Interview Paul Neurath talks about Underworld Ascendant at PCGamesN

[sser]

Unity runs like it's Crysis but looks like it's CRAPSIS.

 

[tuluse]

Just to recap Unreal Engine 3 couldn't handle levels made for the Wolf3D engine which was significantly simpler than the UU engine. Unity is significantly slower than UE3.

I hope you like your linear corridors.

[citation needed]

http://www.rpgcodex.net/forums/inde...e-of-the-triad-game.74783/page-5#post-2800438

In a post on the Steam board, the developer says that most of the framerate problems are caused by the developers trying to be clever with an engine that is no longer built for clever developers. All the levels are built out of modular components, but UE3 is normally used for environments built of large individual models and stuff, so loading all the individual components with something called "drawcalls" is just too much for UE3's tight little anus, even though it's just the commands that bring it to its knees. It sounds almost too stupid to be true, but who knows!!!

http://infinitewrench.com/2012/09/07/unity3d-vs-unreal-development-kit/

[Unity] isn’t expected to create as ambitious a set of projects, it cuts a lot of corners that can end up crippling a project. There isn’t native level-of-detail support, you have to code that for yourself. There isn’t native culling support, it renders everything. If you put together the exact same environment in both engines, UDK will come out way ahead in terms of framerate and performance because it doesn’t cut these corners.

 

[Infinitron]

Oh, btw, assuming they chose Unreal, why wouldn't it be UE4?

 

[tuluse]

Oh, btw, assuming they chose Unreal, why wouldn't it be UE4?

I suppose it would, but I wouldn't expect much improvement for geometry in UE4. AAA video games don't care about complex levels they care about post processing effects which I expect it to be loaded with.

 

[Excommunicator]

Just to recap Unreal Engine 3 couldn't handle levels made for the Wolf3D engine which was significantly simpler than the UU engine. Unity is significantly slower than UE3.

I hope you like your linear corridors.

[citation needed]

http://www.rpgcodex.net/forums/inde...e-of-the-triad-game.74783/page-5#post-2800438

In a post on the Steam board, the developer says that most of the framerate problems are caused by the developers trying to be clever with an engine that is no longer built for clever developers. All the levels are built out of modular components, but UE3 is normally used for environments built of large individual models and stuff, so loading all the individual components with something called "drawcalls" is just too much for UE3's tight little anus, even though it's just the commands that bring it to its knees. It sounds almost too stupid to be true, but who knows!!!

http://infinitewrench.com/2012/09/07/unity3d-vs-unreal-development-kit/

[Unity] isn’t expected to create as ambitious a set of projects, it cuts a lot of corners that can end up crippling a project. There isn’t native level-of-detail support, you have to code that for yourself. There isn’t native culling support, it renders everything. If you put together the exact same environment in both engines, UDK will come out way ahead in terms of framerate and performance because it doesn’t cut these corners.

That's really not proving your specific claim about Wolf3D levels that "couldn't be handled"

I'm well aware that Unity3D has performance problems - that isn't in question - but I have a feeling you've worded it that way in order to exaggerate the extent of the problem.

If you load up everything from a wolf3d level as is and it can't be run then I'll be very surprised.

If on the other hand you took a wolf3d level and then detailed it to modern expectations and with present-day assets and that didn't run, then I won't be surprised.


The point I think is that modern engines are built for modern (see: less ambitious) mapbuilding techniques and older engines for older techniques. They're following the market in that regard.

It follows logically that modern techniques will be used with a modern engine - that is, there will probably more loading screens between sections of an otherwise large map. This is not inherently a problem, but also not ideal. The important thing is that the sum of those parts delivers the same core experience. I think they can still do that with Unity, even if it won't run as well as UE3/4. This isn't the only factor coming into their decision.

 

[tuluse]

Linear corridors are also a modem technique.

 

[Infinitron]

Well, the maps of the original Underworld can fairly be described as "corridors" for the most part. Computers in 1992 probably couldn't deal with the huge chambers, outdoor areas and verticality you got in Thief and Deus Ex.

They were hardly linear, though.

 

[Untermensch]

Just to recap Unreal Engine 3 couldn't handle levels made for the Wolf3D engine which was significantly simpler than the UU engine. Unity is significantly slower than UE3.

I hope you like your linear corridors.

[citation needed]

http://www.rpgcodex.net/forums/inde...e-of-the-triad-game.74783/page-5#post-2800438

In a post on the Steam board, the developer says that most of the framerate problems are caused by the developers trying to be clever with an engine that is no longer built for clever developers. All the levels are built out of modular components, but UE3 is normally used for environments built of large individual models and stuff, so loading all the individual components with something called "drawcalls" is just too much for UE3's tight little anus, even though it's just the commands that bring it to its knees. It sounds almost too stupid to be true, but who knows!!!

sea explained it a few pages forward:

EDIT: Found the original developer post.

If I'm reading this right it means that they are using, say, 10 meshes for a wall instead of 1. That isn't clever. That's stupid. In an engine like Unreal (and most other modern game engines) it's far more useful and performance-friendly to simply have 1 mesh for a wall, barrel, etc. and use CSG for any flat plane (like a solid floor). A smart developer would have simply created a wall mesh and then used decals and other detail objects placed on that wall to hide the fact that it's the same wall used over and over. Unreal also allows you to scale objects along all axes, flip them, rotate them, etc., so you can use the same mesh in many places where you couldn't normally (i.e. if you have a wall that's 10 feet, you can scale it down to 5 feet or up to 15 feet).

This isn't an "Unreal thing". This is something that virtually every modern 3D engine will have trouble with.

In other words, developers either were not aware of the technical limitations of their engine when they built their game, or they (more likely) ignored them. Instead of picking an engine that let them do what they wanted or modifying it, they just forged on ahead knowing that there would be major performance problems.

UE4 could easily handle a UU-type game.

 

[sser]

I personally wish more RPGs would take a crack at using the Source engine. I'm assuming Valve has a pricy licensing tag on it, though.

 

[Infinitron]

I personally wish more RPGs would take a crack at using the Source engine. I'm assuming Valve has a pricy licensing tag on it, though.

The Source engine is notoriously hard to work with, apparently. (one of the major focuses in "Source 2" is increasing usability, IIRC)

 

[WingedPixel]

The engine they're using doesn't matter nearly as much as the developer's familiarity with the engine, and of course the talent/skill of the developers. Talented devs can figure out any engine given enough time.

 

[Jack Dandy]

Hey mindx2 !
https://twitter.com/OtherSide_Games/status/559967140938137603

 

[Burning Bridges]

I wonder if three dimensionality of the dungeons can even be handled by Unity.
That was a special feature of UW that for example Wolfenstein did not have (it was flat).
Programmer instinct would be to make a prototype based on OpenGL with polygons for straight walls and meshes for the caves that is generated from random noise and/or imported from laser scanning.
Then a special physics simulation to adds and distribute the realistic water everywhere in the three dimensional dungeon, including waterfalls and underwater caves and air filled areas that can only be reached by diving through water.
But they choose the engine that has the easiest 3d modelling support I bet they probably plan (read: are only able to) to edit the dungeons by hand.
I am already pretty sure this will not turn out well.

 

[Zeriel]

The engine they're using doesn't matter nearly as much as the developer's familiarity with the engine, and of course the talent/skill of the developers. Talented devs can figure out any engine given enough time.

There's a reason AAA studios use a small number of engines. Intensive scenes that push a lot of polygons don't just automatically run well no matter what, the big name engines have years of dedicated development time put into pure efficiency. No amount of magical "developer familiarity" will ever make Unity perform at the same level as Unreal, Idtech, or Frostbite. I'm not speaking of attractiveness here, either, just performance.

(Of course, worth mentioning that Unity never had "high performance" as one of its design goals. You just shouldn't expect fully 3D games in Unity to ever feel professional.)

 

[WingedPixel]

No amount of magical "developer familiarity" will ever make Unity perform at the same level as Unreal, Idtech, or Frostbite. I'm not speaking of attractiveness here, either, just performance.

You can make your game perform shitty in any engine. Familiarity would mean the developers understand the engine's limitations, so they wouldn't build their levels in a way the engine can't handle well.

(Of course, worth mentioning that Unity never had "high performance" as one of its design goals. You just shouldn't expect fully 3D games in Unity to ever feel professional.)

I agree with the first part but not the second. The polish of the game has little to do with the engine and more the skills, experience of the devs and the time available. There may however be a strong correlation between Unity users and inexperienced devs.

For what it's worth, I don't think Unity3D is the right engine for a game like Underworld, but more due to the CPU limits it's simulation will probably need. UE4 at least lets you develop in c++ easily.

 

[mindx2]

Hey mindx2 !
https://twitter.com/OtherSide_Games/status/559967140938137603

I'm on it!

 

[Infinitron]

Hey mindx2 !
https://twitter.com/OtherSide_Games/status/559967140938137603

View attachment 4374
I'm on it!

mindx2: Keep in mind that they're going to be starting their Kickstarter very soon. You might be able to get more answers from them if you're able to visit within the next week. After that, they might be very busy. Though to be honest, they're probably already quite busy now.

 

[mindx2]

Hey mindx2 !
https://twitter.com/OtherSide_Games/status/559967140938137603

View attachment 4374
I'm on it!

mindx2: Keep in mind that they're going to be starting their Kickstarter very soon. You might be able to get more answers from them if you're able to visit within the next week. After that, they might be very busy. Though to be honest, they're probably already quite busy now.

I was thinking to go after the KS when there might actually be something to see and discuss. Going before doesn't seem like it would result in anything new that hasn't been heard/seen already. Not really interested in the pre-KS hype. The Codex wants/needs to hear and see what this game is not what it's promised to be.

 

[Infinitron]

I was thinking to go after the KS when there might actually be something to see and discuss. Going before doesn't seem like it would result in anything new that hasn't been heard/seen already. Not really interested in the pre-KS hype. The Codex wants/needs to hear and see what this game is not what it's promised to be.

I'm pretty sure they already have everything they're going to show in the KS pitch prepared, man. It doesn't magically materialize on the day the Kickstarter begins.

The idea is that you might be able to get an exclusive glimpse of that before everybody else.

 

[tuluse]

They're supposed to have a working prototype so you could also experience it in person instead of just watching a video.

Spoiler

A hands on demo if you will

 

[mindx2]

I was thinking to go after the KS when there might actually be something to see and discuss. Going before doesn't seem like it would result in anything new that hasn't been heard/seen already. Not really interested in the pre-KS hype. The Codex wants/needs to hear and see what this game is not what it's promised to be.

I'm pretty sure they already have everything they're going to show in the KS pitch prepared, man. It doesn't magically materialize on the day the Kickstarter begins.

The idea is that you might be able to get an exclusive glimpse of that before everybody else.

Just don't see it happening before their KS starts. We've got a huge snowstorm going on right now with more storms forecast for Friday and Monday. Don't think they'd want to be interviewed on the weekend but that would be the only way and still leave me time to write something coherent before the KS began.

 

[Infinitron]

Just don't see it happening before their KS starts. We've got a huge snowstorm going on right now with more storms forecast for Friday and Monday. Don't think they'd want to be interviewed on the weekend but that would be the only way and still leave me time to write something coherent before the KS began.

Oh, you wouldn't have to complete writing about it before the KS began.

It's your call though.

 

[mindx2]

Just don't see it happening before their KS starts. We've got a huge snowstorm going on right now with more storms forecast for Friday and Monday. Don't think they'd want to be interviewed on the weekend but that would be the only way and still leave me time to write something coherent before the KS began.

Oh, you wouldn't have to complete writing about it before the KS began.

It's your call though.

I'd rather wait until after the KS (or, unless they aren't overwhelmed, during it) when we all have a clearer picture of what the dream hype game will entail. This way I can also elicit the burning questions the Codex Hivemind will undoubtedly have when more info has been released. I'll gather all those questions and narrow it down to the ones to be asked. I also like the idea of afterwards as they may have more actual game to show off but that may mean waiting until late Spring or early Summer. Make no mistake, I really want to do this (and will do it if you guys can arrange it). Not only for the glory of the Codex but because UU is at the top of my own personal "Best Computer Game of ALL TIME" list and Looking Glass Studios being on my "Best Game Studio EVER!" list as well. This will be an exciting opportunity* that I just can't pass up.


*

Spoiler

Just have to reign in my raging fanboyism that day and act like I should as a member of our Prestigious Magazine/Forum... bitterly cynical with a smattering of sarcasm and nihilism...