The general software thread

Software company Autodesk is launching its own game engine

If you play video games, you’ve probably seen the effects of Autodesk’s work. Autodesk owns the Maya and 3ds Max modeling software, the Scaleform rendering engine, and a variety of animation and texture-making tools. But the meat of the game will (usually) be made in an engine like Unity or Unreal, both of which have spent years courting developers with low prices and approachable tools. On August 19th, though, they’ll see a new competitor: Autodesk’s own engine, called Stingray.

Stingray was first announced in March, but it’s technically existed for several years as an engine called Bitsquid, which Autodesk acquired in 2014. Bitsquid is behind, among a handful of other games, chaotic cooperative shooter Helldivers. The core of Bitsquid, says Autodesk’s head of games Frank Delise, isn’t changing much. But among other things, it’s using a different interface and is more tightly integrated with the rest of Autodesk’s design tools. “We really liked that the engine was separated from the user interface, so we can make very user-specific versions of the technology,” says Delise. It’s hard to tell exactly how good Stingray games will look — the major test for any engine — but the screenshots don’t look too bad.

Autodesk bought the company that made its engine last year

The theory is that while it may not be easy to learn the finer workings of Stingray, anyone on a game development team — like an artist, animator, or texture designer — should be able to preview their work simply and in near-real time, especially when Autodesk owns every step of the process. Make or modify a character in Maya, for example, and it can (in a controlled demo, at least) show up almost instantly in the Stingray editor. There’s also a visual, drag-and-drop programming tool to perform simple tasks without learning the Lua scripting language that Stingray uses.

“If you’re a AAA developer, what you do is buy 3D tools, and then buy other tools that you need or make them in-house, and then you either buy a game engine or you build your own engine, and then you get your teams to customize this massive pipeline around it,” says Adams. “What we’re trying to do is create more of that out of the box.” Autodesk is also promoting the Creative Market, which it acquired last year, as a place for developers to buy and sell models or other game assets.

“Free is always such a weird thing.”

One of the most obvious differences between Stingray and its competition is that the general public can’t get it for free. Like its other products, Autodesk sells subscriptions to Stingray; it costs $30 a month for “basic” support, with discounts for longer commitments and a free option for students. Unity, by contrast, offers its basic editor and engine for free, with a $75 monthly subscription for team licenses and additional features. Epic charges a 5 percent royalty for Unreal Engine 4 games that make more than $3,000 per quarter. “Free is always such a weird thing. I like this way because it’s more basic,” says Adams. “You don’t have to worry about suddenly owing a whole lot of money to somebody else once your game takes off.” He notes that Stingray will be included with a subscription to the lightweight Maya LT later this year. This still aims it at developers that are closer to the professional side of things, not people learning how to make games for the first time.

Easy access to engines like Unity helped make high-gloss game design accessible to more developers, lowering the bar of entry for our current boom of indie and mobile games. The actual staying power of Stingray has yet to be seen, but if nothing else, it’s introducing more competition into the field.

3 Likes

That could backfire if Unity and Epic start promoting MODO and Blender because of it.

Backfire only for Autodesk of course, which is always a good thing :smile:

1 Like

This is the “Bitsquid” engine back in 2013:

This is Autodesk Stingray now:

2 Likes

Autodesk is a cancer.

9 Likes

Looks like a solid attempt to enter the market. I think their PR effort is going to be very important to them here in the beginning.

One thing of note: as far as I can tell, they aren’t launching with a Mac OS editor or platform support. I at least couldn’t find reference to it on their site.

Very interesting, for it to be any sort of success it’s really going to depend on age old issues. Does it have decent documentation (hopefully better than Maya does)?

Will it have plenty of tutorials? How extensible is Lua and how good is the debugger? What Navmesh / AI features does it have? Does it have decent occlusion culling systems? Does it have a decent animation state tree? What’s the terrain system like? How simple is the UI to use? Does it have speedtree integration? Does it have a decent particle system, does it have a material editor? Does it support substances? How flexible is the engine to modify? Etc. Etc.

It looks quite pretty so I gather they have a decent rendering pipeline? Although it has some pretty stiff competition, so if they manage to integrate Maya LT’s workflow pipeline directly whilst being able to do what Unreal can (for the most part) they’d be onto a winner for indie’s.

The perfect solution is a more lightweight version of UE4 (that doesn’t require silly hardware) and a scripting API, pretty much a mix of Unity and UE along with their tool integrations. That would be IMO perfect.!

End of the day, we finally have some great choices. CE (which uses LUA too for some things) is getting better, UE4 is great and so is Unity. It’d better be good :).

I’m so looking towards paying for it

…NOT!

1 Like

Wonder where it will be in 2 years.
If they integrate Maya within the game engine, it is bye bye to every other game engine for me.

2 Likes

For all you know it could be a great engine that improves productivity 10 fold, also increase your chance of success which makes that $30.00 totally worth it. Don’t knock it until you’ve tried it…

Also Maya LT is IMO pretty damn good, so if you get a sub with it. Then that’s pretty cool.!

2 Likes

Woa, I don’t know about performance or stability but in terms of features it blow all the competition away already.
Love the live link to devices!

I was expecting Autodesk to lean more heavily on integration with their middleware, but I’m not actually sure it does significantly more in that direction yet. Hard to tell. Are there specific features you have in mind here?

One potential weakness with the ‘glue middleware together’ approach could be that they’re relying on the strength of middleware code, written by several different teams, without knowledge of this engine. If the documentation often says “See X product’s documentation for more”, that could be trouble.

But if they’re been silently updating these different packages to be ready for this engine over the last few years, that could also be a strength. I wonder how hierarchical things are over at Autodesk regarding integration into this engine.

Couldn’t imagine coding a whole game in Lua, no thanks.
Too little info to draw conclusions on anything else, but the limited marketing material I’ve seen doesn’t make me think of games, more archvis.
The integration with Autodesk products is good, but would prefer them just to fix their damn FBX exporters!

1 Like

Can’t really tell much about it at this point.

I can say from first hand experience that Autodesk is indeed not UT or Epic. Autodesk has tunnel vision when it comes to its products. They are a bloated, fragmented monstrosity that does what they want irregardless of user input. If you think UT or Epic has poor support then you most certainly are not interested in using an Autodesk product.

I seriously doubt they will be able to deliver a formidable product that grows in the long term that competes with its peers… But, I guess we’ll see.

5 Likes

So true, the number of redundant, legacy features that have never been removed that clog up 3DS Max astound me.

1 Like

All I want from this is for Unity and Unreal to take action and look for ways to improve the syncing/linking of the game engine with the digital content creation tools. Or make a deal with MODO.

This article seems to explains things better (also kind of contradicts slightly the one the op posted): http://www.engadget.com/2015/08/03/autodesk-stingray/

There is also a visual scripting tool and it does have a material editor, which is nice.

@Frpmta

I’d settle for a happy medium between Unreal and Unity, Unreal is a great engine but it’s clunky (options within options, disparate systems from editor) and slow as hell to do anything (shaders, code (not BP), lighting tweaks, import) plus it doesn’t scale well… Also I’d take Beast over Lightmass any day…

Unity could do with a polish and some additional tools, I get that they’re working on it but I wouldn’t hold my breath on it being right for quite a time to come.

I’m excited by the news TBH.

More video here, the video above doesn’t explain much:
http://www.autodesk.com/stingrayengine
They are using their middleware:Beast(light baking), HumanIK(animation retargeting and IK ), Scaleform UI.

The cool part is that it’s all immediate, Live link to devices, live coding(make changes to the code in real time without stopping), live shaders. Zero time loss.

2 Likes

https://vimeo.com/134982672

3 Likes

The content creation workflow is actually pretty cool, Unity should add an option to “send” .fbx files to a user set editing software.

And streamline importing, maybe even creating some plugins for the popular modeling suites.

1 Like

Thanks! I’ve been geeking out on that site as well. Not hating here, but a few caveats/thoughts:

  • They do take advantage of their middleware, but I’m not sure how their integration is significantly tighter here than in some other engines- the immediate importing is something we’ll all be familiar with, as well as Beast and Scaleform (if you’ve used their plugin). It is middleware after all.

  • Live link (to hardware devices) does look cool as a short term debugging tool, and maybe it’s more than I’m thinking, but I’d be cautious about using it in place of real deployment to test for performance and stability. For fast, iterative level design, I definitely see a boon there, like Unity’s previous iOS streaming tools.

  • any live updates to coding sound like they would be limited to lua scripting. Any updates to non-interpreted code would have to be recompiled, which would restart the game by necessity (unless they went way out of their way to separate their code into multiple assemblies, and hot load small, quarantined bits). That’s fine if you’re interested in using an interpreted language (and lua’s a good one), but there are some significant drawbacks to using it as your primary game logic language.

  • The shader integration is one of the coolest things I’ve seen here, if only because it uses the exact same interface and node library (probably) as in their 3d packages. The idea that an artist can set up a shader in Max/Maya, then perfect it in engine without needing to rebuild things is awesome. One reality check here is that those 3d packages don’t include the same rendering engine as Stingray (yet?), so artists probably won’t be perfecting their shaders in Max, then expecting them to look exactly the same in SR. There will still be some polish going one, for sure.

  • This is more of a question: The latest Beast integration- does it have dynamic features these days, or is it primarily still a lightmap/backing solution?


Anyway, I do think there’s a lot of future potential for Autodesk. If it can get this engine off the ground, maybe it can leverage it’s various middleware teams to add features specifically for SR. But that’s a big if- It would imply a very efficient level of collaboration between teams. It would also require a lot of central control to be held by the Stingray team, which could turn Autodesk primarily into a game engine company, over time.

1 Like