All language subtitles for 13. Creating the final composite

in this lesson

we will finish our composite

and review what we have Learned

in our previous lessons

we've made it up to this point

where we have our CG renders composited

we've made holdouts

we've added a rain particle system

and integrated it with deep

and we've added fog

to replace our bad atmospheric volumetric render

from Render Man

I'll show you what this looks like now

we have our background

and then we have all of our CG

deep holdouts

and over here on the right

we have our rain and our fog holdouts

and this is where we left off last time

with this node tree

I've organized this a bit since

so as we come down here

all the different holdouts are being merged together

they're labeled they're merged with an over

it's important to mention that I'm merging all

so under also merge all

and if remember

a great way to keep track of what you're doing

is to use labels

so not only am I labeling

what the element is called upstream that were merging

but I'm also adding in the little expression value

and then the knob

in this case also merge

which of course

allows us to see beyond

RGBA what's being merged which is very helpful

and this script follows uh

the theory of beast blind compositing

which is something that I haven't discussed yet

but now is a great time

so in Beast blind compositing

it's something that you may be familiar with

basically the theory is that in nuke

when you disable a node

it disables the a input

and the B input becomes a pass through

and you tend to work from back to front

so the root of this composite is actually

the background

so if I disable this merge

and the others separately

then you'll just see the background

if I enable it now

the CG from the apipes flowing through

so to make this a little bit more articulate

I'll step through it like this

so we have on this

just the character

they're gonna merge the ground

large vegetation 1

large vegetation 2

vegetation small and trees

so as you can see if you just toggle the disabled

everything else still works

if you were to connect these in the other way

let me swap the inputs here

it would still look the same

but if you disabled it

it would follow the B pipe

which means that

only the small vegetation would become visible

you can see it's highlight yellow

because it's where the data is flowing through so only

only this element would now be visible

so this is really efficient

because as you go through your composite

you can just toggle things on and off

toggle different layers and very intuitively

manage what's being rendered

so once you put together the

the Grillo CG

it goes into this template here

now some of you might be familiar with CG

multipass compositing

basically what's happening here is the CG was rendered

in a rendering engine called redshift

which is a uh

fairly new uh

GPU accelerated

uh bias render

and what this template does is it takes the standard

a 0 VS that are output from that render

and it allows you to rebuild the beauty pass

the whole purpose of this

is just having the flexibility

to be able to change colors or

or more likely

change the amount of specular reflections

refractions

so I won't go into technical detail

about how this is particular to redshift

it's fairly similar in theory

to other rendering engines like Render Man

but every rendering engine is different

so for your situation

this might not necessarily apply

something to point out is that here

check for missing passes

basically what's happening is

all of these passes are being combined

they should reconstruct the beauty but

and they're being subtracted

from the incoming beauty pass

so if all the passes don't exist

they won't all be subtracted

which means something will be left behind

in this operation

and then you'll be able to see which a O V

was not included in the template

and which needs to be

and then finally the alpha channel gets copied back in

because the AOVs

we don't use these alpha channels

when we merge these together

these just RGB

and then I strip out everything except RGBA

the object IDS

and the normal

the position and the depth because those

are things that could be useful downstream

and I actually do use them in a few cases

so what I decided to do is because we've added

that it rain and atmospherics

I thought we should make this look

a little bit more cloudy

overcast gloomy like a rainy day

so the first thing I did

was just decide to try to get there through grading

that's pretty fast

so this grade

basically just pulled some of the green out

made it more blue

made it a little bit less contrasty

because if it's an overcast day

you don't have direct sunlight

so there's a little bit less contrast

additive in yet just a simple roto shape

this really brought out the uh

the composition

to really focus your attention at the center

and make things a little bit darker around the edges

really help things look a little bit

more like they're in shadow

at the end here

I'm adding some grain

to help it look more photographic

I'll get the focus at the end because it's PG bouquet

and it's very slow to render

so something that I wasn't too happy with is

I wasn't too happy with

the way that these plants looked

so I went ahead

and I used the object IDS that we made before

to isolate them

and I graded them down

made them a little bit darker

a little bit more blue

so these are just some grades they do

the same kind of grading that's being done overall

but just specifically to these plants

and I mask them with the object IDS that we're using

so you can see here

this is the ground vegetation

I just thought

this made everything just sit in a lot better

and then I thought that

the highlights on Girlo's eyes

were just a little bit too hot

and a little bit too white

so what I did is I used the

luminance for the eye reflections

to sort of isolate this highlight here

and then I went ahead and

graded this one down

so we just look here

pretty subtle difference

but I think it helps

and in this case

what I did was

I used the position pass to make a position mat

and if you never use position mats before

it just uses the

values of the position pass to generate

a new mat

so this is really cool if you've never used this before

I recommend looking into the technique

it's a bit more pretty subtle as well

just bring down these highlights

they were very bright way over 1

now they're little bit more subdued

so that's starting to look pretty good

and the last thing I did was

grabbed PG bouquet

and I rendered out a deep precomp

took everything every volume made a

made a single precomp of it

and then just fed it back in at the bottom here

and the appager value

I decided to use was F11

for the camera

I thought that was a nice photographic value so this is

where we end up this is our final composite

we've reached the end of the course

I hope you enjoyed it

good luck with all of your deep adventures

and happy compositing