Game Engine Convergence And The Problem With UE5
- Thread starter BitByte
- Start date
From that statement.Now it just gives off the "low color depth" vibe
I like the 2013-2023 one best but I'm biased.
Frenetic Pony
Veteran
Now it just gives off the "low color depth" vibe
I like the 2013-2023 one best but I'm biased.
It's a retro game engine now, banding and low precision vibrating triangles are features!
Wasn't Frostbite pretty bleeding edge when EA made it's decision? How has progress been and why isn't it comparable to UE and the preferred choice? Or is it, and this is more a case of the effort of getting new staff on board to use a proprietary engine versus picking up any of the existing UE users?
It is invariably going to be engineers - if you go on linkedin, you can see they bled engineers the last few years to many studios.
Silent_Buddha
Legend
These are inter-related. The pool of quality engine engineers (or hell almost any engine specialized software engineer) are a resource in the industry that appears to be shrinking (retirements, moving on to other industries, etc. greater than incoming people qualified or interested in being an engine architect) leading to poaching from larger and more well funded development studios and/or needing to offer larger salaries in order to attract interest in those positions.
DICE while being an EA studio has funds to attract engineers if they want (no clue how much they are offering right now) but they lost a lot of key engine architects a few years back when Repi and others left DICE to form their own studio. IMO - the engine was better technically (in terms of stability, tech relevant to X time period, etc.) before they left. For a while they were able to get by as that legacy framework was so good that even minor improvements still allowed for good looking and performing BF games. The latest game, however, was a technical mess and, IMO, didn't really stand out compared to competing FPS games as much as previous DICE games. I'm interested to see how the next BF game turns out. Will it reverse the trend or will it continue to get worse?
Deck 13, as noted suffered due to being a smaller studio with a more restricted budget that was unable to match competing offers from studios willing and able to offer more money to poach their engineers. I really like the studio and have loved seeing the progress they've made WRT game development, polish and presentation over the years, so I'm kind of sad to see them bleed so much talent.
Regards,
SB
Frenetic Pony
Veteran
What decision? I'm a bit lost here as Deck 13 doesn't have a whole lot to do with EA
Regardless Frostbite was bleeding edge, better than UE4, but suffered a major loss of engineering talent for a while. EA just kind of stopped making games other than sports ones mid way through last gen (other than Respawn). As I understand it, it was this terrible set of proscribed rules around performance reviews and bonuses and etc. for projects. They'd strictly budget absolutely everything then decide what your entire team was worth solely based on absolute maximum ROI.
The end result was games cancelled left and right, no Dragon Age sequel, Andromeda shoved out fast (sure they had "years", under very bad management then sprinted to actually make the game in 18 months or so), several Star Wars games started then cancelled, etc. etc. by nervous middle managers who didn't think any game at all would have enough ROI, so doing nothing was better than actually making any project at all.
Vice president of gamedev went off with most of the engineering talent to form Embark, which now actually makes games like The Finals, which already appears to be super successful. The end result is EA has had little in the way of games outside Respawn and Sports, and relatively little work on Frostbite until recently. They've sprinted to try and get back into shape, there's finally a WYSIWYG editor for a lot of functions in Frostbite instead of external DCC tools, finally a modern streaming architecture that came on with Dead Space remake, etc. But they're still vastly behind UE5, thus internal EA projects like Iron Man just going to UE5.
I'm not entirely sure what the future of Frostbite is going to be. I know the internal research group has switched almost entirely to machine learning tools, all of which could just be done as proprietary plugins for UE5 or Frostbite. Frostbite still has an internal dedicated team and the new Dragon Age (first one in a decade) looks great. But I don't know what the future really holds there.
EA just announced devs no longer need to use Frostbite and can use other engines if they'd prefer.
I do wonder if an alternate timeline they'd decided not to build the new Battlefield around 128 players and ultra-sized maps and a 'live service' sort of thing that was expected to be 'the' Battlefield game for foreseeable future(and so might have actually properly prepared for work on a new next gen entry after this initial cross-gen title released), that things might have ended up more promising. Not only do I think they could have made Battlefield 2042(or whatever it might have otherwise been) more technically impressive due to more restrained gameplay scope, but might have also given them more motivation to keep Frostbite at the cutting edge.
I expect these things likely had as much or more to do with the state of things than simply people leaving. Feels a lot like a similar situation to Halo Infinite. These devs/pubs keep wanting to stop having to develop new games and instead just make one, long-term game that brings in endless season pass+microtransaction money. I get the attraction, but it really seems to be hurting studios and publishers who try and make this switch more often than not.
I expect these things likely had as much or more to do with the state of things than simply people leaving. Feels a lot like a similar situation to Halo Infinite. These devs/pubs keep wanting to stop having to develop new games and instead just make one, long-term game that brings in endless season pass+microtransaction money. I get the attraction, but it really seems to be hurting studios and publishers who try and make this switch more often than not.
It's not just game design. I'm sorry, but this is superficial thinking.
Picking a physics engine and exploring each possibilities of it and making it work like it did for sure required a cooperation between engineers and game designers. Also, I would guess that the physics engine was modified by Nintendo. Devs had already told that TOTK was done by 2022, they did not specify what they spent a year working on, but for me it's clear that they spent time making sure everything was working well on that side. That part is not with the game design staff, this is made by the engineering team, on the technical side.
Using Havok (maybe a modified version of it) does not make what they made on the game less impressive.
This is an acccomplishment on the game design side and the technical side of gaming.
Frenetic Pony
Veteran
Nah, it's just game design. Game design is the hardest problem there is in, well, games. But the physics in Tears of the Kingdom has been done to death by games of all shapes and sizes down to indie titles like Besieged, which is basically Tears of the Kingdom's machine building but years earlier. 0 special about its implementation and 0 customization is needed to achieve anything like it.
Zelda has had top tier game design for decades now. Tried playing Mass Effect 2 (remastered) just yesterday and damn is that awkward, but I can boot up my Switch and Link's Awakening (now 30 years old, soon to be+) is still a perfectly enjoyable game.
I was curious as to whether 3dmark is doing anything interesting in its latest Speed Way benchmark and came across the somewhat sparse description of the āengineā.
support.benchmarks.ul.com
They describe their GI solution as being based on dynamically updated light maps in texture space. It sounds very similar to Lumen but likely less advanced as the benchmark consists of a very small, mostly static indoor scene.
The days of 3dmark being a cutting edge benchmark seem to be long over. There are actual games shipping with more advanced graphics than 3dmark can muster.
Speed Way engine
In 3DMark Speed Way, we have improved our existing illumination pipeline by removing its reliance on any precomputed illumination data (e.g., lightmap and probes), making it completely dynamic. To highlight dynamic illumination, we limit further ...
support.benchmarks.ul.com
They describe their GI solution as being based on dynamically updated light maps in texture space. It sounds very similar to Lumen but likely less advanced as the benchmark consists of a very small, mostly static indoor scene.
The days of 3dmark being a cutting edge benchmark seem to be long over. There are actual games shipping with more advanced graphics than 3dmark can muster.
How much money is there in 3DMark? Presumably it doesn't warrant the kind of investment game engines do.
Well whatever money there is depends on people finding the benchmark relevant and therefore willing to pay for it. If the benchmark isnāt useful/relevant then there eventually wonāt be any money at all.
It's wild that they're using ROVs to render OIT since that goes against several IHV recommendations ...
I was curious as to whether 3dmark is doing anything interesting in its latest Speed Way benchmark and came across the somewhat sparse description of the āengineā.
Speed Way engine
In 3DMark Speed Way, we have improved our existing illumination pipeline by removing its reliance on any precomputed illumination data (e.g., lightmap and probes), making it completely dynamic. To highlight dynamic illumination, we limit further ...
They describe their GI solution as being based on dynamically updated light maps in texture space. It sounds very similar to Lumen but likely less advanced as the benchmark consists of a very small, mostly static indoor scene.
The days of 3dmark being a cutting edge benchmark seem to be long over. There are actual games shipping with more advanced graphics than 3dmark can muster.
I think it compares well to current games in terms of tech.
Can we make a list of all the games released that use Unreal Engine 5 so far with a reference on whether they are using Lumen and/or Nanite?
That would help us evaluate adoption of the engine, if its main key features are used and get also an idea of the quality we are getting in the visuals.
Tech demos are super impressive. But games that look just as impressive that use UE5 are scarce.
That would help us evaluate adoption of the engine, if its main key features are used and get also an idea of the quality we are getting in the visuals.
Tech demos are super impressive. But games that look just as impressive that use UE5 are scarce.
Similar threads
- Locked