Unreal Engine 5, [UE5 Developer Availability 2022-04-05]

[Andrew Lauritzen]

The developers that made Remnant 2, Immortals of Aveum just don't have that history and don't have that time.

They are also much, MUCH smaller companies (and often shorter timelines). I know ultimately to end users the only question is the delivered value of a given game, but if you are discussing tradeoffs of engines and the quality achievable by a team of a given size in a given time, obviously those sizes are a big part of the equation.

You really do have to compare like for like rough game budgets here at some level. If you're going to compare massive AAA games on other engines, you kind of have to wait until those exist for UE5 (assuming you are ignoring Fortnite for whatever reason). I think part of the reason people feel like they can compare stuff like Remnant 2 to games made by much larger teams is because it does punch somewhat above its budget in terms of visuals, even if they are uneven. Similar for Talos 2, and definitely for Ark SA.

 

[Andrew Lauritzen]

Meanwhile even small studios like The Outsiders (Metal Hellsinger) or Moon Studios (Ori games) or Unknown Worlds Entertainment (Subnautica) or Studio MDHR (Cuphead) can make Unity (an engine with a reputation worse than it deserves) sing and produce incredibly impressive results. But again, there are also lots of Unity games that are hot buggy messes. Same goes for Unreal Engine.

Yup and I think Jusant is a perfect and timely example of this on the UE5 side. Smaller game from a small dev team that runs great and looks great.

 

[Andrew Lauritzen]

Yeah, the first game was initial very rough on console, so I'm not expecting the second game to be much better in terms of performance and adjustments to graphics settings to make it work on console at least initially. Not sure where Ark finally landed WRT performance and graphics by the end, however.

Yeah it started super rough on both performance *and* graphics on all platforms, even PC. Performance got somewhat better over time but some of that is just hardware improving. The visuals still look bad in Ark SE with lots of stuff just never addressed or improved. Ark SA is very heavy on performance right now (at least at top settings), but at least it looks good this time around.

 

[BitByte]

Exaggerated hyperbole doesn't help your case. I mean Gears 4 and 5 immediately come to mind as great looking and great running UE4 games.

Regards,
SB

Is it really exaggerated? When we compare the list of good to bad ue4 games, the scales tip heavily in the favour of bad. So much so that we even have bad ue4 games on ps5/xsx. Platforms significantly more powerful than the original target set forth by epic. It doesn’t really matter why it’s bad be it traversal stutter, shader compilation stutter, development issues, single threaded nature of ue, etc. There’s one unifying theme, it runs on ue.

Look at UE5, I’m still waiting for a good release. DF released a console video comparison for Robocop, the latest ue5 game that leverages all the features. Ignoring the massive rendering differences DF missed during their comparison…. Ideally should lead to the video being deleted and replaced but ignoring that…. Sub 30 fps in a 60fps performance mode? Can’t hold 30fps is quality mode? Very very poor image quality? Watching the video on a 77 inch Oled was quite revealing. Other than Fortnite, UE5 has yet to deliver a solid product. I mean, are we going to argue that all the developers suck? You know, I don’t shy away from such things but in this case, it’s the engine that’s the problem. It cannot be all these devs.

 

[arandomguy]

Is it really exaggerated? When we compare the list of good to bad ue4 games, the scales tip heavily in the favour of bad. So much so that we even have bad ue4 games on ps5/xsx. Platforms significantly more powerful than the original target set forth by epic. It doesn’t really matter why it’s bad be it traversal stutter, shader compilation stutter, development issues, single threaded nature of ue, etc. There’s one unifying theme, it runs on ue.

To actually go with this line of reasoning I'd think you would at least need to compare the ratio of games with what I guess you'd term as the technical performance faults on UE against the the non UE game population. It's not like non UE games are devoid of stutter, development issues, primarily single threaded, and etc.

This similarly also applies to people who always bring up the good old days. Plenty of games in the past had technical challenges as well and that was a much higher rate of hardware advancement to brute force through those issues. You're often just selectively remembering the good games of the past. Everything mediocre or bad (unless it's hilariously bad) is often just forgotten.

 

[Scott_Arm]

Is it really exaggerated? When we compare the list of good to bad ue4 games, the scales tip heavily in the favour of bad. So much so that we even have bad ue4 games on ps5/xsx. Platforms significantly more powerful than the original target set forth by epic. It doesn’t really matter why it’s bad be it traversal stutter, shader compilation stutter, development issues, single threaded nature of ue, etc. There’s one unifying theme, it runs on ue.

Look at UE5, I’m still waiting for a good release. DF released a console video comparison for Robocop, the latest ue5 game that leverages all the features. Ignoring the massive rendering differences DF missed during their comparison…. Ideally should lead to the video being deleted and replaced but ignoring that…. Sub 30 fps in a 60fps performance mode? Can’t hold 30fps is quality mode? Very very poor image quality? Watching the video on a 77 inch Oled was quite revealing. Other than Fortnite, UE5 has yet to deliver a solid product. I mean, are we going to argue that all the developers suck? You know, I don’t shy away from such things but in this case, it’s the engine that’s the problem. It cannot be all these devs.

One of the problems here is "UE Game" becomes a category and it gets held up against the best of everything else, and not EVERYTHING else. What's the ratio of "bad" games to "good" games in total vs UE? There are so many variables: budget, technical experience, release schedule ("runway"), game engine, establishment (brand new company vs old company), culture (is it an executive driven company or a developer driven company) etc

 

[BitByte]

To actually go with this line of reasoning I'd think you would at least need to compare the ratio of games with what I guess you'd term as the technical performance faults on UE against the the non UE game population. It's not like non UE games are devoid of stutter, development issues, primarily single threaded, and etc.

That’s fair I guess but it’s heavily limited by the types and number of games a person plays.

This similarly also applies to people who always bring up the good old days. Plenty of games in the past had technical challenges as well and that was a much higher rate of hardware advancement to brute force through those issues. You're often just selectively remembering the good games of the past. Everything mediocre or bad (unless it's hilariously bad) is often just forgotten.

I guess but we selectively remember things for a reason. The things that are bad or mediocre that fade out of memory are irrelevant. They don’t make a memorable impact on us and our feelings on them aren’t strong. Things that we remember made a strong lasting impression on us and carry more weight. If people negatively remember UE, it’s because the negative impression was so great, we’re forced to remember which is drastically worse than being an inoffensive type of bad.

 

[BitByte]

One of the problems here is "UE Game" becomes a category and it gets held up against the best of everything else, and not EVERYTHING else. What's the ratio of "bad" games to "good" games in total vs UE? There are so many variables: budget, technical experience, release schedule ("runway"), game engine, establishment (brand new company vs old company), culture (is it an executive driven company or a developer driven company) etc

I think it’s two fold. How many devs released a whole engine demo trying to hype up excitement for their engine at the launch of the console generation? Epic has done it twice now and both times they’ve flopped.

The second side of this coin is the everything else you refer to. I think most individuals try to avoid playing bad games. The everything else is mostly irrelevant because the majority of the money goes to the biggest games. The issue is UE features in many of the biggest games so the average person is more likely to come across it. It doesn’t matter if indie dev 1 releases a game using their indie engine and it’s horrific. Most people won’t play it so it’s irrelevant.

 

[Scott_Arm]

I think it’s two fold. How many devs released a whole engine demo trying to hype up excitement for their engine at the launch of the console generation? Epic has done it twice now and both times they’ve flopped.

The second side of this coin is the everything else you refer to. I think most individuals try to avoid playing bad games. The everything else is mostly irrelevant because the majority of the money goes to the biggest games. The issue is UE features in many of the biggest games so the average person is more likely to come across it. It doesn’t matter if indie dev 1 releases a game using their indie engine and it’s horrific. Most people won’t play it so it’s irrelevant.

What I'm saying is the comparison for UE games always seems to be against the AAA best of the industry that have custom engines, but not against a failed AAA game like Mass Effect Andromeda or something. And if you go AA or indie there are just plenty of bad games all around. If there's a bad UE game, it's a knock against UE. If there's a bad game with a custom engine, it's not a knock against anything else. We don't know that these developers would get better results with whatever time and budget they have if they made their own game engine versus what they release with UE. I think the assumption is that it would be better, but that's because people are thinking of the best of bespoke engines vs all bespoke engines of which there are bad ones.

 

[Andrew Lauritzen]

We don't know that these developers would get better results with whatever time and budget they have if they made their own game engine versus what they release with UE. I think the assumption is that it would be better, but that's because people are thinking of the best of bespoke engines vs all bespoke engines of which there are bad ones.

The majority of the games mentioned here would not even exist if not for using 3rd party game engines and technology. The whole discussion of bespoke vs. commercial engines is only relevant for a small number of the very largest AAA game teams these days. The notion that you'd expect a developer with a few dozen employees to write their own engine from scratch is ridiculously disconnected from reality for any suitably complex 3D game.

Maybe some of you would prefer that we had way fewer games, or that the ones that get made were mostly 2D or much simpler, but I for one am happy that stuff like Jusant can exist without requiring some publisher to invest hundreds of millions of dollars, which we all know would never happen for a concept like that. (Or if it did, it would happen exactly one time.)

I'm also exited and happy to see the big AAA studios do their thing whether that be with Unreal, Unity, Godot, something bespoke or something else entirely.

These two things are not mutually exclusive.

I’ve honestly lost the premise here for these games. If the argument is they would have been better with Unity, cool, that can at least be a discussion. But arguing that a game shouldn’t exist because it doesn’t meet your personal criteria is stupid. And let’s be clear this is effectively what people are saying when we’re talking about these lower budget titles.

 

[Remij]

I just want games to work properly, regardless of the engine they use. At the end of the day, you work and design within the constraints you're given and your goal should be not to exceed it at the expense of the user's experience.

 

[techuse]

Plenty of developers with employee counts in the 100s made games on UE4 and the vast majority were technically troubled in very similar ways. Andromeda was a very bad product but its problems were unique to that game. Unless I'm remembering wrong, Frostbite engine games very rarely had issues in the visuals and performance category, where as UE4 games almost always had issues in those categories specifically. It's the trend that I am drawing conclusions from. The jury is still out on UE5 I would say.

 

[Scott_Arm]

Plenty of developers with employee counts in the 100s made games on UE4 and the vast majority were technically troubled in very similar ways. Andromeda was a very bad product but its problems were unique to that game. Unless I'm remembering wrong, Frostbite engine games very rarely had issues in the visuals and performance category, where as UE4 games almost always had issues in those categories specifically. It's the trend that I am drawing conclusions from. The jury is still out on UE5 I would say.

Battlefield V, Battlefield 2042, Anthem come to mind. Also massively rocky launches of Cyberpunk, Fallout 76 are other examples of big AAA games with big issues at launch though not UE. Those are just off the top of my head.

 

[techuse]

Battlefield V, Battlefield 2042, Anthem come to mind. Also massively rocky launches of Cyberpunk, Fallout 76 are other examples of big AAA games with big issues at launch though not UE. Those are just off the top of my head.

Those first 3 games all had unique problems and widely known about troubled developments. There wasn't a trend of the same problems facing the majority of Frostbite games. Again, it's the trend of UE4 games all having the same or at least very similar issues despite coming from a ton of completely different developers. Poor CPU utilization and performance, very high GPU demands for some mysterious reason and a very poor DX12 implementation.

 

[Andrew Lauritzen]

Those first 3 games all had unique problems and widely known about troubled developments. There wasn't a trend of the same problems facing the majority of Frostbite games.

I've worked on Frostbite a bit and like a lot of things about it too, but I think you're falling victim to motivated reasoning here. Star Wars Squadrons got a lot of flak for being "unoptimized" in VR as well (despite it being yet another case of people just comparing to games with baked lighting and not understanding the costs of dynamic lighting... sound familiar?). Together that list represents a lot of the games on Frostbite in recent history... what "majority" of other games are you thinking of? NFS (not sure those are going to be a good example for you)? The sports games? :S

Again, it's the trend of UE4 games all having the same or at least very similar issues despite coming from a ton of completely different developers. Poor CPU utilization and performance, very high GPU demands for some mysterious reason and a very poor DX12 implementation.

Frostbite also got a lot of flak for DX12 being slower than DX11 (despite being one of the major drivers of low overhead APIs) in every game that supported both until they dropped DX11 and made further comparisons impossible. (example )

Ultimately it's fine to look at patterns to some extent, but your phrasing betrays the narrative that you are fitting everything into at this point: obviously not *all* UE4 games had the same issues. Several were mentioned in this very thread that have been great. I'll add Satisfactory (also recently updated to UE5) and Chivalry 2 to the list, but there are plenty more especially if you go back a bit to similar time periods as those Frostbite games (several good Borderlands, Batman, Biosock, Darksiders, Mass Effect, etc. games). And those are just the popular ones I played off the top of my head... there are hundreds more (https://en.wikipedia.org/wiki/Category:Unreal_Engine_games) many of which didn't exhibit any of the problems you are concerned about here. Are you just going to call those games the "outliers", but then reverse that logic for non-UE games?

(And yeah I'm aware you can go and find a reddit thread bitching about literally any game ever, but I think it's fair to say broadly that if you apply the same criteria between different games the ones I've listed - and many others - would fall on the upper end of technically sound releases.)

I'm gonna be honest, I don't think this discussion is very productive. Everyone is just going to have their own biases and motivated reasoning, myself included. If the discussion is really to be about the games and not caring what engine they use, great. If we want to talk about engines let's stick to the tech and keep the sweeping generalizations out of it IMO.

 

[arandomguy]

I guess but we selectively remember things for a reason. The things that are bad or mediocre that fade out of memory are irrelevant. They don’t make a memorable impact on us and our feelings on them aren’t strong. Things that we remember made a strong lasting impression on us and carry more weight. If people negatively remember UE, it’s because the negative impression was so great, we’re forced to remember which is drastically worse than being an inoffensive type of bad.

But what I mean with that is how many of these games you're currently complaining about are even going to be remembered at all? Just with the current UE5 launch wave for example other than Fortnite is there any game that can achieve anything other than a niche cult status years from now? The likely noteworthy UE5 games years from now haven't launched yet.

Just for example looking back to UE3 and the games list initial games in the first few years prior to 2009 were -

Alliance of Valiant Arms
American McGee's Grimm
BlackSite: Area 51
Brothers in Arms: Highway to Hell
Frontlines: Fuel of War
Gears of War
Hours of Victory
Legendary
Mass Effect
Medal of Honor Airborne
Rise of the Argonauts
Stranglehold
Tom Clancy's Rainbow Six Vegas
Tom Clancy's Rainbow Six Vegas 2
Turning Point: Fall of Liberty
Turok
Unreal Tournament 3
Warmonger: Operation Downtown Destruction

Of course there's some extremely notables in there like Gears and Mass Effect but also I'll be honest at least for me the majority I can't tell you whether or not they had issues as I have no impression of them at all much less ever think of them these days. But I'm guessing you'd associate UE3 with Gears, Mass Effect and I guess UT3 (albeit not necessarily for the game intrinsically).

 

[arandomguy]

Plenty of developers with employee counts in the 100s made games on UE4 and the vast majority were technically troubled in very similar ways. Andromeda was a very bad product but its problems were unique to that game. Unless I'm remembering wrong, Frostbite engine games very rarely had issues in the visuals and performance category, where as UE4 games almost always had issues in those categories specifically. It's the trend that I am drawing conclusions from. The jury is still out on UE5 I would say.

Was that really the case? I think there was "complaints" (well not sure how to frame it) with DA Inquisition as well stating that Frostbite was not suitable for what they wanted to do. The engine was said to not have a lot of flexibility outside of what it was used for with Battlefield. Which is likely why Battlefield type games and due to the vehicle inclusions games that used those elements (eg. racing) did I guess work well (? not to sure, not a fan of those types of games). I believe the sports games sides did have complaints as well (and to this to this day) in particular also with respect to the animations, which was also a complaint about the RPG usage.

 

[dobwal]

Curious then, if UE is such a disaster and have been for the last 3 generations, why does it get used? It must have some properties that appeal to whomever makes the decision to use UE as their engine?
Or maybe its about money and no metric that "we" as super picky players on this forum cares about shows up on the radar for the money people. IE UE delivers on the economic front?
Or nobody got fired for buying IBM once, maybe nobody gets fired for choosing UE5?

Because it’s an easier avenue for developing games. It like asking why a Honda accord bought off the car lot can’t win many track races. It can if you are willing to invest in better racing parts that will increase performance. But not many devs are willing to invest in going under the hood of UE to soup up the engine for their games’ specific needs while competing on the cutting edge. They just need competent performance to produce modern visuals. UE does just that at a cost that’s feasible for devs that rather spend their budget on story, art and level design.

 

[techuse]

I've worked on Frostbite a bit and like a lot of things about it too, but I think you're falling victim to motivated reasoning here. Star Wars Squadrons got a lot of flak for being "unoptimized" in VR as well (despite it being yet another case of people just comparing to games with baked lighting and not understanding the costs of dynamic lighting... sound familiar?). Together that list represents a lot of the games on Frostbite in recent history... what "majority" of other games are you thinking of? NFS (not sure those are going to be a good example for you)? The sports games? :S


Frostbite also got a lot of flak for DX12 being slower than DX11 (despite being one of the major drivers of low overhead APIs) in every game that supported both until they dropped DX11 and made further comparisons impossible. (example )

Ultimately it's fine to look at patterns to some extent, but your phrasing betrays the narrative that you are fitting everything into at this point: obviously not *all* UE4 games had the same issues. Several were mentioned in this very thread that have been great. I'll add Satisfactory (also recently updated to UE5) and Chivalry 2 to the list, but there are plenty more especially if you go back a bit to similar time periods as those Frostbite games (several good Borderlands, Batman, Biosock, Darksiders, Mass Effect, etc. games). And those are just the popular ones I played off the top of my head... there are hundreds more (https://en.wikipedia.org/wiki/Category:Unreal_Engine_games) many of which didn't exhibit any of the problems you are concerned about here. Are you just going to call those games the "outliers", but then reverse that logic for non-UE games?

(And yeah I'm aware you can go and find a reddit thread bitching about literally any game ever, but I think it's fair to say broadly that if you apply the same criteria between different games the ones I've listed - and many others - would fall on the upper end of technically sound releases.)

I'm gonna be honest, I don't think this discussion is very productive. Everyone is just going to have their own biases and motivated reasoning, myself included. If the discussion is really to be about the games and not caring what engine they use, great. If we want to talk about engines let's stick to the tech and keep the sweeping generalizations out of it IMO.

I’ll drop it but if I may respond to the frostbite points. I can’t speak to Squadrons in VR, but the regular version screamed in terms of performance. Other games would be BF3, BF4, BF One, BFV, both Battlefront games etc. DX12 was bad but you didn’t have to worry about using it. DX11 ran great and scaled to as many cores as needed. To this day performance scales great on new CPUs and GPUs in all of these titles. BF2042 was the rare FB game were performance was all around dreadful. The UE games you mentioned were UE3 which always performed well. People really only complained about the similar look a lot of titles had in the beginning.

Was that really the case? I think there was "complaints" (well not sure how to frame it) with DA Inquisition as well stating that Frostbite was not suitable for what they wanted to do. The engine was said to not have a lot of flexibility outside of what it was used for with Battlefield. Which is likely why Battlefield type games and due to the vehicle inclusions games that used those elements (eg. racing) did I guess work well (? not to sure, not a fan of those types of games). I believe the sports games sides did have complaints as well (and to this to this day) in particular also with respect to the animations, which was also a complaint about the RPG usage.

Problematic for gameplay systems as the engine was not designed for RPGs.

 

[cwjs]

I’ll drop it but if I may respond to the frostbite points. I can’t speak to Squadrons in VR, but the regular version screamed in terms of performance.

50fps at 4k on a flagship gpu released 3 years after the game's target console generation launched... Doesn't look too different from ue5 game perf (on 3 year past console launch cards) to me. On 2028 gpus I bet you'll find games ue5 games targeting the ps5 scream too.