All language subtitles for 03. Discussing the history of deep compositing

in this lesson

we'll discuss the origins and history

of deep compositing

so the story of deep compositing begins in 2000

at Pixar Animation Studios

so Pixar was searching for a solution to rendering

self shadows and cash shadows for hair for

and volumetric renders

now this is a Cigraph paper that Pixar published

describing their new Deep Shadow Map technology

so normally

what you would have to do

if you rendered something like a fur or hair object

this is a example as a ball with 50,000 hairs

is that you'd have to use a shadow map

now shadow map has a finite number of pixels

and those pixels have a memory footprint

so if you look at something that you might

find more desirable in production

something like a 5 12 by 5 12 pixel shadow map

you see the result is coarse and grainy

and not really satisfactory for production

so you increase your samples

to something like a 4K by 4K shadow map

and you get a satisfactory image

but the memory hit

and the computational hit is unsatisfactory

it's very large

and if you're doing this all the time

it's going to impact your production

so

this fourth image

you see a 5 tail by 5 tail deep shadow map

now the Deep Shadow map

unlike a normal shadow map of course

has deep slices so it has multiple samples

and

this allows you to get the same self shadowing

and cash shadowing quality

but with a smaller map

so

what Pixar did

he has developed the technology that ended up

becoming deep compositing

to solve a somewhat related issue

because when you think about

the same reasons that we use deep compositing

it's really

the same reasons that Pixar uses deep shadow maps

they use it for dealing with motion blur

dealing with Volumetrics

and dealing with complex occlusions

so at the end of this paper here

you see they give examples of

a shadow map for a heavily motion blurred object

self shadowing and volumetric shadows

so these are really the same examples of

of the types of renders that we would use to

composite with deep anyway

so can see that it's really

the same problem they're trying to solve

and this is something that

that comes into play

and people in the compositing world recognize this

so

Pixar after they publish this paper

they go ahead and they implement

deep shadow maps into their render man software

so deep shadow maps become commercially available

and it uses the format

Detex it's a deep shadow a deep deep texture

and this is the format that Render Man still uses today

so

we jumped to what

a digital in 2,008

and a man named Calm Doncaster

he recognizes what Pixar has done with deep chetamaps

and he thinks that that kind of

depth slicing from a 3D scene

would be a great way to control holdouts in compositing

so he goes and he builds

an implementation of deep compositing in shake

so the time whether was still shake

this would be for nuke

and they use it on the scene

in the day of your stood still

where the truck is dissolved

and it's a volumetric render

and it's a really good example of where you'd use deep

so this was a successful test

so weather decides to keep using deep on

The Lovely Bones in 2,009

and then in a larger capacity on Avatar

now its use on Avatar gets a lot of attention and

people start hearing about you compositing

and this leads to Johanna Sam who

has been aware of Collins work

and he decides to write implementation of

of deep in nuke as nuke 5

and so what Johannes does is

he makes his own nuke plugins

so he has deep nodes

and he publishes videos of him using them

and demonstrating how deep compositing works online

so now everyone

in the industry gets to see how deep compositing works

before they had just heard about it

unless they were witnessing at firsthand at UETA

so now deep compositing becomes

sort of a big deal because now people see the benefits

so what Johannes and Collin

and two other people

Chris Cooper and Daniel Heckenberg

what they do is they decide to write a cigarette paper

that describes deep compositing and

surprisingly

this paper gets rejected by Cigraph

they decide that

it's not gonna be part of the conference

because they say that it's too similar to

Pixar's technology

and it's not

it's not an important development

now the irony here of course

is that deep compositing

ends up becoming a very popular technique

but this paper

is the first time that

people really get to see deep compositing described

from a technical standpoint

so what now

continues to use deep compositing

but they now switch over to nuke in their pipeline

and they use it to a large

large degree on rise of the Final Apes in 2,011

and they get a lot of attention by making a

deep compositing demo video

which some of you may have seen about their work

on rise of the planet of the apes using deep

now at this time

Weta is still using their custom deep pipeline

they're using a format for deep called Open EXR

Deep Z or ODZ

probably ODZ is how they pronounce it

and they use this because they didn't wanna be tied to

the render man native format

they wanted a format that they could use

for other rendering engines as well

and a way to store it um

in an open source format

so at this point

Pixar decides to add deep image output as a feature

natively in Render Man

before it was still a hack

it was still just taking d text deep shadow maps

and using them from the camera's perspective for images

now they add official sport for it

in Render Man Pro Server 16

of course Render Man still uses the DTX format

so now at this point

after Pixar adds official support

the foundry

in Nuke 6.3 decides to add native deep support to nuke

so they make their own deep Nodes

and the first iteration of deep Nodes

only really supported Render Man's Dtex files

they didn't support what they do now

and a lot of studios that used

nuke decided to also started

start to use deep compositing

so now deep compositing is really

exploding in the industry

and critically

the open E XR file format

gets to version 2

and a major part of their update was adding support

for deep image data

so now it's not just Detex files

or other proprietary deep formats

now there's one single format that pipelines can use

to support their deep image data

and in Nuke 8

the foundry had support for Open ex R 2.0

they also had support for

rendering deep image data with the scanline render

and outputting deep data with deep right

prior to this

you could only import data into nuke

you could not output deep data from nuke

so it was really just a one way in

but not out

now go back to the irony of the cigarette paper

being rejected in 2,014

the developers of Deep Compositing

actually won a Technical Achievement award

or a Psychtech Oscar

and this was a huge recognition for them

and it really underscores the contribution

that the industry feels Deep Compositing has had

now today deep compositing is

fairly common in larger studios and high end pipelines

and is supported by most major rendering engines

things like Arnold Houdini's Mantra

Maxwell render

the new red shift pixels

render man of course

V ray through the light

uh there's a few rendering engines

that is not supported by

uh mainly uh

the largest one being Mental Ray

uh and from a compositing standpoint

there are really only two applications

that support deep compositing

and that's of course

nuke and then fusion

um no other 2D applications support deep compositing

so it's still fairly

dedicated to high end pipelines but

maybe in the future

things will spread out and you'll have more options

but we're gonna focus on a nuke in this course

and hopefully this history lesson

is gonna help put things into context

about why certain things are the way they are

why formats are the way they are

why work why workflows are the way they are

and hopefully in your experiences in

in different studios

you might understand sort of

how they got to the pipeline

they got to