Unreal Engine 5

If he make shenmue 4 he will use ue5 for sure since ue4 project will be easy to port to ue5

Exactly.
Probably It would be easy to port projects from UE4 to UE5, it would also work in a similar way (no need to make things too complicated for developers), YSNet would need to learn only how to manage the new features...

Now I only hope that UE5 will not have the same streaming issues of UE4 like we seen in Shenmue 3 and Final Fantasy VII.
 
It's not that easy, but certainly possible (unless they are using a lot of dependencies that will be changed/slashed).
Sadly I don't think the timing will be right.
Hopefully they will.
 
It's not that easy, but certainly possible (unless they are using a lot of dependencies that will be changed/slashed).
Sadly I don't think the timing will be right.
Hopefully they will.

In any case it would be a lot easier than learning to use a totally different engine.
 
Yeah, for sure.
And they'll try to make it as easy as possible to update, so who knows.
The timing is the thing, if they are already doing work for S4 targetting 2022, it'll be kinda crazy to upgrade that late.
 
View attachment 7843
Ue5 will have blueprint and everything that ue4 does but with new feature and more powerful so i really dont think he will stay with ue4 View attachment 7844
Thank you for the information. This means that if Yu Suzuki and Ysnet are already working in Shenmue 4 on UE4 that they can easily port their work to UE5 or even port the assets from Shenmue 3 to use in Shenmue 4 to UE5.
 
Mighty impressive. Would've been nice to see an environment that wasn't just varying shades of brown, though. Some of it came across as almost untextured. Still, it's just a demo, not a real game. The possibilities are exciting.

The only downside to devs leaning heavily in this "micro geometry" feature is 30fps. The dream that we'd escape 30fps with PS5/XSX at resolutions above 1080p, and 60fps would once again become the standard, doesn't seem likely. There's this, and ray-tracing...these techniques just take too much juice. My bet is we'll end up to a very similar performance split we have now.

Will be interesting to see if you can use this strategically throughout a game instead of using it for the entire thing.
 
Now I only hope that UE5 will not have the same streaming issues of UE4 like we seen in Shenmue 3 and Final Fantasy VII.
This is not a UE4 issue, this is a hardware bottleneck issue. If you played S3 on a PC (and pretty much any UE4 game for that matter), you will not encounter texture streaming issues.

In any case it would be a lot easier than learning to use a totally different engine.
It is using a totally different engine. It requires a totally different workflow; S3 relied heavily on textures and UE5 relies heavily on geometry, UE5 has a brand new lighting engine, and a new particle system (which admittedly is in the latest version of UE4). All of these require massive shifts in development pipelines and/or learning curves to take advantage of.

This means that if Yu Suzuki and Ysnet are already working in Shenmue 4 on UE4 that they can easily port their work to UE5
I think there's a programmer or two who would disagree with this...
 
The assets are always porte
It is using a totally different engine. It requires a totally different workflow; S3 relied heavily on textures and UE5 relies heavily on geometry, UE5 has a brand new lighting engine, and a new particle system (which admittedly is in the latest version of UE4). All of these require massive shifts in development pipelines and/or learning curves to take advantage of.

It's not tho, they said the codebase will not drastically change, there will be changes on APIs for sure, but there are every minor release (and I don't think S3 really delved too much into UE source).
The lighting will just be an improvement, they'll probably keep the static lighting anyway, but it will streamline the whole workflow.
Niagara has been in UE for a while.
It still supports textures and lods, just that with that much geometry there's no need for normal maps.
There will be learning curves, and probably will not take full advantage of the tech (nobody uses raw 3D for games), but I don't think they'll make it that much harder to update (well, they said the contrary).
Do you really think they'll leave thousands of users in the dust ?
 
(nobody uses raw 3D for games)
Yet... If Epic is legit about this (and I don't see why they wouldn't be), that would actually save time in certain pipelines because you wouldn't have to re-topologize and bake maps. But I think that they're viewing UE4 less as a game engine and more of an all-purpose real-time rendering engine. There's no way they're not targeting VFX houses with the emphasis on high poly geometry.

Do you really think they'll leave thousands of users in the dust ?
Well the engine won't see an official release until 2021 with partners likely only getting it early if they're high profile and/or going to push the limits (like, say, Square), plus the issues of early adoption that I raised before; I just don't see it happening. There's no advantage. If they were starting development in 2021 then maybe, but then don't expect to see a release until 2024-25 at the earliest.

From a technical standpoint; it's not a huge shift but from a development pipeline standpoint, especially for a relatively low-budget sequel; there's a lot of headache.
 
Taking the conversation from here: Yes, UE4 projects can be ported to UE5. There's still the need to see a need for Shenmue IV though. As far as UE4 goes, this isn't what limited Shenmue III's visual fidelity. Visuals, models, textures, draw distances, crowd density. This isn't on UE4 if Shenmue 3 ended up where it is. It was more of a matter of devs being able to do those things while retaining good performances rather than UE4 not being able to process that. Now there could be other advantages when it comes to ease of developpement. But when it comes to visual, what we got was what the team could bring. Now of course with Qixel being included with Unreal Engine, that could be another matter for environnements... althought it'd mean using base assets.
 
This is not a UE4 issue, this is a hardware bottleneck issue. If you played S3 on a PC (and pretty much any UE4 game for that matter), you will not encounter texture streaming issues.


It is using a totally different engine. It requires a totally different workflow; S3 relied heavily on textures and UE5 relies heavily on geometry, UE5 has a brand new lighting engine, and a new particle system (which admittedly is in the latest version of UE4). All of these require massive shifts in development pipelines and/or learning curves to take advantage of.


I think there's a programmer or two who would disagree with this...
Im an game dev on both unity and a ue4 and yes porting asset to different version of ue4 is hard sometime some asset are broken because of.that but with ue5 they are working on it and it wont be so hard since a lot of game are currently using ue4 for the next gen and they will migrate to ue5 for sure even themself with fortnite and no its not really a different engine its like a new version of ue4 but with a lot of new feature and improvement but if you know of to use ue4 you will know how to use ue5 just like when unreal engine 4 was release it was similar to ue3 (udk) or when unity lunch unity 5 it was also similar to unity4 but with a lot of new thing , the physics engine will be the same , blueprint will be here for scripting , the particule system will be also the same etc.. but with a lot of new feature as i said
 
Yup so as Kenzo stated this isn't the same as the move from UE3 to 4. It is a upgrade - also development could be scaled to target the UE5 engine. But the auto optimisation is a huge deal - if thats legit (and epic claim its real, running on a PSV dev kit and they simply captured the HDMI output) - its a massive change for mid-sized development teams. The lighting is also fantastic but we sort of knew it was coming. (Get ready for UE5 "remasters" of every UE4 PS4/XbOne title!)

Also lets be honest here folks Shenmue IV will probably only be coming to next gen consoles as the game is a minimum of 2 years away. Especially considering the issus they had scaling S3 down to run on current gen (it runs like a dog in places), its actually very work load intensive on current gen to optomise games to run - there's arguably more of a chance of getting Shen4 on current gen with UE5 than UE4.

I think this could be a huge leap for smaller teams to produce more photo realistic work and could even setup Shenmue for a return to a more "real" character style (A man can dream - right?).

I think looking at those improvements with a lot of the game mechanics written in Unreal engine already it would be foolish not to target the step to UE5 for Shenmue IV. Unless Suzuki is planning a VERY quick turn around (which I'm not really sure is possible without another team working on a game).
 
Im an game dev on both unity and a ue4 and yes porting asset to different version of ue4 is hard sometime some asset are broken because of.that but with ue5 they are working on it and it wont be so hard since a lot of game are currently using ue4 for the next gen and they will migrate to ue5 for sure even themself with fortnite and no its not really a different engine its like a new version of ue4 but with a lot of new feature and improvement but if you know of to use ue4 you will know how to use ue5 just like when unreal engine 4 was release it was similar to ue3 (udk) or when unity lunch unity 5 it was also similar to unity4 but with a lot of new thing , the physics engine will be the same , blueprint will be here for scripting , the particule system will be also the same etc.. but with a lot of new feature as i said
It's certainly possible, but I'm hoping Suzuki and co. are already in pre-production on S4, I would hope/expect development to be well under way by the time UE5 releases next year especially if:

Also lets be honest here folks Shenmue IV will probably only be coming to next gen consoles as the game is a minimum of 2 years away.
If they change to UE5, you can absolutely forget about a 2022 release date, that's a 1 year turnaround after porting to a new engine. Unless nothing breaks and everything works perfectly (which happens all the time in game development /s): that's impossible. 1 year out from release, you're starting to ramp up press releases, trailers, etc. not changing the engine.

I'm just speaking from experience having actually gone through this process. Game development, even for small teams, is a very slow moving ship when you actually have deadlines and budgets and publishers to answer to.

I think looking at those improvements with a lot of the game mechanics written in Unreal engine already it would be foolish not to target the step to UE5 for Shenmue IV.
Another thing to consider is, the further the release date is pushed (say 2024), the more Shenmue runs the risk of looking even worse by comparison. 4 years from now we're going to be getting the latest Naughty Dog and Rockstar games showcasing the power of next gen consoles. If Suzuki really wants to court a new audience, that's going to be tough to do with such stiff competition. Lots to consider.
 
Last edited:
It's certainly possible, but I'm hoping Suzuki and co. are already in pre-production on S4, I would hope/expect development to be well under way by the time UE5 releases next year especially if:


If they change to UE5, you can absolutely forget about a 2022 release date, that's a 1 year turnaround after porting to a new engine. Unless nothing breaks and everything works perfectly (which happens all the time in game development /s): that's impossible. 1 year out from release, you're starting to ramp up press releases, trailers, etc. not changing the engine.

I'm just speaking from experience having actually gone through this process. Game development, even for small teams, is a very slow moving ship when you actually have deadlines and budgets and publishers to answer to.


Another thing to consider is, the further the release date is pushed (say 2024), the more Shenmue runs the risk of looking even worse by comparison. 4 years from now we're going to be getting the latest Naughty Dog and Rockstar games showcasing the power of next gen consoles. If Suzuki really wants to court a new audience, that's going to be tough to do with such stiff competition. Lots to consider.
In my oppion there is no way Ysnet and Shenmue can compare or compete with powerhouses like Rockstar or Naughty Dog and their games and nor do I want to.
If Yu Suzuki wants to attract a wider audience for Shenmue 4 and the series he first needs to tell a great story with memorable moments and with great characters. For it to have a great impact Ysnet must improve the facial animations and someof the character models. He also needs to improve the gameplay and the fighting in general.
QTE's must also be improved and must definitely branch out like in the previous titles.
It also needs to be more variety in terms of mini games and arcades; he also needs to bring back the sega arcade classics.
All of these features must be modernized but at the same time retain the Shenmue Dna.
Shenmue is never going to be again at the cutting edge of technology both in terms of graphics and gameplay and I'm fine with it.
The new players that will eventually fall in love with the series will almost certainly love it because of the same reasons we did 20 years ago and that is mainly because of the story the amazing cast of characters and its slow paced gameplay, but without being stuck in the past.
 
@iknifaugood - Yes it does depend on two things -

1. How much of a "new" game are they trying to make - they could do a Shenmue 2 esque same engine, new places and not much in the way of new mechanics just fixing some flaws from the first game. If thats the case then UE4 away, depending on where they were with the bits they could have a lot of assets etc. in place.

2. How easy it is to transfer up to the new engine, I know its marketing spiel but this is meant to be more like moving to a new version of UE4 than the previous generations which did require a lot of "retooling" (Im sure there is a better word than that but it slips my mind!). Its a trade off, if they don't have to work on LOD maps etc. for the whole world and let the engine handle it theres a question to be had about banking that time and use it to move to UE5.
 
Another thing to consider is, the further the release date is pushed (say 2024), the more Shenmue runs the risk of looking even worse by comparison. 4 years from now we're going to be getting the latest Naughty Dog and Rockstar games showcasing the power of next gen consoles. If Suzuki really wants to court a new audience, that's going to be tough to do with such stiff competition. Lots to consider.
Attempting to compete with AAA games (with 10 times the budget and teams of 500+ people) seems like a fool's errand. In that regard I agree with @Sergeynest. They should put a focus on bringing the characters to life, on writing, on stronger combat mechanics...all of that should come before utilising cutting-edge graphic techniques. Look at games like Vampire the Masquerade: Bloodlines 2 -- the characters looked like ass during that Xbox Series X showcase, but MS chose to show it anyway, because they know the graphics aren't the selling point of that game.

However, I don't agree that the game needs branching QTEs or SEGA arcade classics. QTEs are cool but they're extremely expensive and not worth pumping extra money into for branching paths. They just aren't a selling point anymore. Make them easier so you don't die on 90% of them -- problem solved. As for SEGA arcades, I love OutRun as much as the next guy, but including those same games again feels like a token gesture more than anything. It'd be nice fan-service. Is it needed now we have fishing, herb-collecting, possibly other new side stuff? Not for me, personally.
 
Attempting to compete with AAA games (with 10 times the budget and teams of 500+ people) seems like a fool's errand. In that regard I agree with @Sergeynest. They should put a focus on bringing the characters to life, on writing, on stronger combat mechanics...all of that should come before utilising cutting-edge graphic techniques. Look at games like Vampire the Masquerade: Bloodlines 2 -- the characters looked like ass during that Xbox Series X showcase, but MS chose to show it anyway, because they know the graphics aren't the selling point of that game.
In my oppion there is no way Ysnet and Shenmue can compare or compete with powerhouses like Rockstar or Naughty Dog and their games and nor do I want to.
Oh I agree and I wasn't saying that S3 should attempt that level of fidelity; exactly the opposite. What I'm saying is that we know Suzuki wants to court a wider audience with S4, we know that S3's graphics are already behind the times, and presumably they're going to want to use the same assets for S4. Factoring all that in, it seems a mistake to me to push the release date into 2024+ (where the assets will look even more dated) just to take advantage of some of UE5's features which may or may not even be worth it for this particular game.

QTEs are cool but they're extremely expensive and not worth pumping extra money into for branching paths. They just aren't a selling point anymore.
This is very true. It's kind of built into Shenmue's DNA but I would rather they focus on the fighting system and incorporate QTEs that way (maybe as counter attacks or context sensitive moves a la Doom's glory kills), that way they'll be less expensive and they can be reused. That being said, I was shocked replaying S1 and 2 how much better QTEs are in those games than most other games (even recent ones). But (like all things) they shouldn't be in future games just because they were in the old games.
 
Last edited:
Seeing as the team had such trouble learning to use UE4 for S3 (their own admission, not my criticism), it would be asinine to try and learn UE5 -- cost, learning time involved, and so on-- for S4.

That said, this looks delicious. (It better be used in the next version of Monopoly; I wanna see Ventnor Avenue pop!)


They did a better job than Final Fantasy VII Remake did
 
It's also worth mentioning that as soon as you increase quality to a certain level (much less this level) everything else needs to match. Could you imagine how out of place the facial animation and stiff movement would look if everything else was basically photo real?
Well this tech demo itself showed stylised models can work in a realistic background, that female character model clearly has an exaggerated anime esque face which I thought was an interesting choice for the designers to go with since the backgrounds are ultra realistic. But yeah the current models of Ryo and Shenhua would look completely out of place in the tech demo's environment though, even with the oriental influence.
 
Well this tech demo itself showed stylised models can work in a realistic background, that female character model clearly has an exaggerated anime esque face which I thought was an interesting choice for the designers to go with since the backgrounds are ultra realistic. But yeah the current models of Ryo and Shenhua would look completely out of place in the tech demo's environment though, even with the oriental influence.


No doubt their models would get upgraded if put through UE5
 
Back
Top