Support NeoGAF

[CamHostage]

Unreal Engine is good, but I just wish they would focus more on performance. The fact that the matrix demo didnt run flawlessly tells us all we need to know about this era. Because the engine is so easy to create for, optimization becomes secondary. Developers add as much shit as possible from teams around the world and just knda hope it works. They gotta put more effort into lowering the performance hits

Failure for us 60 fps connoisseurs.

*cries in OLED*

That's Lumen though, I think? If a developer is using dynamic GI and lots of high-quality, bouncing light and virtual shadow maps for a photorealistic, natural look, it seems it will chug. Use more traditional lighting techniques and tricks (or even just use Lumen more selectively , bolstered by Nanite's ability to handle all the polygons you can throw at it (though it's ultimately recalculating the mesh for visibility, not actually displaying "infinite polygons",) and you should be able to have a solid framerate again with a better look still than you could achieve in UE4.

There are inherent differences between UE4 and UE5, and UE5 is still young (and UE5 Early Access didn't have full access to all the tools and functions,) so the story will probably change as more developers work in UE5. Still, UE4 isn't going away (and the big AAA games are often already not targeting 60FPS as it is, so I don't know what would change in their approach,) and UE5 is perfectly capable of 60 or even 120FPS if a developer targets that framerate.

 

[Reizo Ryuu]

Looking good, definitely will try to make some super small games with it.

 

[ClosBSAS]

LMAO the subtle message where they are saying epic is buying them to go EGS exclusive on PC.

 

[SF Kosmo]

I completely cannot deal with what a terrible public speaker Tim Sweeney is.

 

[Notabueno]

Failure for who?

Read my answers below

 

[Notabueno]

Doesn't Nanite solve precisely the complexity and optimisation of assets? I mean, not having to generate LODs manually is a huge time saver in itself.

Nanite static meshes are a complete game changer because most games dont actually have destructable environments.
So basically not needing to budget for your static meshes changes quite alot.

Name me one game that uses Nanite? Exactly...the promise is great, it's not viable for now. But also it's far from sufficient for the overall production.

 

[CamHostage]

A user running through City Sample (the Matrix Awakens map and character set, without the opening with Keanu/Carrie, also without the playable action scene,) running in the UE5 Editor. He runs it as a playable game (30FPS on a RTX 3090) and also pulls the project file apart a bit to show some of the assets and the scripting. (For sure other people will do a better teardown of this project file as the weeks go by, so look for more inside looks at the Matrix Awakens demo now that it's out on PC and available to edit.)

 

[Notabueno]

Addendum: a native project-wide procedural mesh and script repository system is exactly what UE5 needs, 10x more of that actually

 

[EverydayBeast]

Try to get unreal tournament out there, it’s an easy to pronounce name, not high risk.

 

[EverydayBeast]

Unreal tournament should get another chance people allow shovelware and it’s essentially built into unreal technology.

 

[vpance]

Fascinating how they proc gen the city

 

[Bernd Lauert]

Someone tell me what it takes to run Matrix demo at 60 fps.

 

[ckaneo]

That's Lumen though, I think? If a developer is using dynamic GI and lots of high-quality, bouncing light and virtual shadow maps for a photorealistic, natural look, it seems it will chug. Use more traditional lighting techniques and tricks (or even just use Lumen more selectively , bolstered by Nanite's ability to handle all the polygons you can throw at it (though it's ultimately recalculating the mesh for visibility, not actually displaying "infinite polygons",) and you should be able to have a solid framerate again with a better look still than you could achieve in UE4.

There are inherent differences between UE4 and UE5, and UE5 is still young (and UE5 Early Access didn't have full access to all the tools and functions,) so the story will probably change as more developers work in UE5. Still, UE4 isn't going away (and the big AAA games are often already not targeting 60FPS as it is, so I don't know what would change in their approach,) and UE5 is perfectly capable of 60 or even 120FPS if a developer targets that framerate.

I wasnt talking about 30 or 60 or 120

I just want better performance at any framerate.

 

[onesvenus]

Name me one game that uses Nanite? Exactly...the promise is great, it's not viable for now. But also it's far from sufficient for the overall production.

Are you aware that UE5 has released today? How do you expect that?
I'm sure that before the year end or early next year we will have games that use Nanite and Lumen.

Why do you think it's not viable?

 

[Tripolygon]

Name me one game that uses Nanite? Exactly...the promise is great, it's not viable for now. But also it's far from sufficient for the overall production.

You are asking someone to name one game that uses Nanite, a feature of UE5 that was just officially released today? Seriously?

 

[vpance]

Can't load the Matrix project for some reason and got a black screen with text. Says 'Nanite is used in the scene but not supported by your graphics hardware'. 6800XT and latest drivers installed.

 

[Werewolfgrandma]

Can't load the Matrix project for some reason and got a black screen with text. Says 'Nanite is used in the scene but not supported by your graphics hardware'. 6800XT and latest drivers installed.

Getting everything for it now and I'll give it a go. No clue what I'm doing but I figured out the valley demo. Cross my fingers

 

[bitbydeath]

We are less than 24 hours until the official launch of UE5.0 and we being able to download and run all 3 UE5 demos on the PC using only ~7-10 total ram without any superfast SSD, god tier I/O (direct-storage), or decompressor chip (RTXIO). Contrary to popular beliefs.

Let that sink in!

Don’t forget the PS5 demo was downgraded to run on PC.

 

[vpance]

Getting everything for it now and I'll give it a go. No clue what I'm doing but I figured out the valley demo. Cross my fingers

This is what I see after loading the project. Maybe I'll have to try it in my Win 11 install.

 

[Juza]

"Our goal to push the envelope of fidelity and to deliver the high-quality cinematic action-adventure experience"

 

[vpance]

Got it working but I had to switch to Vulkan instead of DX11/12 in the project settings. Now I have to wait for 24,000 shaders to compile lol

 

[Pedro Motta]

Fascinating how they proc gen the city

There was one user here that argued with me that Houdini wasn't being used wide spread in game dev.

Sigh.

EDIT:
https://www.neogaf.com/threads/the-...unreal-engine-5.1614111/page-5#post-264243703

 

[CuNi]

There was one user here that argued with me that Houdini wasn't being used wide spread in game dev.

Sigh.

EDIT:
https://www.neogaf.com/threads/the-...unreal-engine-5.1614111/page-5#post-264243703

Not to disagree with you, but using the creator of said tech as a proof that the industry uses it is quite pointless.

Its like saying since Astrobots playroom used PS5 adaptive triggers, that tech is used industry wide.

I think it comes down to what you and the other person clarify as wide spread. For me, it would mean adoption of more than 50% on games that could be using that tech, which it also isn't rn but obviously time will tell as more and more games being build on UE5 need to release to definitely make a statement.

 

[Pedro Motta]

Not to disagree with you, but using the creator of said tech as a proof that the industry uses it is quite pointless.

Its like saying since Astrobots playroom used PS5 adaptive triggers, that tech is used industry wide.

I think it comes down to what you and the other person clarify as wide spread. For me, it would mean adoption of more than 50% on games that could be using that tech, which it also isn't rn but obviously time will tell as more and more games being build on UE5 need to release to definitely make a statement.

What the hell are you saying? DO you even understand what are you talking about?

Houdini is a tool developed by SideFX, it has nothing to do with Epic Games. It's a DCC app that uses procedural workflows to achieve higher levels of detail in a very short amount of time.

EVERY studio either big or small is using it or at least considering using it.

 

[CuNi]

Houdini is a tool developed by SideFX, it has nothing to do with Epic Games. It's a DCC app that uses procedural workflows to achieve higher levels of detail in a very short amount of time.

EVERY studio either big or small is using it or at least considering using it.

Do you have anything to back this up, especially your claim that every studio big or smal is either using it or considering using it? Because when I Google for games that use houdini, you barely get a list of 11 with the games on said list being spread across at least 5 or even more years.

1 game a year would not be "every studio" nor "widespread use in game development".

 

[Tripolygon]

Not to disagree with you, but using the creator of said tech as a proof that the industry uses it is quite pointless.

Its like saying since Astrobots playroom used PS5 adaptive triggers, that tech is used industry wide.

I think it comes down to what you and the other person clarify as wide spread. For me, it would mean adoption of more than 50% on games that could be using that tech, which it also isn't rn but obviously time will tell as more and more games being build on UE5 need to release to definitely make a statement.

Houdini is used across the industry. It is just about the gold standard in terms of procedural systems to build a variety of assets and buildings. It is also used to create baked simulations and visual effects.

 

[CuNi]

Houdini is used across the industry. It is just about the gold standard in terms of procedural systems to build a variety of assets and buildings. It is also used to create baked simulations and visual effects.

Again, I'm not claiming it's not what you guys say since I have no industry knowledge about that, but so far all I get is "trust me bro" arguments. Like I said, when I Google it, I only found one or two lists of games that use it and the releases of those games were spread across multiple years which would average out to 2 or 3 games per year. If that is really all there is to it then it is far from being in widespread use.

So my question is, how do you guys back up your claims?

 

[CamHostage]

Do you have anything to back this up, especially your claim that every studio big or smal is either using it or considering using it? Because when I Google for games that use houdini, you barely get a list of 11 with the games on said list being spread across at least 5 or even more years.

You wouldn't generally list your work as a "game that uses Houdini" if you were a studio, I wouldn't say? You would just use it. It's not like say a physics engine where it's a defining aspect of a project or something you need to promote in order to use it, where it's like has to have slapped "Powered by Havok" on the front screen of the game. Maybe Houdini would come calling for a chance to promote its engine using your work with it and so people would know, but otherwise it's a tool in the toolbox. And I don't know what evidence you'd need as proof since it's not something talked about unless you get deep into like Gamasutra post-morts or GDC tech talks or deve forums , but it's in a lot of professional toolboxes.

Game Development | SideFX

Game developers face a growing demand to build huge immersive worlds populated with vast amounts of content. In order for studios to achieve great gameplay while meeting deadlines and containing costs, a procedural approach using Houdini offers a smarter way to create quality games.

www.sidefx.com

It's hard to tell what point Black_Stride was trying to get at, maybe it was more in conversation or in favor of an alternate offering. (Side FX has a bunch of different products in the Houdini family so I don't think it matches up, but Blender is probably an alternative people have heard of and may be what Black_Stride was saying is what he considers as the mainstay in gaming?)

 

[Pedro Motta]

Do you have anything to back this up, especially your claim that every studio big or smal is either using it or considering using it? Because when I Google for games that use houdini, you barely get a list of 11 with the games on said list being spread across at least 5 or even more years.

1 game a year would not be "every studio" nor "widespread use in game development".

Oh gee I don't know, maybe because I'm their client for over 10 years and use the software every day? Maybe because I've been to countless conferences and presentations about the new versions and am aware of their client statistics? Maybe because I watch lots of GDC presentations and each year more and more game studios are using it?

I can assure you it not 1 or 2 per year. Almost all Sony studios use it, Kojima productions uses it, From Software uses it, and these are just high profile ones.

Houdini is used across the industry. It is just about the gold standard in terms of procedural systems to build a variety of assets and buildings. It is also used to create baked simulations and visual effects.

It's not worth it.

 

[Notabueno]

You are asking someone to name one game that uses Nanite, a feature of UE5 that was just officially released today? Seriously?

STFU

Are you aware that UE5 has released today? How do you expect that?
I'm sure that before the year end or early next year we will have games that use Nanite and Lumen.

Why do you think it's not viable?

No. We've been testing it for a year and privileged/major publishers for a bit longer, especially since this is based on virtualized primitive shaders which is a bit older, here's a random topic from a year ago: https://forums.unrealengine.com/t/nanite-and-displacement-tesselation/231500?page=2

It's not going to be viable for a few years unless using solely Quixel assets, the reason why being the pipeline technology is not streamlined and integrated in ways that are viable in terms of yield, production or distrib.

Especially the texture "virtualization" in terms of buffering and "streaming" from SSD cache. It's like the majority of Nvidia vaporworks, say Flex or VGXI which seemed like great solutions ootb but of course...they were nowhere near viable for a actual shipped product.

 

[onesvenus]

No. We've been testing it for a year and privileged/major publishers for a bit longer, especially since this is based on virtualized primitive shaders which is a bit older, here's a random topic from a year ago: https://forums.unrealengine.com/t/nanite-and-displacement-tesselation/231500?page=2

It's not going to be viable for a few years unless using solely Quixel assets, the reason why being the pipeline technology is not streamlined and integrated in ways that are viable in terms of yield, production or distrib.

Nobody would release a game with a beta version of an engine, I'd say that's common sense. Expecting games to use it now is quite crazy IMO.
And about not being viable: Stalker is releasing in December. I'd bet money that we'll start seeing UE5 games being released next year.

Let's say that I don't share your pessimistic view at all

 

[Notabueno]

Nobody would release a game with a beta version of an engine, I'd say that's common sense. Expecting games to use it now is quite crazy IMO.
And about not being viable: Stalker is releasing in December. I'd bet money that we'll start seeing UE5 games being released next year.

Let's say that I don't share your pessimistic view at all

Everybody will switch to UE5, because it's just a new update of UE4, and a compatibility aside, there's no reason not to. But for the purpose Epic set, the over-advertised new tech and the need of the industry, UE5 fails to be the solution is overhypes itself to be, and this for at least not just 5 years but maybe a decade.

 

[VFXVeteran]

Name me one game that uses Nanite? Exactly...the promise is great, it's not viable for now. But also it's far from sufficient for the overall production.

I agree with you here since there is a lot more R&D to be done with actually incorporating it inside of runtime game loop with physics and collision detection. However, you are out of place for speaking so negatively about UE5 since it's the only hope for next-gen console gamers that declare they will see a giant leap in graphics fidelity this generation.

 

[Notabueno]

I agree with you here since there is a lot more R&D to be done with actually incorporating it inside of runtime game loop with physics and collision detection. However, you are out of place for speaking so negatively about UE5 since it's the only hope for next-gen console gamers that declare they will see a giant leap in graphics fidelity this generation.

Look to be precise: we'll start seeing Nanite, and maybe Lumen, in a about 3-5 years inside of shipped game. Not only it doesn't change THAT much graphically but also it doesn't solve much of the growing hurdles of creating video games for small or big studios.

UE5 needed 100x more UX tooling research and conception for it to really be a change. I think that's what they're trying to do with the procedural mesh system, but they haven't show much of it, and it's not that much more either.

 

[VFXVeteran]

Look to be precise: we'll start seeing Nanite, and maybe Lumen, in a about 3-5 years inside of shipped game. Not only it doesn't change THAT much graphically but also it doesn't solve much of the growing hurdles of creating video games for small or big studios.

Yes, but I don't think UE5 is supposed to create every tool that a studio is going to need to make full use of it's engine. The studios have positions for pipeline developer for a reason. The hurdles are big though. The budgets are blowing up to film production budgets and that's obviously going to slow things down on any development.

UE5 needed 100x more UX tooling research and conception for it to really be a change. I think that's what they're trying to do with the procedural mesh system, but they haven't show much of it, and it's not that much more either.

I also think that the GPUs aren't there yet for any real significant changes in graphics fidelity. The 2D texture space pipeline is dead to me and needs to be dumped. It's just that going full 3D RT is too much for even the highend graphics cards (i.e. 3090). We need more bandwidth and more VRAM for any real big change in graphics fidelity. The path tracing demo that was just released with UE5 is the kind of quality that I see as being a big jump. Not a game like R&C for example.

 

[onesvenus]

Look to be precise: we'll start seeing Nanite, and maybe Lumen, in a about 3-5 years inside of shipped game

I'd bet money that's not true. Even if mesh fidelity doesn't grow, it makes no sense not to use Nanite on static objects.

 

[Stamps1646]

I really enjoyed this presentation.

UE5 is looking very nice, I love it when gaming engines evolve.