awesome. design and art looks incredible, but I suppose it doesn't really show much from the engine, apart from size and detail. 

but I suppose it doesn't really show much from the engine

I dunno, it looks like it's prerendered. No jaggies in sight, great reflections. Motionblur is very convincing. UE4 has an amazing realtime lighting system (voxel cone tracing). 

gaaaaammmmeeeepplllaaayyyy.

(seriously tho, if anyone has evidence that the "next generation" of tentpole action games aren't just going to be even shorter, even dumber versions of what we have now, I'm all ears...) 

Close-up quick time events where you see sweat running down the named protagonist's face, blood on his bulky armor and slow motion death cameras that rotate closely around models. 

We can't sell engines based on gameplay videos. How would that work? Here's an innovative and amazing game idea! It ... comes with the engine?

:) 

That said, here's a video showing what you can do with the new Kismet system in UE4:

http://www.youtube.com/watch?feature=player_embedded&v=IReehyN6iCc 

Damn nice looking movie engine there, too bad Unreal Engine won't be used to make games anymore.

:P 

We can't sell engines based on gameplay videos. How would that work? Here's an innovative and amazing game idea! It ... comes with the engine?

Completely disagree. As a designer, I want to see what possibilities the new technology provides for player interaction with the world.

I want to see what sort of awesome physics is possible. I want to see new possibilities for dynamic environments and procedurally generated content, to name just a couple of things.

All that demo showed me is improvements in real-time rendering in a cinematic sequence.

That kismet system looked wicked cool tho, for non-programmers to make some game logic I guess. 

Gameplay does not sell games, visuals and brand names sells games. People love pretty screenshots, full cinematic cut scenes. Obviously there are exceptions like retro/pixel art but they have to still look good (yes there is bad pixel art).

If you make a game which does not have the latest bling or a tried and tested brand name then it will be a huge mountain to climb to get anyone interested. People love shiny stuff! 

"All that demo showed me is improvements in real-time rendering in a cinematic sequence."

Right, and that's what we're offering you. The potential for top end visuals with a great tools pipeline and a fantastic scripting system in the Kismet/Blueprint features.

The gameplay is up to you. 

Spiney said it really. Lighting looks spectacular and everything looks super smooth, no jaggies, no sharp stencil shadows. Looks completely pre-rendered. It's achieving that weird x-factor where I can't work out how it's being rendered. Uncanny valley here we come! There are exciting times afoot with the next-gen consoles round the corner methinks. 

yes it does. The game industry just lost all the customers that actually cared for it years ago? 

His point is more that sex sells. You can't attract customers with gameplay. You can, however, attract them with great graphics. Then once you have their attention, you can show them your gameplay. 

it totally doesn't work for me that way. I never would watch graphics-/cinematic-only trailers if i wouldn't actually be doing CG for a living.

Gameplay videos actually attract my attention. If there's something E3-like, I watch gameplay videos, as well as the top 5% of the graphics only stuff (but i would never buy those) 

it totally doesn't work for me that way. I never would watch graphics-/cinematic-only trailers if i wouldn't actually be doing CG for a living.

They would for me if they weren't filled to the brink with Hollywood cliches presented as pubescent power fantasies... 

Besides what has been already said is how fulfilling and rewarding succeeding feels in this game. After finally getting past area XYZ or killing boss ABC, its not like "oh well, okay" like I feel in most other games, it's more like "HOLY FUCKING SHIT YEAH BITCH" and literally almost jumping from joy. 

As long as it works for enough other people who care more about eye candy than they do about the graphics. Come on, this is obvious. Also, I think Willem was saying that they're trying to sell their engine more than they're trying to sell their games. 

Graphics are important. People chase the next big thing in computer graphics. If you are making a game, then regardless of what you want the gameplay to be like, you want the graphics the be good. You also want the tools to be good. So this video shows that both the graphics and tools for the engine are awesome. Seems like a win-win to me. 

SleepwalkR

Right now, yes, because we're not promoting a game. :) We're an engine provider as well as a game studio so this demo is to get people interested in licensing the engine. Which everyone should! It only seems right... 

all I see in those videos is a trillion dollar artist budget. I'm not at all familiar with the unreal technology (is it still called that?), but..

Where are the new rendering techniques? Where are the "simple art" test cases that show the pros and cons, where are the performance breakdowns? Where are the awesome features your engine has that nobody else has? Voxel cone tracing maybe? But where are the details of what is actually implemented there? When does it break down? How does it handle near-field? I heard there are problems with that... in the original paper. What about your implementation?

If I would have to make a decision about spending a lot of money on engines, those would be what I want to base my decision on. All I see in most engine videos is more particles, more deferred lights, more post processing, etc., and of course, lots of artists. Those don't come with the engine, I heard =)

Maybe that's my academic view, and that's because I see the papers a long time before an actual product uses that stuff. Maybe managers just buy stuff, because they don't even know what engine vs artist means. Or it isn't even about them, it's about the public that in turn buys everything with unreal engine "because it just has to look as good as the tech video".

oh, and i didn't even watch it, just the graph ui advertisment. :-)

okay, i just did. the glossy reflections look indeed nice, but there does seem to be only one type of glossy material (puddles/wet) that actually uses it. Why's that?

What else is there? Nice volumetric lighting, but not enough time and bad video compression to actually see what the problems with it are. Or how fake it is. Then... there's lots of particles, lights, artist budget :-) How many "VPLs" are actually hand-placed? Model/Animation of the main character with dread locks looks really bad (the joints?!), but I'm not sure if that's just artist fuck-up (but how did it pass QA then, if it was easily fixable? See the problem?) What shadowing there is looks really nice, but some stuff doesn't seem to cast any.

And it's all really pretty. 

what's your code quality, how easy is it to modify stuff, plug my own stuff into it, what abstractions are there... I'd like a video about that =) 

I expected some nice honest advertisement but that natural and spontaneous infomercial made me cringe. So much bullshit bingo phrases. I swear I was able to do what I want and let the creative energy flow into the code and leverage the intuitive interface and feel unleashed with that FPS creator thing that was advertised in magazines in 1996. You can be a game designer too! But only with this new product. Not any of the competitors. It is so new and innovative! 

It's a marketing video, yes, but look at what's being shown. That's no bullshit. Shane created all of that stuff entirely on his own and he knows nothing about coding. He's an artist, through and through, and he's making games entirely on his own in UE4. 

They did release some paper last year:

http://eat3d.com/blog/metalliandy/siggraph-2012-technology-behind-unreal-engine-4-elemental-demo

Also some walkthrough of the editor and some tech info:
http://www.youtube.com/watch?v=MOvfn1p92_8

And yes, million dollar art budgets and such. I totally agree. I also tend to get more excited about NPR and innovative art direction nowadays. Nevertheless, impressive stuff.

One thing I wish would get solved once and for all in this generation is transparency. 

I hope something is coming out of that Samaritan demo ;) I loved the noir cyberpunk setting.