken Age in terms of camera and character animation control.

That is impossible to pull off in Visionaire. Or at least, almost impossible without some hard-core workarounds. The engine needs much better 3d/2.5d control and features to pull something off even remotely close in the same league.

So for now I'll be using something else for this project I've started work on, and return when I need a more traditional 2d adventure engine again.

Hi Simon, please check out Broken Age, and the 2d characters as well as the dynamic camera that behaves more like a traditional cartoon/film. Compare that with Pillars of the Earth, and you'll realize how dynamic and freeflowing Broken Age is compared to the very static looking Daedalic's camera and scene movement.

Today I tried to import the sample character of the Blender COA tools add-on found here:

The sample blend file is stip.blend in the sample file directory.

I couldn't get it to work. I tried Collada (old original Blender exporter), but things are missing. I tried the newer Godot Engine collada exporter for Blender, but nothing shows up. (get the new Collada exporter here: https://godotengine.org/download/windows)

Basically, I just want to be able to import flat animated cut-out characters with multiple animations, similar to Broken Age, and preferably have things such as lip sync done more or less automatically (this last thing is probably not possible, but I can deal with that in Blender if need be).

Having better camera control would be nice too. I wasn't aware of the shader toolkit, and will investigate further - but a quick glance tells me that every camera move must be coded. Ideally Visionaire would allow for real controllable 2d animateable camera objects. It would be nice if the camera could work in layers for parallax effects. Again, check out Broken Age.

The other thing I'd say Visionaire would truly benefit from, is an animation timeline to animate and tween anything (characters, background items, objects, cameras, and so on) and fire events, similar to Godot's animation timeline, or Unity's timeline - even simple rigged 2d characters. But now I am dreaming, I suppose.

I believe Simon intends to create a built in sprite-part animation feature at some point, but when that will be? who knows...

You talk about the camera in Broken Age, but I'm betting the programmers of Broken Age had to manually code in those camera movements themselves. It's probably the same with most engines?

Broken Age was apparently developed with an engine called Moai. I've never heard of it, but based on what I found out about it via google & the github page, it sounds like a framework based engine only - like Löve2D & Ren'Py.

As Simon said (aisus), you should be able to use shaders to create some nice dynamic camera movements with panning, zooming & rotation; & all should allow easing too. There are some camera functions Simon added into the shader toolkit, but you would probably be better off coding your own camera stuff if you know how to write glsl - if you do, then you are much better off than I am as I have zero clue where to start.

Moai is more than just a framework-based engine - or at least, Double Fine extended it with their own narrative game building tools. Broken Age consists of 5 hours of cutscenes and dialogue, and scripting all that to a high quality would have taken ages and mentally exhausting.

What if I told you that those five hours of scripted cutscenes were done by only one animator? ONE! With a tremendous amount of dialogue.

So how did they accomplish this?

1) the characters and various other animated elements in the scenes are segmented objects, animated in an external app (they used Maya at the time)

2) these canned animations are then imported their custom cutscene editor and combined in the scene. This is all visual, and done with a nice timeline which includes full camera control and allows for layering and parallax.

This cutscene editor dopesheet timeline is shown @27 minutes in this video:https://www.youtube.com/watch?v=iWEVY4ujyI4

This allows for real-time previewing and editing of cutscenes and dialogues.

3) another tool, the Animation Toolbox, was used to organize sets of animations for various joints in the characters (allowing for facial expressions, automated lip sync, etc.). Lip sync was automated.

Anyway, the point is that no scripting or coding was used to animate and control these cutscenes, and the same tool could be used to control animations of characters in the playable scenes as well.

The way I see it, it would be Visionaire's next step forward if it would allow us to:

1) create animated elements (flat 2d rigged characters, animated background elements and items) in Blender, Spine, any 3d app

2) export to Collada

3) import those elements and sub animations into Visionaire for characters, backgrounds, and items.

4) Visionaire would include a scene-based cutscene editor dope sheet timeline that allows for visual camera control, NPC movement, simple tween-based animation of elements, placement of scene components, and simple choice of various imported elements to cut and paste a scene together, including dialogue trees. Allow for simple parallax layering and placement. Quick placement of objects and characters, other elements on various layers.

5) multiple timelines could be added to a scene, and timelines can consist of nested timelines.

6) easy visual selection of sub-animations.

Scene editor timelines are nothing new, really. Most major game engines have these built-in now: Unity, Godot Engine, Unreal, Flash, and so forth.

Visionaire, a narrative game building tool, has none of this, and still builds on the simple frame-based animation that were introduced in the 80s and the beginning of the nineties.

PS the entire Broken Age development documentary is free to watch here:

Really interesting up and close documentary about the making of a modern point-and-click adventure game, and a nice insight into the minds of two of the most seasoned classic adventure game designers (of Lucasarts fame!) Tim schafer and Ron Gilbert.

Interesting, though that's way too much documentary for me to watch.

The popping camera he mentioned, you can do that easily with the shader toolkit camera functions. You should also be able to easily pan, zoom, rotate & follow too - you may need to code a safeguard in to prevent the camera from being able to pan past the scene background; unless Simon already sorted that out?

Lip sync is easy too as Simon implemented support for Rhubarb Lip Sync which uses the hanna-barbera mouth shape system, which compromises of mouth shapes A-H & X for the idle mouth shape. You can also create custom tsv lip sync files manually if you want & assign animation frame numbers instead of the mouth shape letters in case you want to add special gestures or mood based talking animations.

I too would really like to see an internal sprite part/bone rigging animation system. If VS had one built in then it would be perfect - I doubt it would be anything as complex as Spine with mesh warping & all that, but even a basic bone rigging & keyframe animation system with the ability to swap out sprite parts at run time would likely be more than enough, I think? Here's hoping it gets implemented at some point. In the meantime, maybe Simon could look into implementing actual support for DragonBones models for people that don't have the money to purchase Spine - I don't have that kind of money either.

Wasn't aware of that lip-sync option! Nice.

I disagree having to code camera movements. In particular when doing narrative games, intuitive camera controls are quite important, and coding these movements will lead to rather mechanical looking results - unless one spends so much time to fine-tune, that it just isn't feasible within a reasonable time limit.

On the other hand, a timeline with one or multiple camera support solves this immediately. You see what you get, instant feedback. And with visual keying it becomes second nature to create nice looking shots. But coding? No. And you can tell from one look at Pillars of the Earth - camera control feels very static and unnatural. Coding doesn't allow for quick experimenting either. You need this option in the scene editor, and with realtime feedback for the best results.

I agree having an internal sprite bone rigging animation system in Visionaire would solve many issues, especially with us having to rely so much on external tools for even the most basic animations.

Add a (multi)camera system with parallax layering, a 2d cutout character/item/background system, and hook up that lip sync part, timelines to animate all of that nicely with curves, and Visionaire would be up to date, and users would be able to churn out some radical results.

One can dream. While I dream away, I have been investigating other options today. So far it seems I will have to (partly) write my own adventure framework, though, or re-use others' frameworks. Godot has Escoria, for example. I've been testing Unity, but I really REALLY dislike Unity: always have. So that one's out. Blender has a new built-in game engine called Armory, which looks quite promising so far. No exporting of assets, everything is done in Blender, and there's even visual node-based scripting.

But you know, I would stick with Visionaire if only it would be able to import those 2d cutout animated characters straight out of Blender. I'm going to give it one last try tonight.

Having built in camera controls does sound interesting. I didn't watch enough of the pillars game to get a good idea of how the camera works. You could program in some functions for different kind of camera techniques maybe to reduce the amount of work. You should technically be able to create smooth looking camera movements if you get the x,y timings & easing right - it's likely incorrect timing that makes the camera looks stiff/static.

With a loop, you could probably create quite a nice handheld camera feel. You could also probably simulate depth of field using blur while zooming in/out.

Parallax layering is already possible. Check the effects tab for scene objects. There's 2 options that let you control the scroll factor percentage.

VS has a builtin plugin system where you could create some new Actions which could handle that. 

As VS is build i doubt that the top-to-bottom order to create cutscenes will change, but all in all you can handle similar stuff like Doublefine did in their cutscene editor.

Lipsync will be done automatically for your current speaking character as far as a lipsync file is available (.tsv). 

Doing gestures while talking may need tweeking of these files as AFRLme already mentioned. 

Parallax stuff alreaey works and zooming with the shader toolkit can do a lot of stuff,too. Also for these functions you could probably create a plugin to make new actions which control the lua part and the user just selects either characters /objects/coordinates and a zoom level in that action...