10 replies to this topic

[Invertex]

Unfortunately, Unity 4 is not very good at handling the huge and complex lighting that is currently in this game, as we've all seen. And the lack of multi-threading puts a serious damper on performance as well.

So I'm curious if Red Thread has plans to transition to Unity 5, which has implemented multi-threading in its job scheduler, and has significant advancements to its higher quality 3D rendering features.

[Aj17]

I don't see them doing that, unless they plan on re-releasing Book One on the same engine..wouldn't make sense. They're better just optimizing the game engine better as they go.

[ShadowNate]

If by next release you mean one of the next chapters, and considering that those will come as patches to the first book, then they would pretty much have to port everything to the new Unity.

 

I am not sure how easy this is. They already are using Unity (I am also unsure about the actual version/builf), but typically moving to a newer major version of the same engine is not as an easy task as someone would expect (I am thinking the Source engine here). 

 

And porting to a new engine (or I guess a new version of an engine) will be time and resource consuming in any case and also could break some custom code that RTG used to extend the particular Unity version they were using (they already said that they were doing that, pushing the engine to its limits)

 

It is generally not advised for a game to do this mid-development. I think when such thing is ventured, it seems to work better when it is done months/years after final release, whereas it works quite catastrophically when done mid-development. 

 

Someone from RTG would be more appropriate to answer the question, but my guess would be "no plans for that at the moment". 

 

Edit: It seems that according to the official Unity3d site, the Unity 5 engine has not yet been officially released and that additionally someone would have to buy a new/upgrade license to work on it (they mention something like $600 as a limited offer). But, transitioning to a to be released engine? This guarantees a definite no in my opinion. 

[Jemobulas]

Agreed. RTG started with what they decided to be best (or with what they were limited to with) & should finish it like that. Changing engine in mid-development will lead to too many problems & will take too much of their attention away from the actual game content. They've already placed the foundations of DFC & now they're building up. I don't see it right to start switching the foundations at that point. Plus as ShadowNate said - "to be released engine".

[Invertex]

Well, Unity 5 is in beta, and it's generally not hard for a studio to get access to that. But I didn't know the future chapters were coming as patches and not separate clients; that would change things a little bit. And with it still being in beta, I guess it would be a bit unrealistic to expect (though studios coming out with AAA games on unreleased new versions of Unreal is definitely not unheard of, but that's irrelevant)

 

As for actual porting difficulty, well, it definitely is hard to say without knowing what kind of code RTG has created, but in general, Unity does make it very easy to move your projects up into new versions, especially with how well its object oriented scripting works. If new functions are created to replace old ones, the old ones are still left there working, you are given a warning about them being deprecated and told what replaces it, as well if it doesn't work at all. This is partly a result of the highly cross-platform nature of Unity, most of the changes made to the engine are done under the hood so to speak, as the game code is more generalized and reinterpreted on compile for each target platform. (also, I wouldn't really call that "changing engine", it's the same game engine, just an updated version, nearly all the code should still work).

 

I do hope they can further optimize though. 

 

Also, something seems funky with the SSAO implementation, it has a pretty huge performance hit even on low, and the differences in visual quality between all 3 levels is almost imperceptible.

[Tina]

Additionally, RTG have already made the basics for all the books, so they are able to "play" them through (they had done this before the decision to go episodic). It's mostly pre-alpha, and grayboxing, if I've understood it correctly, but all of the basics for the next four books have already been made in the current engine.

 

I would think that  RTG probably will use Unity 4 on the rest of Dreamfall Chapters (to not add more time between episodes etc), but maybe go over to Unity 5 for Draugen and any other upcoming games.

[Leonard Zelig]

Couldn't the modify the Unity 4 engine , so that they can benefit from some Unity 5 enhancements? Why didn't they use the Unreal Engine 3, wouldn't that have been a lot easier?

[agirlnamedbob]

As far as modifying the engine, I'm guessing that's a lot more trouble than it's worth. I could be wrong. I don't know exactly how that works, but I would imagine it wouldn't be easy and would require the game itself to be retooled in ways to take advantage of those enhancements, assuming it's even possible. 

 

As far as why Unity vs. Unreal... The Unreal Engine is great, but the Unity Engine is was a better deal for them and it seemed to be a fit for what they needed. (At least at the time of the Kickstarter. I heard that Unreal changed some of their licensing stuff, but I don't know all the details.) 

 

I'm also not sure if it would have been "easier" per se. Maybe it wouldn't have had some of these specific issues, but it may have had other issues. It's also easier in hindsight to be like "Yeah, we're really pushing this engine to the limits." And again...every engine has it's pros and cons. 

[Philip]

Earlier this year, a couple (?) of game engines changed their licensing model from something which only appeals to big studio to something much more affordable. Wikipedia says about the Unreal Engine:

 

On March 19, 2014, at the 2014 Game Developers Conference, Epic Games opened Unreal Engine 4 to the world, releasing all of its leading-edge tools, features and complete C++ source code to the development community through a new subscription model. Anyone can sign up for UE4 for PC, Mac, iOS and Android by paying $19 per month, plus 5% of gross revenue resulting from any commercial products built using UE4.^[79][80] CEO and founder of Epic Games, Tim Sweeney, said that the new business model is a reflection of changes in the industry. Epic Games has traditionally made its Unreal Engine available to large AAA game development teams at a cost of millions of dollars. But as the industry has evolved, Epic has had to "really rethink our whole business as to how we make the engine available to teams."

 

So apart from all other things that might have played a role when deciding about the engine, the Unreal enging just wasn't available for RTG at the time they started developing. Also keep in mind that there were a couple of scenes already shown in the kickstarter pitch video, so they had to start working with the engine even before the kickstarter went public.

[Flo]

Switching engine revisions mid-development only makes sense if doing so saves them time or resources on upcoming Books.

[Roxie]

I'm finding I can be much more productive in Unreal 4, and it's codebase is, well, a work of art in the areas I've been playing.  Then again, Unity is quite capable as well.

 

As far as unity 5 upgrades, well, highly unrealistic to make those changes.  A significant portion of the assets will have already been created, including AIs, shaders, camera behaviors, and controls which may require code updates.  It might be reasonable to port some of the performance related stuff, especially if it's 'low hanging fruit'.