Curious about UE5 and Halo Infinite

Why do some want this to be the engine for Halo?

What will it do for this franchise? :thinking:

I know it can churn out some pretty realistic graphics, but won’t that negatively affect performance?

2 Likes

I imagine most people that think UE5 is a good fit are people who only superficially understand game engines, to be a little blunt. I think moving to UE5 would negatively impact a Halo title. That’s not to say that the current iteration of the engine is perfect (clearly, it isn’t), but rather that the look and feel of a game come from the engine and how it handles physics, shaders, AA, AI, user input, etc and that UE doesn’t do it the way any other Halo game has done it, and thus would not be able to provide a Halo experience.

6 Likes

Maybe not if you develope for next Gen concoles only. I have to imagine that many of their set backs have to do with trying to get features to run on the Xbox One

2 Likes

I’m not really hip to all of these engines and what they can do. The devs really keep all of this hush hush for no real reason.

1 Like

I would imagine that this belief is derived from how long the Slipspace Engine took to develop, not to mention the potential impact on workflow that the creation of an entire engine might have had on Halo Infinite.

Would I actually advocate for using UE5 for Halo? Well… no, since we already have a next-gen engine. Could the next Halo game feature UE5 or it’s future successor? Maybe, it depends on how it works.

A lot of the time it’s an engineering decision on which engine to use (coupled, in Halo’s case, with input from folks on the gameplay systems and franchise/artistic side who are concerned with look/feel, I imagine) and so it’s definitely not something that the marketing side of a game studio really understands well enough to make any hoopla about probably. But yeah, even in engineering blogs, devs usually focus on what the new engine does better, but not why it’s important they chose the engine that they chose.

I’d suspect, for most devs (not Halo specifically), the choice of what engine comes down to a mixture of things - how expensive is the license for the engine if it’s not one you own, how hard is it to maintain the engine if it’s an in-house engine, how easy is it to integrate with middleware applications, is the look/feel right for the artistic direction, how easy is the workflow for the engine for the various disciplines, and on and on.

1 Like

Well, for starters:

  • You don’t have to maintain the engine. Epic does it… in a great way.
  • You have literally thousands of capable artists around the globe knowing, using and mastering the engine. No need to get them up to speed with slipspace.

Slipspace seems to be a complete mess. Add one thing, break 2 other. Hiring new people you need to give them time to learn how to use it.

IMO, in house engines are nice if you are well structured and led by a good management… like naughty dog. Or have no real time pressure like Ninja Theory. But for 343, an easy to use, well maintained engine would be absolutely essential.

7 Likes

I think one things with this viewpoint that is often missed is that the entire current work force of 343 is trained on the current or past iterations of the Blam! engine (Slipspace is, effectively, a heavily updated and reworked Blam!). You would have to retrain the entire 343 dev force on UE in order to make the transition (or replace them with people who have shipped UE games). It would also likely require changes to the vast majority of workflows and code management processes that are all designed to operate with how Blam!/Slipspace works. It would likely take years to make such a transition from a business process standpoint, and then years on top of that to get the first product out the door. It’s something that is technically possible between games, but nigh on impossible in the middle of a game’s lifecycle.

Well, arent most of the employees contractors for a maximum of 1.5 years? Isn’t this the big problem?

But of course, you are right when it comes to retrain people. Though I have to say UE5 is extremly user friendly and has a steep learning curve. Im coming from a motion design and 3D Design background but we are starting UE5 to implement in our workflow and it’s pretty intuitive and a lot of fun.

IF 343 decides to change engines, they would have to to a reboot of the game entirely. Would it be worth? No.

So my hope is they through their 10 years plan out of the window as soon as possible, take the feedback from infinte as long as it is fresh and start working on Halo:whatever its called as soon as possible using UE5.

I know they do use contractors (though, I also know most large studios and large software companies do as well, in fact my company does for several of our products), but I don’t know if we have confirmation on how long they stay (we have assertions on the Internet that MS in general does 18 mo contracts, but not 100% sure if that applies to 343) and we don’t have any official documentation around the percentage makeup of contractor vs permanent employee.

On that note, I do work at a very large company (~80,000 employees worldwide) and software/hardware products for enterprise IT/networking solutions is what we do. Our company runs a split team for major software updates/dev work - 1/3 to 1/2 of the team will be permanent employees and the rest will be contract workers. Permanent employees build the software spec and scope the work, contractors with C++ experience are hired (Slipspace is C++ as well, so it’s not entirely unfamiliar to others in the industry, IMO), and the contractors are given the spec and let loose to make it happen with permanent employees effectively acting as team leads and escalation points. I imagine MS/343 likely operates similarly, given their scale.

In my personal opinion, given my experience with being around similar dev practices, I don’t believe that the contractor system is Infinite’s problem. Personally, I think that likely it’s more to do with design oversights for game systems, bugs, and issues with getting new tools and processes pushed out to the various workgroups with adequate training internally, probably all coupled with a lack of time. Supposedly, post-Halo 5, re-tooling interfaces and tools for many of the artist workgroups was part of the big push on Slipspace. From my experience, even at a company many times the size of 343 with a very large number of devs that work on internal tools, new tools always have bugs and issues and people never attend trainings and it takes forever to get new tools working smoothly and people utilizing all the features of the new tools to make work faster.

2 Likes

Blam and Slipsace are notorious for being hard to work with. Bad netcode seems to run at its core. It also doesn’t help that the contract work means that the constantly rotating employees have to pick up a proprietary and infamous engine. But 343 stuck with it because it is the Halo engine, just like how idTech is the Doom, Quake, and Wolfenstein engine (though I hear much less issues from them and the latest idTech engine is a optimization powerhouse.)
Unreal Engine is used by many different companies and is openly available, so more people know how to use it and because it is standard, there are (theoretically) less long-term issues.
Even talks went on behind the scenes to swap to Unreal and many fans think that if they did make the swap to unreal we would be much further ahead than we are now, regardless of how late the Unreal talks came into development.

Players who weren’t following pre-release rumors and hype probably don’t know that 343 apparently originally tried to move Halo onto UE4 after H5.

Supposedly it took~1.5 years of concepting/development and 343 eventually decided to create their own engine (Slipspace) instead. The main two reasons known are

  1. Difficulty recreating “Halo feeling gameplay”.
  2. XBO support for the campaign would have been impossible

Here’s just a couple of examples of what some fans have done with UE4 and UE5…

https://www.youtube.com/watch?v=jqh5fyu3Lb4

https://www.youtube.com/watch?v=F-4VmN18Nkc

https://www.youtube.com/watch?v=ryEAYC9Zz-c

https://www.youtube.com/watch?v=yoyxL5CozUo

1 Like

To be fair 343 as a whole can’t seem to provide a proper Halo experience.

The only way I can think of putting this is like this:

UE is like buying a standard PC case and all the components you need to make it run, each part can be changed whenever you want to achieve what you’re looking for.

Slipspace is like building the PC case and every component needed for it to run from scratch, even once its built you’re going to be fighting with it and making modifications to try and keep it running properly. If one part is broken you can’t just swap it out because this PC is completely unique and requires new parts to be made specifically for it.

UE5 is a great game engine that is made for ease of use with all sorts of different projects.
They have the experience in updating and creating new game engines that spans all the way back to the mid to late 90’s. What you get with UE is something that almost guarantees a great end product, not to mention the modular nature that UE has had for over 2 decades is a benefit.
There’s also the fact that the UE devs have already solved and dealt with many of the learning curves that come with game engine development.

Slipspace is 343i’s first attempt at creating their own in house game engine as far as I know. They aren’t using the modified Blam! stuff made for them from the Bungie era. They have no documentation when it comes to ironing out kinks or adding features, they have to learn as they go which is inherently going to cause achingly slow development.
When you throw in the fact that game engines have to be so complex to even come close to competing with other engines used by AAA developers currently and the fact that those other AAA developers are using game engines that are already modern and already have experience developing with, you get the situation 343i is in now. Achingly slow dev time and incredibly long times between updates.

343i making their own Engine and not having a side project using the old engine (Similar to Halo 3 ODST) to fill the void during development of Slipspace is a bit rough.

I honestly believe that if they started developing Infinite with UE4 we would have ended up with a great Halo game and most of the expected features on release instead of the predicament we have currently which is a dwindling player base less than a year after release along with many, many upset fans…

lol I like this, but I would say Slipspace is like buying a Dell Alienware pre built PC (watch Gamers Nexus review on it).

Slipspace is not a brand new engine built from the ground up. It’s just Blam polished up a bit with a new name.

1 Like

Seems that way.

1 Like

I agree with this to a point, though. Features are features, this isn’t really unique to an engine but I do believe Halo would run a lot better on UE. The features/content are up to the developers and as far as the split screen goes, Digital Foundry did a video on that and it looked completely playable. Even on the OG Xbox One. I think there’s more to the story then just canceling it.

This is just speculation but if you ask me, I think either the whole team or a good chuck of that team responsible for split screen left 343. Before they left, they added a bug to be able to play split screen. Most of these “bugs/glitches/workarounds” are actually leaked by someone from the inside. That’s my theory anyway.

As far as the core game play of infinite, I really enjoyed it. It felt like Halo, how Bungie would of made it. It has proper sprinting. The sprinting in 4 and 5 were horrible, 5 not as bad but 4 was dog shi.

Gears 4 and 5 crap all over Infinite regarding performance and fidelity. Both run on UE4. They should’ve switched to UE years ago…

1 Like

Depends on settings. If you turn up the SSGI settings in Gears 5 it actually runs worse than Infinite does on PC. For all of Slipspace’s flaws, it does GI with less of a performance impact that TC’s iteration of UE4. And TC’s iteration of UE4 for Gears 5 (at least on PC) has WAY worse TAA breakup in fast motion than does Slipspace (Slipspace’s TAA is fuzzier with slight movement, like waving grass).

1 Like