Amazon Spent Big Bucks On Crytek's Engine
polycounter lvl 12
Just seen this.
http://kotaku.com/sources-amazon-spent-big-bucks-on-cryteks-engine-1696008878
Soo... CGI movies for their network or games ? Maybe both ?
70 million is hellova money. It sounds like they made a complete buyout of the engine with rights and all. Like RSI for Star Citizen.
Well whatever the deal is I only pray that Amazon puts up it's own version of the engine out there for indies with decent tutorials and community love.
Juts leaving this here.
http://kotaku.com/sources-amazon-spent-big-bucks-on-cryteks-engine-1696008878
Soo... CGI movies for their network or games ? Maybe both ?
70 million is hellova money. It sounds like they made a complete buyout of the engine with rights and all. Like RSI for Star Citizen.
Well whatever the deal is I only pray that Amazon puts up it's own version of the engine out there for indies with decent tutorials and community love.
Juts leaving this here.
Replies
Could also just be a push into the hard core market? Or VR.
can't say too much here, but research is being done. However it's not just as straight forward as in "let's replace our render farm with an engine". Requirements and resulting workflows are still different for pre-rendered and games. Current research is mostly aimed at using real-time for TV series style quality, as far as I'm aware.
But I often wonder why the same studios that make the hours upon hours of in game cutscenes don't make in engine movies?
And even when they do they even outsource them to other companies for something that often doesn't look as good or has as good of a story. We know game companies have all the tech they need to make a damn good movie like they do in-game.
To me the difference in story and quality might be like the animated marvel and DC movies (not tv series) compared to the live action ones.
It's just confusing that this has been held off for so long. We know it's be tried and pitched in the case of Enslaved. But the only reason that got shot down was because it didn't need a super high budget XD.
its easier, faster, more itterative and cheaper.
in game cutscenes only make sense because you need to produce the assets for the game anyways, so you might aswell use them for some cutscenes.
Then why do people do CG or animation at all? Film isn't "that" much easier, considering the scale of most games.
I doubt it's cheaper too, if you're talking about a full length movie.
Short film? Sure, but a "quality" full length movie could cost as much or multiple times a complete "quality" game.
I doubt a game of just cutscenes would cost more, and certainly less money and time than a full CG movie considering render times and such.
Have you tried to get CineBox? I think that Virtuoso is big enough to have opportunity to lay its hand on it, so if you know something you WOULD share it with us, right?
nope. We're dealing mostly with Unity and Unreal. The idea is to make use of existing knowledge and tools. Also, hiring people who have those skills is much easier.
Man even I don't remember it exactly
http://games.amazon.com/
The most common reasons for not using engine movies are, as far as I know, the following:
Hiding load times.
With an engine movie, you'd first have to pause to load all the assets for the cutscene - models, textures, animation, sound etc. - then after the cutscene, you'd have to load the next level's data.
However, if you pre-render the cutscene, you can stream it into memory and still have enough bandwidth to load the next level while the movie is playing. So the player doesn't have to look at any load screens; whereas with an engine movie, it'd be two long load screens.
Better quality from the ingame assets.
Pre-rendering the cutscene allows for the developer to dial up the engine's settings for antialiasing, shadows, lighting etc. and even use higher quality character models and textures compared to what's possible in real-time.
Better quality cutscene overall, by using offline CGI.
As advanced as current game engines are, their abilities are still nowhere near what's possible with offline rendering. Vastly increased scene complexity, more detailed assets, higher quality shading and lighting, much more complex effects for things like fire, smoke, destruction and so on, and of course much better image quality.
Outsourcing the cutscene production - even to an internal CG studio of the publisher, like Square's Visual Works - also frees up resources at the development studio, so a lot of people can focus on making the game itself better.
This is a matter of personal opinion, so it's hard to argue here. You believe this, others may believe differently.
With that said, know that the game development teams are always heavily involved in the production of outsourced cutscenes as well; usually they write the script, but sometimes they also do animatics. They supply the concept art for the assets and sometimes even models and textures for reference, and they're always giving a lot of feedback during production as well. Basically, the outsource studio pretty much acts like an extension of the developer.
So if you're not content with the results, you should probably blame them and not the contractor.
It's certainly possible to do a lot with game engines nowadays, especially compared to what we've seen 10 or 20 years ago. But offline CGI is always going to have an edge, because they're not limited by the amount of memory and the processing power of the game's chosen platform. Even a small CGI studio has hundreds of powerful CPUs in their render farm, and they can spend hours and hours on computing a single frame - compared to a few gigs of memory and a single CPU and what can be done in 33 or 16 milliseconds. This is something I can't see as arguable.
If you truly believe that an offline rendered movie looks like an animated Marvel film and an ingame cutscene looks like a feature film, then this debate is pointless...
Apologies for bringing up a personal issue... One of our latest jobs was to produce 20 minutes of CG movies for COD:AW. I believe they had the budget to do the whole thing as an engine movie, it probably wasn't a limitation. So I believe they must have had their reasons to chose us.
While I am very much interested in the appeal of just doing actual real time movies.
I mean just made in engine not displayed realtime. I'm asking why we don't see those pre-rendered but made in engine cutscenes in anywhere but the games themselves and unofficial machinima? I'm aware there's limits with loading times and such for true real time.
I just think there's a desire for people to get "episodes" or movies using a visual presentation fans are used to.
These two I can understand, it depends on need. But like i said before i'm definitely talking about offline.
But is it impossible to render a movie in say unreal engine 4 with each frame taking 1 second instead of 1/30th of a second? I know they said the openworld kite demo was rendered in real time, but what if a studio would take it a step further and ignore real time and just use the engine to render?
This can definitely be the case for those movies. But I look at stuff like that heavenly sword movie that was released and wonder how that kind of thing happens and if Ninja Theory was theoretically able to do it themselves would it be done better?
no thats not what I meant, I should have been more clear. I mean the animated movies are usually more cannon and more closely represent the stories and characters shown in the comics, I personally enjoy these more. The films on the other hand usually take a lot of liberties and the storytelling or sometimes characters is very different.
I'm saying the engine rendered movies made by the studio's might feel more true to the source than movies often do.
Thats interesting, but yeah I don't expect it to be very common, I just wonder why it hasn't been done at all beyond a few short films.
I used Unreal to do exactly that (it was for machinma used within a live action TV pilot) about five years ago. I would render everything at maximum settings and output individual uncompressed frames. Then I could give it to a compositor and editor to use within the show. The reason to use Unreal was to make it look like a real game and authentic.
As i remember it was used only for previs.
Some things that game engines barely or don't even consider:
Motion Blur
Subdivision surfaces
Displacement
Fur and Hair
Post-processing
Sub-surface scattering
Volumetrics
Efficient memory usage
Have you happened to check out the epic kite demo from GDC, they are considering many of these things. Funnily enough for UE4 I think a lot of the realistic rendering stuff is being pushed by the visualization industry rather the game industry.
Also the last thing on your list "Efficient memory usage" what do you mean by that... and how the fuck do offline renderers do well with this?
All of this is addressed by Cinebox. Plus Alembic.
On CGTALK there is a guy that claims that his company used Cinebox to previs and FINAL quality render of more than 120 shots for "Maze Runner".
Crytek used Cinebox to render commercials for TV straight from Cinebox for their game, Ryse Son of Rome.
The public cryengine forum is a horrible experience, there is no evidence of anything worthwhile being done with the software by indie devs in all the years it has been available. There is one alpha pre-release survival game and an admittedly decent looking fantasy rpg that is in development. The rest is mostly just some teenage kids stealing other peoples work or placing some default assets in an empty level with no gameplay mechanics and calling themselves a game studio.
Crytek seem to not be interested in engaging with the community, with one crytek dev recently saying he doesn't even feel like reading the cryengine forums. There are long running requests for features, bug fixes and even simple requests for ease of use and clear documentation that have gone ignored for years.
With no income from royalties (not that there are any profitable games being made) and only making $10/month from at best a few thousand subscribers, I am not surprised that they don't want to waste time or resources on the cryengine eaas. But why keep it around at all? Just kill it off and get rid of all the negativity and bad publicity and stick to doing multi million dollar licensing deals.
http://www.fxguide.com/featured/cryteks-cinebox-an-update/
Most recent video of a Cinebox project I could find taken from a Cinebox thread on the Crytek forum:
https://vimeo.com/channels/782595
Looks like standard game quality to me (nothing wrong with that, of course).
Kite Demo is fantastic and artistically amazing. It definitely is a huge step forward for real-time rendering.
I agree that the game industry, or at least the game industry outside of AAA, has moved past realism and is more stylistically diverse.
Yeah, I have seen it. But those are previs only. Kinda hard to guess what can be done with it just based on this.
Here is article about RYSE pipeline => http://www.fxguide.com/featured/the-tech-of-cryteks-ryse-son-of-rome/
There was a page on Crytek about Cinebox, which I can't find right now, where they mentioned supported features. Also from what I heard they use a little different algorithms there for couple things (GI).
Yeah pretty much this
Also to add many older user base people are waiting 3.7 to make a final decision about the engine of choice. If it flops quite a bit of people will leave for UE4 or Unity.
Aand I guess we pretty much summed it up.
By the way I want to say this, I am completely fine with teenage people using ready assets and placing on maps. This is how things start for people out there. It is neither wrong nor must be looked down on.
EDIT:
By the way many people creating actually fun and good stuff are outside the community for obvious reasons.
Lately I am hooked to this guys work. He renders his MMD's in CryEngine, and I love good MMD's and CryEngine
https://www.youtube.com/channel/UCVKw4LIPdmVLkdAdP0u8FSQ
I have actually been a fan of cryengine for a decade and was always amazed at the games that were made with it in the past. It is just really disappointing to see it not getting the treatment it probably deserves.
Naah, don't worry. I feel ya mate. Don't be sorry
Also I feel exactly the same. I also want to see skillful people working with it and a very good community after all the games/ demos/ videos. But alas only good work of CryEngine comes from outside the community and forums.
It's kinda hard to find a good reason to waste your time on engine that 2 years ago didn't had even FBX support. You needed to use their 3dsmax/Maya exporters. If had never version than they supported or other program, well... f*ck
I think right now it changed a little and FBX is there, but in the age of super user friendly Unity import and almost as friendly UE4 import, who cares?
I don't.