Larian Studios Forums Baldur's Gate III Baldur's Gate III - Suggestions & Feedback Unreal Engine 5 disccusion

Moving forward, should Larian use Unreal engine 5 for development BGIV, D:OSIII ?

Reasons:
development costs, QOL assurance, multiple teams linked together, collaborate work with 3rd party studios like Anshar, faster project development, faster bug detection and resolution ...

Changing engine this close to release?

Sounds like recipe to disaster.

Is unreal engine even a good choice for turn based RPGs? Isn't a custom engine alike the one currently used more adapted to those specific needs?

I've written for BGIV, D:OSIII, not BGIII.

You can adapt Unreal as you wish. Reasons for using Unreal are more on QOL, productions costs and technical side, in my opinion.

I do not think it is the right question, because not all engines are capable to achieve what the developers need or want for their games, which is why they create their own. There are also other reasons too.

But no, I do not think they should.

That is untrue, for exemple making a voxel game on unreal is at best a miserable experience. UE isn't a silver bullet.

Oh!

My bad im sorry

I am not a dev and I do not know where UE5 is at in terms of handling something like a turn based CRPG that has the complexity of a Larian game.

That being said if they can get a CDPR type of deal where the Epic engineers are going to help them put it all together, it wouldn't take horribly long to do and would ultimately permit them to add more systems to their game while saving time? I don't see why not - if the time saved in development can offset the cost of working on their own engine.

But the questions around that are all pretty big and nuanced... and I am guessing the vast majority of people who post here have no answers for them. But maybe there are some devs lurking around with thoughts.

If it weren't true, then companies would use a single engine for all their games and they would all perform the same. But they don't, which is why they use different engines. Just like you don't use one type of canvas to paint every painting, you use a specific type of canvas for the specific type of painting you wish to paint. Same with games.

When a game engine cannot do or achieve what is required, wanted or needed for a project, they develop one that can or move to one that can. Because games are an ever-inovating breakthrough technology and each one requires an engine that can fulfill the vision of the game's designer within the time they want it in. And it's not just about gameplay reasons, there are so many reasons inside and outside of the scope of the game on why a company would use an engine.

For example with Bethesda games;

• Dishonored 1 using Unreal Engine 3.
• Dishonored 2 using Void Engine.
• Doom 2016 using id Tech 6.
• Doom Eternal using id Tech 7.
• Fallout 3 using Gamebryo Engine.
• Fallout 4 using Creation Engine.
• Starfield using Creation Engine (even though people are by now fed up with it and begging Bethesda to kill this engine off as it is outdated and obsolete)

As for Unreal Engine, it is a terrible engine that is only good for eye candy, at the cost of terrible performance as it's notorious for being a stutterfest whenever it has to load in assets. And Unreal Engine games are for the most part never in-depth complex games, mostly action-adventure games. Unreal Engine is for companies who don't know how to code and optimize to save their life, so they use that engine and their games always suffer as a result. Just because it's popular, doesn't mean it's good. Darksiders 3 (UE4), Star Wars Jedi Fallen Order (UE4), Batman Arkham Knight (UE3) are great games, but also some of the worst performing games due to their stuttering and they're coming from 3 different companies.

So no, I do not think they should use Unreal Engine nor do I think they need to, if their current engine is achieving everything they need it to.

It depends what functionality it has out of the box. Traditionally Unreal is being used for First Person Games, and from what I understand that is what it is best geared for. Firaxis created both XCOM1&2 using Unreal, so it is possible, but according to Jake S. they did a lot of work to make it happen. That was a reason why they didn't update to newer version of Unreal for XCOM2 and decided to build on what they already made for XCOM1.

One can always run into the problems Bioware had with Frostbite - where allegedly a lot of fundamental functionality needed for their title just wasn't there.

They should use unreal engine 5 because 1) unreal engine 5 is great (recent update 5.1 is just magical) 2) their own engines suck

They should not make bg4 though.

That is some interesting point of view there, care to develop any of it?

Excellent question, OP!! And a resounding 'yes' from me.

Every major problem with BG3 can be traced to the limitations of Larian's home-brew engine: cannot pause the game during real-time exploration; no day-night cycle; toilet-chain party movement; inability to do reactions properly; etc. Everything is because of the Larian engine's limitations. But Larian fans love it because "cool graphics," supposedly.

UE5 has no significant limits on what you can do with it. And it's graphics fidelity blows every other engine away, including Larian's by a wide margin. It can do first person, third person, or iso; RT, RTwP, or TB; single-player, co-op, or multiplayer; solo play or party-based. All of these things and more are possible with UE5, and that's why pretty much all RPG developers large and small are flocking to it - except Larian.

One thing to keep in mind is that UE comes at a hefty cost, from their official page :


5% is a crazy amount of money on such numbers. If possible, it absolutely makes sense to reuse the engine that you already made profit with and you already have developers familiar with. Especially when you can have some paid partnerships; like they probably did with Nvidia to integrate their cutting edge features.

Is it, though? What makes you think that if Larian were working with other engine, they wouldn't still design the game with multiplayer first in mind. Game with far smaller budget has stuff you mentioned - seems more like a design decision then an engine limitation.

But Larian's choice of MP first was not there in my list of BG3 game design limitations due to their engine (the very part of my post that you quote in you post).

It seems like Wormerine's point is that Larian's desire for MP first that is the cause of these problems you list, not engine limitations.

In an interview, Larian's reason for no day-night cycles was: "We're a multiplayer game and day-night cycles in multiplayer becomes incredibly complicated." I.e., since time can pass differently for players in TB vs real-time exploration, creating perfectly realistic time-dependent NPC schedules is too difficult.

In fact, I'd attribute most of your list to design decisions, not engine limitations.
- Toilet chain is a combination of Larian's historical preferences ("don't fix what ain't broken") and to make the experience "better" for console players.
- Automatic reactions are to make combat quicker & smoother. Related to the Engine: BG3 already has pop-ups---tool tips and quest rewards--I could easily see these pop-ups being extended to reactions during combat.
- I don't have direct evidence that Larian's engine can handle a game-wide pause, but I similarly don't have evidence that it can't. It's likely that Larian hasn't implemented this for similar reasons as for day-night cycles: "it'd introduce problems for time passage in multiplayer." (ignoring the solution of one person pausing causing a full pause for everyone)

(Everything above is being argued from Larian's POV: e.g., their ideas of what is better. Almost all of which I disagree with.)

Ah okay. Fair enough.

I still stand by my assessment. I recall even seeing articles from way back where someone from Larian said something to the effect that not being able to pause in exploration and the party movement mechanic (at least those two things) were because of engine limitations (I tried re-finding those articles but couldn't).

In any case, the point about Larian's game design choices being driven by MP over SP is itself an excellent point, one with which I complately agree and have said so quite strongly right from their very day announcing the game. What Larian is claiming they're trying to do is to create effectively two games in one: a SP classic cRPG videogame experience, and also a coop MP tabletop D&D sim gaming experience. But, even though they won't admit to it, the latter clearly is dominating the former. Meanwhile, across the RPG genre with most major developers, the trend now is to go for one or the other, SP experience or MP experience, and not try to have both in the same game. All the big-name, big-budget RPG projects from Bethesda, Bioware, Obsidian, inXile, etc. are declared to be exclusively or mainly SP games. Even CDPR has now declared they're going to make three separate new Witcher games (the first game in the new trilogy plus two other standalone games), two as SP games and the third as a MP game, having explicitly abandoned their previous plan to create one big SP+MP game. So Larian is pretty much alone as a major RPG studio still pushing the "SP and MP games in one" idea. And in the end, my prediction is that it will leave both groups of fans disappointed.

But returning to the OP's question, 2023 is shaping up to be a banner year for RPGs. And especially among AAA big-budget RPGs (but even some small-budget RPGs), pretty much all of them are using UE5, except for BG3. So people will surely compare and judge on that basis, I certainly will, and BG3 will end up being found wanting in many areas.

Ah, well if that's the case then yeah, their engine sucks!

Oh, nice! I had only heard of CDPR planning to make The Witcher Remake. Is that one of the "two other standalone games" you mentioned, or is that in addition to those plus the Cyberpunk stuff they're doing? Although in either case, Witcher Remake + 3(2?) other Witcher games + CP2077 expansion + Cyberpunk Project Orion seems like way too many projects to be working on at once...

(A quick google says CDPR is planning on making future titles in UE5, but if discussion of Witcher games/CDPR is too off-topic here, I guess reply to the... P:WotR thread? That seems to be the catch-all thread for non-BG3 games.)

No, not including any remakes. They have actually announced plans for FIVE brand new Witcher games. Three of them are a new trilogy set in the Witcher franchise but completely new and unrelated to the story of Geralt which is done. And so obviously the trilogy starts with its first game, codenamed Polaris, which is what is currently being worked on. Then, they have two more separate standalone games being made using the Witcher franchise. However, those two games have been farmed out to other studios. Codename Sirius is being made by The Molasses Flood, which I understand is a separate studio owned by CDPR. This is the MP game. Then there is codename Canis Majoris, to be made by a third party that has not been revealed.

Additionally also revealed are: Project Phantom Liberty, an expansion to CP2077; Project Orion, the next new Cyberpunk game; and Project Hadar, a game using a completely new IP that CDPR is creating.