How volumetric video brings a new dimension to filmmaking Diego Prilusky

I love making movies.

Motion pictures have been in existence
for more than a hundred years.

Filmmaking hasn’t changed
for the dimensional mindset.

Placing the camera in a scene
and pressing “record” hasn’t changed.

Filmmaking is still a frontal experience,

and creating the film has the possibility

to follow the same direction
of the content creation.

We still stand in front of a flat image,

watching the fiction.

There’s nothing wrong with it.

I love watching movies
and going to the theaters.

The experiences can be
such emotional experiences.

The art and craft of emotional
experiences within a frame

can be so strong
to drive a stronger emotion.

The question we’re asking is,

How the experience of motion pictures
can exist beyond the flat screen.

How can we start creating content

for the next generation
of content experiences?

Traditionally, when we imagine a scene,

we look at the frame and the composition.

We have to think about how we create
depth and parallax

using foreground, background elements
as the camera moves.

With the technology today
and devices of VR glasses,

AR glasses, smart devices,

allowing three-dimensional
and full navigation in space,

we have the possibility
to enable audiences

to experience content
from multiple perspectives.

What we have to think about
is how we take this technology,

all the capabilities,

and enable the experience
to move farther away inside the scene.

Now we’re not talking about video games
or computer-generated actors,

which look tremendously realistic.

We’re talking about real actors
and real performance,

performing onstage.

We have to start thinking
how we capture the actors

and how we capture the real scene

in order to immerse inside.

Now, we’re familiar
with the 360-degree video,

where you place a camera inside the scene

and you can create this beautiful
panoramic image all around you,

but from the same aspect,

filmmaking is still frontal.

In order to emerge fully inside the scene,

we will need to capture the light
from all the possible directions.

We will have to surround the scene
with an enormous amount of sensors,

with all possible capabilities
to capture the light

and enable us to emerge
inside afterwards again.

Now, in this setup,

there’s no more foreground or background

or a camera placed in space

but hundreds of sensors
capturing the light

and capturing the motion
from all the possible directions.

With the new technological advancements,

we can start looking at 3D photography,

capturing the light
from multiple perspectives,

enabling us to reconstruct the object.

This is like photography in 3D space.

Now, with these
technological advancements,

we can record video
not just as a flat image

but as a volume.

This is what we call “volumetric video,”

and it has the capability
to record every action of the scene

as a full three-dimensional volume.

Now, what is a voxel?

A voxel is like a three-dimensional pixel,

but instead of being a flat image square,
staying light and colored,

it’s like a three-dimensional
cube in space,

with x, y, and z positions.

This enables us to create
a full capture of the scene

from any perspective.

Now this renders
a fully light-immersive scene

from multiple perspectives.

This capability requires an insane amount
of information to be processed.

We will have to capture the light
from an enormous amount of cameras

to create this information.

Now, in order to do such a thing,

we would need a setup that would host
a numerous amount of cameras

installed in a stage

and a stage big enough in order to fit
a full cinematic experience.

Now that sounds like a crazy idea,
but that’s exactly what we did.

For the last three years,

we have been building
a huge volumetric camera chamber.

It’s 10,000 square feet of a stage,

enabling to capture the action
from any location.

We have deployed hundreds of cameras,

sending a tremendous amount of information

to a huge data center
powered by Intel supercomputers.

The ability to have this 10,000 feet

enables us to fit any kind of action,

any kind of performance.

It is the size of an average
Broadway stage.

We call it Intel Studios,

and it’s the largest volumetric
stage in the world,

with the objective
of enabling and exploring

the next generation
of this immersive media filmmaking.

Now, to test these ideas,

we were thinking about what we can do
as the first scene to try it out.

So we chose the Western scene.

We brought horses, set designers, dirt,

everything needed to create
the full scene of a Western.

But this time, there was no camera inside.

There was nothing really moving
besides all the cameras

installed outside.

The challenge
of the actors was tremendous.

They have to perform a flawless action
visible from all the directions.

There’s no possibility to hide a punch
or not show the action.

Everything is captured
and everything is seen.

The output of the capture –

this is our future capture –

opened our eyes for
the immense capabilities.

It was like a full 3D scan
of the entire scene.

We were able to move around
and travel in the space.

The thing about this,

it’s not anymore about perceiving
the light emitted from a screen

but now traveling inside the light,

traveling inside the scene.

This obviously opens possibilities
for an enormous amount

of storytelling
and methodologies of creation.

This is the possibilities
of your personal narrative,

the possibility of creating
your own story inside,

or maybe following other stories.

Let’s take a look
at one of the last renders and see.

(Music)

What you’re seeing here
is full volumetric video,

and there’s no physical
camera in the scene.

(Music)

We have the full control

(Music, sounds of combat)

of space and time.

(Music, sounds of combat)

Now, again, no physical camera was here.

Everything was captured surrounding.

Now, this is very nice,

but what if we wanted to see the scene,
maybe, from the eyes of the horse?

Well, we can do that as well.

(Horse galloping)

So what you’re seeing right now
is the same action,

but this time, we’re watching
exactly from the eyes of the horse.

The possibilities are, well, unlimited.

(Applause)

Thank you.

(Applause)

So this is all great
for creators and storytellers.

It really opens a huge canvas

for a different type
of storytelling and moviemaking.

But what about the audience?

How can the audience
experience this differently?

In order to [create] our explorations,

we partnered with Paramount Pictures

in order to explore immersive media
in a Hollywood movie production.

Together with the director Randal Kleiser,

we reimagined the iconic movie of 1978,

“Grease.”

Some of you know it, some of you don’t.

A 40-year-old movie, amazing experience.

And our goal was really to look at how
we can take this iconic action and dance

and bring it deeper into the experience,

bring it deeper into the audience.

Imagine that you can
not just watch the movie

but get inside it
and dance with the actors

and dance with the performance.

Now we’re breaking, really,
the traditional 2D mindset of thinking,

and bringing a much richer
possibility of moviemaking

and content creation.

But why watch it on the screen?

Let’s try to bring these actors
here on the stage.

So they’re not going to really come –

I’m going to use an iPad.

(Laughter)

Sorry.

I’m going to use an iPad
in order to bring in augmented reality.

Now, obviously, these devices
have their own limitations

in terms of the data-computing process,

so we have to reduce
the amount of resolution.

So what I’m doing now,
I’m placing here a marker,

so I’ll be able to position exactly
where I want everyone to appear.

OK.

I think we have them here.

(Applause)

John Travolta, or –

(Laughter)

a version of him.

Let’s take a look.

(Video) Female: Hey.

Male: And that is how it’s done.

Female: Your turn.

Male: Hey, guys! Check this out.

(Song: “You’re the one that I want”)

Danny: Sandy!

Sandy: Tell me about it, stud.

(Singing) I got chills.
They’re multiplying

And I’m losing control

‘Cause the power you’re supplying

It’s electrifying!

(Video ends)

(Applause and cheers)

Diego Prilusky: Thank you.

(Applause and cheers)

So as you can see,

we can watch and experience content
in the traditional way

or in an immersive way.

Really, the possibilities are open.

We’re not trying to change
or replace movies.

We’re enhancing them.

The technologies enable new possibilities
to start thinking beyond the flat screen.

We’re in immersive and really
exciting times in filmmaking.

We’re at the threshold of a new era.

We’re opening the gates
for new possibilities

of immersive storytelling,

and exploration and defining
what immersive media filmmaking means.

We’re really just at the beginning,

and we invite you all to join us.

Thank you.

(Applause)

我喜欢拍电影。

电影已经存在
了一百多年。

电影制作
的维度思维方式并没有改变。

将相机放置在场景中
并按下“录制”并没有改变。

拍片还是一种正面体验

,拍片有

可能跟
内容创作的方向一致。

我们仍然站在一个平面图像前,

看着小说。

它没有任何问题。

我喜欢看电影
和去剧院。

体验可以是
这样的情感体验。 框架

内情感体验的艺术和工艺

可以如此强大,
以驱动更强烈的情感。

我们要问的问题是,

电影的体验如何
能够存在于平面屏幕之外。

我们如何开始


下一代内容体验创建内容?

传统上,当我们想象一个场景时,

我们会看画面和构图。

我们必须考虑如何

在相机移动时使用前景、背景元素创建深度和视差。

借助当今的技术
以及VR眼镜、

AR眼镜、智能设备等设备,

可以
在空间中进行三维全导航,

我们有
可能让观众

从多个角度体验内容。

我们必须考虑的
是我们如何利用这项技术、

所有功能,

并使体验
在场景中更远。

现在我们不是在谈论视频游戏
或计算机生成的演员,

它们看起来非常逼真。

我们谈论的是真正的演员
和真正的表演,

在舞台上表演。

我们必须开始思考
如何捕捉演员

以及如何捕捉真实

场景才能沉浸其中。

现在,我们已经
熟悉了 360 度视频

,你在场景中放置一个摄像头

,你可以在你周围创造出美丽的
全景图像,

但从同样的角度来看,

电影制作仍然是正面的。

为了完全出现在场景中,

我们需要
从所有可能的方向捕捉光线。

我们将不得不
用大量的传感器包围场景,

并拥有所有可能的能力
来捕捉光线

并让我们
之后再次出现在里面。

现在,在这个设置中,

没有更多的前景或背景

或放置在太空中的相机,

而是数百个传感器
捕捉光线


从所有可能的方向捕捉运动。

随着新技术的进步,

我们可以开始研究 3D 摄影,

从多个角度捕捉光线,

使我们能够重建物体。

这就像 3D 空间中的摄影。

现在,随着这些
技术进步,

我们不仅可以将视频录制
为平面图像,

还可以录制为卷。

这就是我们所说的“立体视频”

,它能够
将场景的每一个动作记录

为一个完整的 3D 体积。

现在,什么是体素?

体素就像一个 3D 像素,

但它不是一个平面图像正方形,
保持浅色和彩色,

它就像一个空间中的 3D
立方体,

具有 x、y 和 z 位置。

这使我们能够从任何角度
创建完整的场景捕捉

现在,这从多个角度呈现
了一个完全沉浸式的场景

这种能力需要处理大量
的信息。

我们将不得不
从大量相机中捕捉光线

来创建这些信息。

现在,为了做这样的事情,

我们需要一个设置,

在一个舞台上安装大量的摄像机,

并且一个足够大的舞台,以
适应完整的电影体验。

现在这听起来像是一个疯狂的想法,
但这正是我们所做的。

在过去的三年里,

我们一直在建造
一个巨大的体积相机室。

这是一个 10,000 平方英尺的舞台,

可以
从任何位置捕捉动作。

我们部署了数百个摄像头,

将大量信息发送


由英特尔超级计算机驱动的大型数据中心。

拥有这 10,000 英尺的能力

使我们能够适应任何类型的动作、

任何类型的表演。

这是百老汇平均舞台的大小

我们称之为英特尔工作室

,它是世界上最大的立体
舞台,

旨在支持和探索

下一代沉浸式媒体电影制作。

现在,为了测试这些想法,

我们正在考虑我们可以做些什么
作为第一个尝试它的场景。

所以我们选择了西方场景。

我们带来了马匹、布景设计师、泥土,

以及
创造西部片完整场景所需的一切。

但这一次,里面没有摄像头。 除了安装在外面的所有摄像机之外

,没有什么真正在动

。 演员

的挑战
是巨大的。

他们必须执行
从各个方向可见的完美动作。

没有可能隐藏一拳
或不显示动作。

一切都被捕获
,一切都被看到。

捕获的输出——

这是我们未来的捕获——

让我们看到
了巨大的能力。

这就像整个场景的完整 3D
扫描。

我们能够
在太空中四处走动和旅行。

关于这一点

,不再是感知
从屏幕发出

的光,而是现在在光中

传播,在场景中传播。

这显然
为大量

的讲故事
和创作方法开辟了可能性。


是你个人叙述

的可能性,
在里面创造你自己的故事的可能性,

或者可能跟随其他故事。

让我们看
一下最后的渲染之一,看看。

(音乐)

你在这里看到的
是全体积视频,场景

中没有物理
相机。

(音乐)

我们可以完全控制空间和时间

(音乐、战斗的声音

)。

(音乐,战斗的声音)

现在,再一次,这里没有物理相机。

周围的一切都被捕捉到了。

现在,这非常好,

但是如果我们想
从马的眼睛中看到这个场景呢?

好吧,我们也可以这样做。

(马奔腾)

所以你现在看到的
是同样的动作,

但是这一次,我们
完全是从马的角度来看的。

可能性是无限的。

(掌声)

谢谢。

(掌声)

所以这
对创作者和讲故事的人来说都很棒。

它确实

为不同类型
的讲故事和电影制作打开了一块巨大的画布。

但是观众呢?

观众如何能有
不同的体验?

为了[创造]我们的探索,

我们与派拉蒙影业合作,

以探索
好莱坞电影制作中的沉浸式媒体。

我们与导演 Randal Kleiser

一起重新构想了 1978 年的标志性电影

《油脂》。

你们有些人知道,有些人不知道。

一部40年的电影,令人惊叹的体验。

我们的目标是真正研究
如何采取这种标志性的动作和舞蹈

,并将其更深入地融入体验,

更深入地融入观众。

想象一下,您
不仅可以观看电影,

还可以进入其中
,与

演员共舞,与表演共舞。

现在我们正在打破
传统的 2D 思维模式,

为电影制作

和内容创作带来更丰富的可能性。

但是为什么要在屏幕上看呢?

让我们试着把这些
演员带到舞台上。

所以他们不会真的来——

我将使用 iPad。

(笑声)

对不起。

我将使用
iPad 来引入增强现实。

现在,显然,这些设备

在数据计算过程方面有其自身的局限性,

因此我们必须
降低分辨率。

所以我现在要做的是,
我在这里放置一个标记,

这样我就可以准确定位
我希望每个人出现的位置。

行。

我想我们这里有他们。

(掌声)

约翰特拉沃尔塔,或者——

(笑声)

他的一个版本。

让我们来看看。

(视频)女:嘿。

男:这就是它的完成方式。

女:轮到你了。

男:嘿,伙计们! 看一下这个。

(歌曲:“你就是我想要的那个”)

丹尼:桑迪!

桑迪:告诉我,伙计。

(唱)我感冒了。
他们在成倍增加

而我正在失去

控制因为你提供的力量

令人兴奋!

(视频结束)

(掌声和欢呼)

Diego Prilusky:谢谢。

(掌声和欢呼)

所以大家可以看到,

我们可以
用传统的方式

或者沉浸式的方式来观看和体验内容。

真的,可能性是开放的。

我们不是试图改变
或替换电影。

我们正在增强它们。

这些技术
为开始思考平面屏幕之外的新可能性提供了可能。

我们正处于电影制作的沉浸式和非常
激动人心的时代。

我们正处于一个新时代的门槛。

我们正在

为沉浸式叙事的新可能性打开大门

,探索和
定义沉浸式媒体电影制作的意义。

我们真的才刚刚开始

,我们邀请大家加入我们。

谢谢你。

(掌声)