Support NeoGAF

[In Cold Blood]

I know Unity is around, but that's generally for low level games.

So, why haven't we seen any of the real high level alternatives to UE make a push for that third party market?
Crytek has tried with Cryengine, and that engine can make some stunning games, but the feedback is that its quite difficult to use.

It appears that there is more interest in keeping a first party engine proprietary and not allow others to use it.
Obviously there is a massive market for this, as Epic is making a shit ton from UE adoption, so why no one else?

The only thing I can think of is that a company like id with Idtech, just arnt set up for the support side of things, but if there's money in it surely it would vindicate it long term.

The great Sony engines I can understand not being multi, but you have a number of possibilities for this.
Idtech. Probably one of the best engine going around. There are a number of offshoots from it in the Bethesda studios, such as Arkanes engine.
I mean, even when Avalanche did Rage 2 for id, they used their own engine and not idtech that the first game ran on.

I would have thought that one day MS would have done an engine to go against Epic. They have some of the best engines like idtech, Forzatech, Creation and um...Slipspace.
It just seems like such a Microsoft thing to do. They are Probably the most likely to do it. They have the resources, they have the tech and they have the money to push it against Unreal.

What say you?

 

[AeonGaidenRaiden]

Because John Carmack is making rockets and playing with VR instead.

 

[The Pleasure]

Idtech was only a competitor during idtech 3. All their other engines never quite matched the versatility of idtech 3.

 

[Torrent of Pork]

Unreal seems to be the best generalist engine to use for the biggest variety of genres. You're obviously not going to do everything with it, but you can do most things with it, hence the popularity.

Inevitably a clear winner would rise to the top in the engine wars, and it would be the one that could do the most regardless of style.

 

[bitbydeath]

You don’t want to search for bethesda unreal engine OP. Creation is not thought highly of out there.

 

[Raonak]

Because making an user friendly general purpose engine is super hard,
and some of the stuff that UE5 is doing is so advanced that "catching up" is borderline impossible.

Who's actually gonna invest the amount of time and money required? It'd make more sense investing the effort into making a game instead.
Do you think doing QA for a game is hard? Try doing it for an engine. that's an even deeper level of hell.

 

[lh032]

its unrealistic to do so.

 

[Three]

Unfortunately it's difficult to compete on that front largely due to workforce. Over time Unity and UE has become the standard. When studios look for experience they see a sea of UE talent and not many for the more obscure engines, which only makes competition even more difficult as it drives more studios to UE again. It's only those who have the resources to support an inhouse/bespoke engine that do it.

 

[CamHostage]

The engine isn't what makes a game amazing. A good one helps, of course, but the vast majority of work on a game (including much of the cool stuff that people praise as "the engine" in games they love) comes from hard work designing custom components to address the needs of their game.

Also, the "engine" is just a small part of what development suites like Unreal and Unity provide. The development tools for integrating assets and managing sound and collaborating across the team, the plugin network of providers, the asset store, the support network of professional collaborators as well as the original architects, all of these things that gamers don't really know, they allow developers to get their work done on the project at large. (I'm not sure how this works out with licensing, but some developers have even used components or worked in these development systems while building games even though they ultimately ship on their own internal engine.) Nanite and Lumen look cool, but they aren't the big reasons why studios are choosing to work in Unreal (and if it is a deciding factor, it's not because Epic is doing something nobody else is capable of; it's because they've figured out and have nice implementation for various features which would be a significant investment of time and talent to develop their own version of.)

The ability to write an engine per se isn't something only the elite and most well-funded can do. There are bound to be people on staff at a big studio who could write an "engine", and most of the sexy things a game engine could do with graphics and physics and such are documented and available to learn and integrate. A high-tier studio will likely have completely rewritten or custom-built huge swaths of the "engine" for their game, to the point that some developers license an engine but call their technology stack by its own name, so even if they use an "engine" and its development suite, they're also doing a huge amount of the work of making an engine too. At a certain point, the stability and suite feature set of something like Unreal may make it worth not starting from scratch anymore, but when you're at the level that these big studios are the OP is talking about here, that's only going to get them started; it's not what's going to make the project amazing.

 

[kikkis]

UE development is sponsored by fortnite money. Engine licenses and royalties are at the end of the day quite small compared to initial investment to create generalist engine and tools, with quite small pool of labor of senior game engine programmers. I mean ue5 and unity are both really popular options and duopoly, yet unity is struggling to make the ends meet.

 

[ChorizoPicozo]

one word:

Fornite.

so be grateful.

 

[IAmRei]

because making game engine is hard, and if there is one who make it free, they still need to think how to monetize it and make it accessible to beginner as well, so the money poured into development will recoup and starting to make business worked. unless the maker only want fame, but game engine as powerful as UE, would take lot of effort, to the point it will force people who make it needs to fill their kitchen without disturbing development, a.k.a. need investor. and you know investor usually wants return, unless they are very generous angel investor.

TLDR: it's almost impossible without money burns

 

[Damigos]

I have another question :
Since Unreal Engine is so much better than the others, why dont more developers use it?

 

[ByWatterson]

Don't know much about this, but I remain convinced Sony will eventually license out Decima.

 

[IAmRei]

I have another question :
Since Unreal Engine is so much better than the others, why dont more developers use it?

if it's 5, some said it's still buggy, but again, 5 still relatively new

 

[Damigos]

if it's 5, some said it's still buggy, but again, 5 still relatively new

Not only 5, any UE

 

[Tams]

I have another question :
Since Unreal Engine is so much better than the others, why dont more developers use it?

Because they want complete financially and creative control.

If you create your own engine, then assuming you create everything else for a game, the game is yours to do with a you wish without consulting or paying other parties.

 

[ergem]

Decima engine.

 

[In Cold Blood]

one word:

Fornite.

so be grateful.

I agree that it takes alot of money to support and develop a cutting edge engine.

If we were looking at what company could make one to compete with UE, one would without a doubt could be Microsoft.

Software is what they are good at. Supporting software is also what they do day in day out.
They already have some cutting edge engines in idtech, Forzatech and most likely IW7 in their stable.

Epic might have that Fortnite money, but MS has 100 times what Epic has.
It just seems to fit right in with what MS is all about. Why would you think MS should be in the API buisness? There they are. I'm also pretty sure that there is no money from DX and it just costs them money to put it out.
They have the API, they have the game engines, they make video games, have their fingers in all the pies, so why not this one?

I know there is a massive hate boner on here for MS, but if people put that to the side, it is surprising that they arnt all ready doing it.

 

[Damigos]

Because they want complete financially and creative control.

If you create your own engine, then assuming you create everything else for a game, the game is yours to do with a you wish without consulting or paying other parties.

Maybe thats why some studios go bankrupt, or development time for big games is around 5 years.
By the way, i am pro consumer and i hate monopolies. But with a graphics engine it feels like you re creating your own cpu instead of getting an AMD or Intel one. Creating your own engine feels like creating your cpu

 

[Tams]

Maybe thats why some studios go bankrupt, or development time for big games is around 5 years.
By the way, i am pro consumer and i hate monopolies. But with a graphics engine it feels like you re creating your own cpu instead of getting an AMD or Intel one. Creating your own engine feels like creating your cpu

Something tells me that you have no idea wtf you are talking about.

 

[Hudo]

Because John Carmack is making rockets and playing with VR instead.

actually he moved on to neural networks and AI.

 

[March Climber]

Only the creation engine(Bethesda) and whatever crazy engine Beyond Good and Evil 2 has, are actually unique in what they offer for 2023 AAA games.

John Carmack and idTech gracefully bowed out years ago as third party competition.

So therefore, right now there isn’t a good reason to have a new engine other than for monetary purposes or saving money.

 

[WhamBamThankYouMaam]

Because creating one is fucking HARD.

 

[CamHostage]

If we were looking at what company could make one to compete with UE, one would without a doubt could be Microsoft.

They already have some cutting edge engines in idtech, Forzatech and most likely IW7 in their stable.

It just seems to fit right in with what MS is all about... They have the API, they have the game engines, they make video games, have their fingers in all the pies, so why not this one?

I'm sure none of those engines right now have the level of support staff or documentation needed to let just anybody else use them for their totally different projects, but besides that...

Sure, this is a business MS could be in, and they've taken a bite at that Apple a few times in the past. (XNA Game Studio wasn't quite the same thing as Unreal, but it's probably the most familiar attempt by MS in this field... interestingly, the XNA library lives on in FNA and MonoGame, which is popular with indie 2D games like Celeste and Stardew Valley.) It just never added up as a worthwhile market. MS is built to be a dominant force in the markets it tackles, and if it is just one of many service providers in a field, it's usually not worth investing in that field.

Even though some offices also have their own tools, Xbox Game Studios has already bet big on UE5; sister studios under Bethesda similarly have their own tech (some of which in the past has been used for middleware solutions,) but the pace and demands of game design now would likely impact the actual productions dependent on those game engines by their own internal studios. Unless there was big, big cash to be made in licensing, it's better for them all to do the jobs they're designed to do.

 

[SantaC]

John Carmack is so glorified here. He was revolutionary in the 90s and early 2000, but havent done anything for a game engine in 15 years or so. People should let it go already. His VR stint wasnt the best either.

 

[Thick Thighs Save Lives]

Because gen z don't want to learn anything other than Unreal Engine and Unity.

 

[ahtlas7]

Doesn’t UE provide a lot of tools to help increase productivity and ease of use?

 

[Black_Stride]

Unity is the most popular engine on the market......devs just arent using HDRP so people assume the engine cant do amazing things.
The Enemies benchmark is still the most impressive thing you can run on your PC and even has RTX 4090s begging for mercy at native 4K.


CryTek had that weird period where they were in the shitter, and their support and documentation isnt at the same level as Unity and Unreal, they are improving but I think it might be too late to really claw back any developers.
Unigine just doesnt seem to have any game devs using it, but its used extensively(atleast it used to be) for other realtime applications such as archviz and military simulations

Making and maintaining your own engine is hard work, Sony has the ICE team who help with basically all the engines.
ATG from Microsoft also help with maintaining and upgrading their inhouse engines.
id and idtech are anomalous considering they are a relatively small team, but clearly talented beyond measure so they have an engine thats super performant.....im hoping under Microsoft more teams will be able to use it.

Doesn’t UE provide a lot of tools to help increase productivity and ease of use?

Indeed it does, Unreal Engine has alot of features that help in that aspect, but its far from perfect, Unity is much easier to use/learn.
The Crystal Dynamics team recently moved to Unreal Engine 5(sad face) and they said there were features they loved from the Foundation Engine that Unreal doesnt quite match.....but when you are the size of Epic, alot of features can be implemented much faster than if Crystal Dynamics tried to replicate Nanite in their own engine.

One of the complaints that Crystal Dynamics (and me) had of Unreal Engine was that the material system isnt quite up to snuff compared to offline or even CryEngine........they have since implemented a completely new material system and they will eventually sunset the old material system.

 

[crackajack]

Might be the future for every company, since it seems to become more and more the de facto standard. Sony used it already despite their in house wizards, the Gears guys naturally...
But for now, we still have Cry/Lumberyard, idTech, Decima, Frostbite, Dunia, Anvil, Source, RE, Snowdrop, IW, Glacier, probably some more and at least some are licensable. Unreal and Unity are probably the cheapest and Indie friendliest though with the most people having some or good knowledge and it makes just sense to use those.
Unreal 3, so before Fortnite and also before Unity could be taken seriously for "real" games, just dominated and the few engines that actively sought to be licensed where seemingly only Cry and idTech3, most others where used if you were close partner or something but usually quite proprietary and developed for one game or a specific line of games (Ego probably dies now too?). Unreal succeeded idTech1-3, which was maybe the only other engine family that dominated, to a lesser extend, for a time. Once you have that position it's hard to change the market share I guess, unless Epic fucks up in some colossal way. No idea myself, but Unreal isn't without flaws, but overall must still be better than trying to build something yourself, with full control, but also full responsibility to make it work. CD Project also plans to switch, right?

It's the same with Windows and why no one guided free Linux to the top, which Valve now kinda tries, while Apple is beloved by its fans but they are still distant number 2. Or DirectX where OpenGL and Vulkan kinda are/were better but also not quite, and DX stays the standard.
Replacing DX probably does not make money, but trying to topple DOS/Windows potentially could, but the last company that tried was IBM with OS2. With no success. Solaris and actual Unix kinda disppeared. Like PA-RISC or the Power architecture.
LibreOffice is basically the only competition to MS Office.
Softimage, Maya and 3DSMax will also merge sooner than later, since all are owned by Autodesk.
The CAD/CAM world is divided by a few companies and the only start up that was kinda late was SolidWorks and almost immediately got bought by one of the old top dogs.

There are niche areas for certain specific stuff, but anything that is used by many, one or two programms will win and push out any competition in one way or another.

 

[In Cold Blood]

I'm sure none of those engines right now have the level of support staff or documentation needed to let just anybody else use them for their totally different projects, but besides that...

Sure, this is a business MS could be in, and they've taken a bite at that Apple a few times in the past. (XNA Game Studio wasn't quite the same thing as Unreal, but it's probably the most familiar attempt by MS in this field... interestingly, the XNA library lives on in FNA and MonoGame, which is popular with indie 2D games like Celeste and Stardew Valley.) It just never added up as a worthwhile market. MS is built to be a dominant force in the markets it tackles, and if it is just one of many service providers in a field, it's usually not worth investing in that field.

Even though some offices also have their own tools, Xbox Game Studios has already bet big on UE5; sister studios under Bethesda similarly have their own tech (some of which in the past has been used for middleware solutions,) but the pace and demands of game design now would likely impact the actual productions dependent on those game engines by their own internal studios. Unless there was big, big cash to be made in licensing, it's better for them all to do the jobs they're designed to do.

One of the things that Epic did was buy some of the most important companies who would make UE better.
Qwixel Megascans. RAD toolsCloudgine

Unity is the most popular engine on the market......devs just arent using HDRP so people assume the engine cant do amazing things.
The Enemies benchmark is still the most impressive thing you can run on your PC and even has RTX 4090s begging for mercy at native 4K.


CryTek had that weird period where they were in the shitter, and their support and documentation isnt at the same level as Unity and Unreal, they are improving but I think it might be too late to really claw back any developers.
Unigine just doesnt seem to have any game devs using it, but its used extensively(atleast it used to be) for other realtime applications such as archviz and military simulations

Making and maintaining your own engine is hard work, Sony has the ICE team who help with basically all the engines.
ATG from Microsoft also help with maintaining and upgrading their inhouse engines.
id and idtech are anomalous considering they are a relatively small team, but clearly talented beyond measure so they have an engine thats super performant.....im hoping under Microsoft more teams will be able to use it.


Indeed it does, Unreal Engine has alot of features that help in that aspect, but its far from perfect, Unity is much easier to use/learn.
The Crystal Dynamics team recently moved to Unreal Engine 5(sad face) and they said there were features they loved from the Foundation Engine that Unreal doesnt quite match.....but when you are the size of Epic, alot of features can be implemented much faster than if Crystal Dynamics tried to replicate Nanite in their own engine.

One of the complaints that Crystal Dynamics (and me) had of Unreal Engine was that the material system isnt quite up to snuff compared to offline or even CryEngine........they have since implemented a completely new material system and they will eventually sunset the old material system.

It's actually interesting in that the two best MS in-house engines are Idtech and Forzatech, and both Id and Turn 10 are relatively small studios in the grand scheme of things.

 

[Starbuck2907]

Most major games series by big publishers have custom engines that can be updated and reused for each sequel. The cost/benefit of updating and engine and toolset vs licensing an engine and tools usually isn’t worth it.

The new Mass Effect will use Unreal Engine 5, while the Dead Space remake used Frostbite. I suspect this is because the scale of a ME game is so big that the better tools and engine supporting more environments with less work will result in a significantly shorter development time.

If I remember correctly, don’t a lot of MMO’s use “off the shelf” engines, but we just don’t hear about them as much?

 

[Hudo]

Doesn’t UE provide a lot of tools to help increase productivity and ease of use?

This is probably the biggest reason why Unreal's adoption has increased. And also, the documentation is far better (it's still mediocre, imho) than it was before.

Interestingly enough, the ecosystem and asset pipelines etc. are also the reason why some studios have stuck with their own engines (Capcom, most of EA's studios, Blizzard, Rockstar, Nintendo, ...). Also, some things are a pain to do in Unreal and other things are a breeze. Unreal is not a perfect fit for every use-case.

 

[winjer]

Unity can do amazing things. But almost no one is using it to make AAA graphics in games.
UE4 became a standard but it has so many performance problems, it's a wonder why so many studios choose to use it.
We can only hope that UE5 won't become the same mess that UE4 is now.

 

[The artful mincer]

What about Dreams!!!

 

[SHA]

Historically speaking, it wasn't even fair to start using the engine effectively before the end of console cycle, the cycle makes it obsolete which leaves them with no choice but to start all over again.

 

[ScHlAuChi]

There isnt much competition becasue no other company (aside Unity) is willing to put hundreds of devs on just the engine side.
The only other company that has enough money to do it is Valve, but their tech is a decade behind the competition and Valve cant get shit done anyway.
It is no surprise that even CDPR switched to Unreal as maintaining your own tech to be competitive is a huge cost factor.

 

[Hudo]

I also found it interesting that BlackBird have decided to move from Unity to Unreal Engine for Homeworld 3 (one of my most anticipated games)

 

[Hudo]

The only other company that has enough money to do it is Valve, but their tech is a decade behind the competition and Valve cant get shit done anyway.

Is Source 2 really that far behind? It seemed pretty much state-of-the-art to me in many aspects in Half-Life: Alyx. I even remember being amazed by how one guy at Valve implemented the physics "simulation" (it's not really what I'd call a simulation but just really cleverly done) of fluids in bottles using mostly only the pixel/fragment shader.

 

[Dr.D00p]

If they could only solve the shader compilation issues (PC gaming) I doubt anyone would care, TBH.

 

[ScHlAuChi]

Is Source 2 really that far behind? It seemed pretty much state-of-the-art to me in many aspects in Half-Life: Alyx. I even remember being amazed by how one guy at Valve implemented the physics "simulation" (it's not really what I'd call a simulation but just really cleverly done) of fluids in bottles using mostly only the pixel/fragment shader.

Yes, they basically added some "modern" features on top of their original Source 2 to make it good enough for HLA.
And that worked pretty well for a game like that, but technologically the engine is stuck in early 2000s design - in no shape or form fit for modern development.
UE5 with Nanite and Lumen is 3-4 generations ahead and let´s not even get started how much better the editor and tools are.

Check this recent UE5 video showing off their new tools at around 4 minutes in, this is some real next level shit, while Source still basically works like the Quake Editor:

 

[Krappadizzle]

 

[Shubh_C63]

It's a runaway train. No stopping it

Only Source Engine ever had a chance but Valve never pushed I guess.

 

[ScHlAuChi]

It's a runaway train. No stopping it
Only Source Engine ever had a chance but Valve never pushed I guess.

Source engine never had a chance, it is based on the Quake engine. You would have to write a completely new engine to be able to compete.
The only engines that can compete are propitary unlicenceable engines made by studios like NaughtyDog, Insomniac, Guerilla or like EA´s Frostbite and Rockstar´s RAGE etc.
And then we have Unity and Cryengine that can be licenced but neither can compete with Unreal on the same level.

 

[Hudo]

Yes, they basically added some "modern" features on top of their original Source 2 to make it good enough for HLA.
And that worked pretty well for a game like that, but technologically the engine is stuck in early 2000s design - in no shape or form fit for modern development.
UE5 with Nanite and Lumen is 3-4 generations ahead and let´s not even get started how much better the editor and tools are.

Check this recent UE5 video showing off their new tools at around 4 minutes in, this is some real next level shit, while Source still basically works like the Quake Editor:

From the technical talks and the paper I read about Nanite, it seems neat that they achieved some sort of "silhouette stability" for many polyhedral complexes (that was always a problem in dynamic LOD generation) in an easy-to-use way. Lumen is just marketing, imho. I don't find anything impressive there. But to not take away from your point: Yeah, Unreal is certainly state-of-the-art. Bit overhyped and over-marketed in some aspects, but that's just how business is. The last thing that actually somewhat impressed me (because they managed to get a solid and working implementation) was the mesh deformation based on trained neural networks (I think they call it MLFormer, so the name suggests that they probably use a Transformer for this, which would mean a lot of training data is required (and clever agumentation)), for example for muscle tension under the skin or cloth deformation physics. The "classic" way of solving this was to use some sort of mass spring system and then concentrate all your efforts on an optimizer/numerical integrator that's fast and robust(!) enough for real-time shit. We had an idea to use deep-learning for this as well but lacked the means to create a good dataset. If you happen to know what kinda data they used for training, I'd be very interested to know.

 

[LazyParrot]

UE4 became a standard but it has so many performance problems, it's a wonder why so many studios choose to use it.

I don't think UE is the problem here, at least not for the most part. Considering the conditions many AAA games are developed under, it's not at all surprising that a lot of them release with performance issues. It's just more noticeable with UE because it's the most widely used non-proprietary game engine, so there are going to be more unoptimized releases overall.

It's also fairly obvious why studios keep choosing it over other engines. It's been around for 25 years at this point, so there is a huge pool of potential hires that have years or even decades worth of experience with it. The support and documentation are also solid, easily accessible, frequently updated and available in multiple languages (this is the main reason why so many Japanese studios switched from proprietary Engines to UE - they had everything translated into Japanese very early on).

 

[winjer]

I don't think UE is the problem here, at least not for the most part. Considering the conditions many AAA games are developed under, it's not at all surprising that a lot of them release with performance issues. It's just more noticeable with UE because it's the most widely used non-proprietary game engine, so there are going to be more unoptimized releases overall.

It's also fairly obvious why studios keep choosing it over other engines. It's been around for 25 years at this point, so there is a huge pool of potential hires that have years or even decades worth of experience with it. The support and documentation are also solid, easily accessible, frequently updated and available in multiple languages (this is the main reason why so many Japanese studios switched from proprietary Engines to UE - they had everything translated into Japanese very early on).

Almost every game using UE4, released in the past few years, has major performance issues. From the biggest AAA, to simple indie games.
It's not normal for a game engine to be this bad. And although devs might have some blame, the reality is that the issue is the game engine itself.

 

[TVexperto]

I know Unity is around, but that's generally for low level games.

unity is low level? dafukk?

 

[NahaNago]

It's mostly a matter of money and constantly updating/upgrading the engine that will stop most companies.

I think an engine for a niche genre could work best at this time for indies. Farming, sims, pokemon style, and card games.

 

[Valonquar]

I dream of the day when my character's hair doesn't clip through his armor.