Wired Magazine: Exclusive Sneak Peek at Unreal Engine 4

[NickP]

Note: This article contains screenshots only. There is no video yet.

I woke up today and discovered this little gem in my Facebook feed. Unreal Engine 4 has been revealed to the public; and this is just a taste. Enjoy!

[Baructt]

Here's the link: http://www.wired.com/gamelife/2012/0...all/1?pid=2550

[ClimateChange]

I can't tell what is revolutionary from those static screenshots.

[darthviper107]

Dynamic lighting, high count particles with physics, high poly counts.

[NickP]

Here's the link: http://www.wired.com/gamelife/2012/0...all/1?pid=2550

Did you have a problem accessing the link in the original post?

[systemic]

The article talks about some really exciting stuff! Between KISMET 2 and the removal of the rendering process, I'm salivating. I wonder how long it will be until they release these features into UDK? A year or two?

[Baructt]

Did you have a problem accessing the link in the original post?

I didn't notice. =) My bad.

[NickP]

Polygon appears to have word that we'll see the first public demo at E3 2012.
http://www.theverge.com/gaming/2012/...es-can-keep-up

[RNG]

Ambershee was right afterall (about Kismet 2)

You'd have to living under a rock for the last 2-4 years to call that jaw dropping though It looks great, but not jaw dropping, especially when compared to Samaritan.

[JessieG]

Seriously. No more light baking. That alone is enough to sell me on this.

Also, all those screenshots are from inside the editor itself. Like if you just hit print screen with the editor out and cropped the menus out. THAT'S what is revolutionary about those screenshots. Really excited for June.

Also, Kismet 2 does sound Like it may be a really robust version of what already has a lot of features. Can't wait.

[NickP]

Seriously. No more light baking. That alone is enough to sell me on this.

Also, all those screenshots are from inside the editor itself. Like if you just hit print screen with the editor out and cropped the menus out. THAT'S what is revolutionary about those screenshots. Really excited for June.

Also, Kismet 2 does sound Like it may be a really robust version of what already has a lot of features. Can't wait.

Right?? This really jumped out at me, though:

"When Alan Willard walks the audience through the demo—complete with armored demon, dancing sparks, and rolling balls of light—the room falls still. Then the twist: Willard reveals that both the cinematic scene and the following tech demo haven’t been running off a game file but in real time from within UE4′s game editor. It’s like finding out that the actors on TV are actually tiny people living inside your set. It also helps him show that changes can be made to the game’s design and code, recompiled and executed nearly instantly—a technical feat that has been simply unheard-of in game development. And just like that, the silence in the room becomes reverent. The videogame industry has changed."

[ClimateChange]

GPU power grows exponentially in each generation, but have they already hit the point where every light can be dynamic everywhere? Reflections, refraction, transmission, penumbra, indirect light bounces just about everywhere every frame?

[darthviper107]

As long as it's more than one light at a time it'll be a big improvement

[ambershee]

GPU power grows exponentially in each generation, but have they already hit the point where every light can be dynamic everywhere? Reflections, refraction, transmission, penumbra, indirect light bounces just about everywhere every frame?

We hit that point a long time ago, however console hardware hasn't changed for nearly a decade.

[joshua.fontany]

Wow, impressed. I hope Epic decides that people are better off learning the new workflows and releases a UE4 version of UDK soon. Loved the Landscape shot, and the new Particle, Lighting and Kismet2 functions sound delicious.

[XilenceX]

This article was written so enthusiastic it got me wondering if the author has never heard of CE3 before. xDD Most of the features of UE4 we were presented with in the article are in CE3 already. Having said that I am looking forward to Kismet2 and finally no more building/baking in Unreal Engine! Can't wait to see the video in June I think it will reveal much more of the quality of the Engine than that article could.

[awakeningfromobliv]

There's noway we will get a UDK version of UE4 this year. There going to want the AAA developers to show off UE4 first. If nexgen console come out Christmas 2013, then we won't see UDK version of UE4 until sometime in 2014.

[Kridian]

There's noway we will get a UDK version of UE4 this year. There going to want the AAA developers to show off UE4 first. If nexgen console come out Christmas 2013, then we won't see UDK version of UE4 until sometime in 2014.

Do you know how many modders will "jump ship" if Epic holds off that long?

[Aldaron07]

It is not like Unreal Engine 3 or UE3-UDK will look terrible in 2014. Unreal Engine 3-made games will still look pretty good.

[JessieG]

I wouldn't be surprised if they release a UE4 Version of UDK as soon as UE4 is ready because the hardware for it is available already for PC, and considering that UDK Games can only be released on PC anyways, the power of consoles is really a moot point when it comes to upgrading UDK to UE4 tech.

There are only 2 reasons I can see them waiting a while before releasing a UDK based off of UE4:

1. They'll wait until big developers start releasing games with it so that those games can showcase the engine's potential first, instead of novice users releasing a bunch of videos of their projects and those are the first things people see done with the engine.

2. They'll wait until the price for hardware to run UE4 to its full potential is lower so the average user can afford it and the UE4 version of UDK's user base will be much bigger.

Though even something as simple as time constraints and priority could delay it for a while. Who knows.

[darthviper107]

Do you know how many modders will "jump ship" if Epic holds off that long?

Can't imagine why that would be the case. Also it's not modding, UDK makes full games.

UE4 is nice and all, but for UDK developers it's just a cool thing and not at all necessary for a while.

[JessieG]

Not necessary but definitely handy, especially considering how much faster they claimed the workflow is compared to UE3, would could prove pretty valuable to indie developers made up of a handful of people, or just 1.

The graphics, though I don't mind improvements, aren't the part that has me excited, it's the faster workflow.

[RNG]

Didn't they said something like UE4 will be exclusive to the next gen systems and UE3 will continue to be the standard for mobile/flash/indies?

I highly doubt they'll release UE4 as UDK. Maybe a few years after the release, but certainly not during the launch.

[MonsOlympus]

Do you know how many modders will "jump ship" if Epic holds off that long?

That's just obsurd, UE3's visuals will still hold up and the people who actually pay Epics bills will get their hands on better tools.

[TMichael]

Didn't they said something like UE4 will be exclusive to the next gen systems and UE3 will continue to be the standard for mobile/flash/indies?

That's exactly right. Think about it this way - as an indie developer, do you really want to limit your potential customer base to people who have Nvidia Kepler graphics and better? UE4 will be a boon to AAA developers, but for the rest of us, I suspect it will make much more sense a few years down the road.

Not that I wouldn't love to get my greedy paws on it. Ho hum.............

[Baructt]

They'll probably think like any good entrepreneurs: sell it to as many people as possible (after letting the big boys show off a little). Epic likes UDK and its selling potential, after all.

[iniside]

That's exactly right. Think about it this way - as an indie developer, do you really want to limit your potential customer base to people who have Nvidia Kepler graphics and better? UE4 will be a boon to AAA developers, but for the rest of us, I suspect it will make much more sense a few years down the road.

Not that I wouldn't love to get my greedy paws on it. Ho hum.............

And I though main point of releasing UDK in first place was to hang in more people into UE crowd. You know. More people know the tools, more license sales in future, because it's easier to hire people who know the engine.

I don't see any reasonable reason (;p), why wouldn't be it diffrent with UE4. Especially with new workflows to learn. Mid-guys (where I guess most license sales go), won't be so eager to experiment with new tech without people who doesn't know it, they simply do not have time and money for it.
I don't believe that what they get from so called indies games if considered as stable incom from them. It's nice bonus but nowhere near something to relay on.

Didn't they said something like UE4 will be exclusive to the next gen systems and UE3 will continue to be the standard for mobile/flash/indies?

We should get it ASAP. I'm running Next-Gen system right now. On my desk..

[MonsOlympus]

Is it just me or is everyone missing the fact that the samaritan demo ran on UE3, on 580gtx's? sure its not UE4 but there was a joke made about it being damn near close. Im sure theres plenty of untapped potential left in UDK if the focus wasnt on little mobile titles and more on PC development. I havent seen an UE3 title yet with a DX11 minimum requirement and Im pretty damn sure the next wave on consoles would handle UE3 just fine also so for me the upgrade does seem aimed soley at the top end.

[iniside]

Is it just me or is everyone missing the fact that the samaritan demo ran on UE3, on 580gtx's? sure its not UE4 but there was a joke made about it being damn near close. Im sure theres plenty of untapped potential left in UDK if the focus wasnt on little mobile titles and more on PC development. I havent seen an UE3 title yet with a DX11 minimum requirement and Im pretty damn sure the next wave on consoles would handle UE3 just fine also so for me the upgrade does seem aimed soley at the top end.

I disagree. UE4 wins for purlerly because it have fully dynamic lighting. And this is for me deal breaker.

[Archerx]

I really just want Kismet 2. I wish they would add it to UE3...

[MonsOlympus]

I disagree. UE4 wins for purlerly because it have fully dynamic lighting. And this is for me deal breaker.

Why is it a deal breaker, is dynamic lighting a core gameplay mechanic in your current project?

[MikeClarke]

I just want to remind people we get UDK for free, think of the manpower it took to develop the engine. Think of the old costs of picking up the engine a few years back. This is by far the best free engine in existance for many titles. We get all this power at our fingertips for nothing. They can't just straight to dropping all there new stuff into a free engine it just wouldn't be good business. The engine's reputation is going to be based on the quality of games produced from it and with all due respect most of us won't have the resources at hand to develop a triple A game with it. They need to filter the engine through to us. First they need to get it to the companies paying them a huge sum of money so they can make plenty of cash to cover the development costs. Those companies develop ground-breaking games with the engine and the world of game dev will see what amazing things have been developed. After a short while they'll filter it down into UDK for us to see what we can do with it. The current UDK hasn't been destroyed/rebuilt to its max by the community yet (although there are some awesome projects) so we still have time before this content gets old & unwanted.

UE4 will unleash major improvements specifically in the lighting & kismet department, but its not something we desperately need right now or our projects fail. Patience is a virtue

[OverRide]

Is it just me or is everyone missing the fact that the samaritan demo ran on UE3, on 580gtx's? sure its not UE4 but there was a joke made about it being damn near close.

Samaritan demo took 3 Graphic's cards to run, and the link in the post to the wired Magazine they even stated that the UE4 Demo was running on a single GTX 680 to show what it can do on what people can afford hardware.

[iniside]

Why is it a deal breaker, is dynamic lighting a core gameplay mechanic in your current project?

No. But I'm working pretty much on my own, and I just can't wait for building lighting every time I change something. And I like to change a lot, durning iteration of level. I need instant feedback on what I'm doing. Well you know what provide instant feedback so I don't need to call names. Waiting at best 5 mins, and at worst 1-2h for building lighting is simply not an option.
Another reason. My PC is not able to render llightmaps for bigger maps. 12GB RAM is simply not enough.. (yeah I can get more ram, but I don't really need right now, not just for UDK not worth it).

edit:
From what I read last time, someone on Epic said (maybe even one CEOs), they regret that they didn't come with UDK much sooner that they did.
I'm pretty sure they won't make same mistake with UE4 (unless they like to repeate mistakes they regret )

[MonsOlympus]

Samaritan demo took 3 Graphic's cards to run, and the link in the post to the wired Magazine they even stated that the UE4 Demo was running on a single GTX 680 to show what it can do on what people can afford hardware.

Thats true indeed but Im willing to put money down to say 2x680gtx's would perform just as good as 3x580gtx's, infact since texture memory isnt shared in sli setups no matter how many cards you got the single GTX680 might have more available texture memory. My point is that there is nothing higher in terms of support when it comes to DX versions, ofcoarse UE4 might use it better than UE3 since it seems more focused on DX9 but the samaritan demo is running on a version of UE we have currently available. How many people using UDK even get close to that kind of production values?

No. But I'm working pretty much on my own, and I just can't wait for building lighting every time I change something. And I like to change a lot, durning iteration of level. I need instant feedback on what I'm doing. Well you know what provide instant feedback so I don't need to call names. Waiting at best 5 mins, and at worst 1-2h for building lighting is simply not an option.
Another reason. My PC is not able to render llightmaps for bigger maps. 12GB RAM is simply not enough.. (yeah I can get more ram, but I don't really need right now, not just for UDK not worth it).

That's a fair call but theres no specs listed on the PC they used other than graphics card right? Which mean it could be a dual i7 clocked to the max with 32gb+ of ram. The current editor does try and make it so everything in the viewport is close to production visuals but honestly you shouldnt be at the lighting stage if youre still building the level, it should be done in passes and I learnt this from arch viz work. Not only that but region rendering was a life saver since you didnt want to render a 5000x9000 picture with GI and all the rest of it just to see the minor change.

I dont doubt it'll be useful in that respect but what have we lost to gain the full dynamic lights, GI? A more accurate lighting system. Theres very few technical features listed at all but Im sure they will come in time and when they do I can make a more informed decision. Right now Im very happy with UDK and its renderer, the component system makes things breezy and I barely use any of my 4 cores as it is and my ati card runs the dx11 and its hardly a GTX680.

[Archerx]

Why is it a deal breaker, is dynamic lighting a core gameplay mechanic in your current project?

I decided to bake all the lighting into my models from the 3D app to avoid the annoyance that is Lightmass. I never have to rebuild a level and it's awesome! When I was doing Bizango Blast everytime I moved an object the lighting would have to be rebuilt and when you have 60 levels it's extremely time consuming!

[ambershee]

I don't see why baking lighting in your 3d app and reimporting a load of stuff is better than baking lighting in UDK.

FYI, I can build lighting on DM-Deck on this machine in less than 30 seconds.

[Archerx]

I built the levels in the 3D app, got them to where I like them, baked, exported, placed and voila done. I get the lighting right on the first export.

I can move the stuff around with out having to rebuild EVER.

30 seconds times 10,000 plus object movements sure does add up. Just saying

Also I save memory from the lack of lightmaps, another bonus.

[iniside]

That's a fair call but theres no specs listed on the PC they used other than graphics card right? Which mean it could be a dual i7 clocked to the max with 32gb+ of ram. The current editor does try and make it so everything in the viewport is close to production visuals but honestly you shouldnt be at the lighting stage if youre still building the level, it should be done in passes and I learnt this from arch viz work. Not only that but region rendering was a life saver since you didnt want to render a 5000x9000 picture with GI and all the rest of it just to see the minor change.

Yes. But bear in mind. It's hard to switch, when you worked with everything realtime all the time. I just used to work this when I get instant feedback and I can work on everything I feel right now. More over I see instatly how assets will look in level which also very important, because I know if I can leave them as they or they need more work. You can believe me it's really hard to go back, when you try it.

As side note. I just like moving shadows. It add big deal of atmosphere to level (at least for me) and it's worth having less quality lighting for it.

I dont doubt it'll be useful in that respect but what have we lost to gain the full dynamic lights, GI? A more accurate lighting system. Theres very few technical features listed at all but Im sure they will come in time and when they do I can make a more informed decision. Right now Im very happy with UDK and its renderer, the component system makes things breezy and I barely use any of my 4 cores as it is and my ati card runs the dx11 and its hardly a GTX680.

Yeah. I'm also really curious what are real system spec required to run UE4. And I don't mean spawning milions of particles just for new lighting.
Anyway. I'm not using UDK anyway, so I don't have anything to complain ;p. But will galdy use UE4 when it's come, because of very nice setup of tools.

[MonsOlympus]

I think you'll find that alot of artists have vision, one that allows them to see assets before they are made and how they will look without even being infront of a computer. Myself I have some of that vision especially when it comes to 3d modelling (not so much detailed lighting) and also design, I find myself running test scenarios through my head before my idea is even prototyped. Having the visual feedback and input helps but I wouldnt rely on it too heavily because while having the realtime feedback can be useful for saving time if youre pushing and pulling verts (as I like to call it) all week and never get any real work done youre gonna have to answer to the big guy. If youre working for yourself you really need to question your habits and enhance workflow, I think you'll find that area (and not the tools specifically) is where big companies save their money and get **** done ontime and to a high quality