All language subtitles for 10. Lesson 8 - Render

hi everyone

in this lesson we're going to talk about lighting

lighting is an important element in the creation of

graphic works

while

light can be used to highlight the shape of an object

it can also be used to create atmosphere in the scene

and create accents in the scene

among other things

it's an important mechanism for creating narrative

and emotional content

so since it's such an important topic

let's get right to the point

there's a classic set of light sources in blender

that you can find in any 3D program

and this set is enough to simulate

any physical light source

let's take a brief look at each of them

the 1st light source is point light

the light from its spreads in all directions

although it's extremely easy to work with

because you don't have to watch the direction

of the light source

but that is also a disadvantage

and it's quite difficult to limit

the areas of the three dimensional scene

from its impact

so this type of light source is often used as a lamp

lantern or street lighting

it's fairly easy to set up a light source

and the set of parameters is almost

identical for all light sources

we can set the color of the lighting

with the color parameter

and thus color the areas of the scene

in the shade we want

the next parameter is the light intensity

which is given in Watts

the Watts describe the amount of power

in the luminous flux

this isn't the most convenient unit

if you're more used to working in lumens or Lux

which are frequently used by lighting designers

you'll have to convert one value to another

or you can keep a table of

already converted values somewhere at hand

but in our case we won't do that

we'll just treat this parameter

as a kind of light intensity

the radius parameter shows the size of the light source

which affects the degree of shadow blurring

if the radius is small

it makes the shadow very clear and sharp

but since such shadows look quite unnatural

I always try to keep the radius much higher than zero

the max bounces parameter

sets the number of light ray bounces in the scene

this is a rather specific setting

which can be useful

only in the case of debugging rather complicated

complex and three dimensional scenes

the cast shadow

option allows you to disable

shadows for the selected light source

it's not very photorealistic

but those kinds of dirty tricks can be useful in

certain situations

the multiple importance parameter

allows you to distribute all the rays from

the light source more rationally

and if necessary

increase the number of rays for glossy materials

and materials with reflection

to reduce the amount of noise in the rendering

besides these basic functions

we can activate the nodes system for each light source

thanks to it

we can expand the functionality to a large extent

we work with nodes in the shader editor window

for example we can use nodes black body

it describes a physical model of a black body

and allows you to set the color of the source

using a color temperature value

this is a more natural method of working

was the light hue

as an example

a value of 3500 Kelvin

would roughly correspond to the color of a setting sun

and 5400 kelvins would correspond to a photo flash

you can find these values on the internet

or check with the lighting designers

now we'll look at the next source of light

namely the sun

this kind of luminaire imitates the source of light

which is infinitely distant from the object

like the sun or the moon

if we talk about the dark time of day

also it has two important properties

namely the lack of fading of the light intensity

and the parallelism of the shadows

note that this

light source illuminates the entire scene

no matter where it's located

and for its adjustment

the only important thing is the direction of light

which we can set with a rather handy manipulator

the next light source is spotlight

this light source is like a point light

which has the possibility to set the

direction of the light

besides the basic parameters

we have the possibility to control this cone of light

and to make the values of the cone harder or softer

such a light source can be used as

any kind of lanterns or spotlights

another light source

that is available to us is area light

this is the most convenient and universal light source

and depending on the settings

it can be used in the role of those

light sources that we've

already discussed with you

this source has a surface area of light emission and

is used like a soft box

in the shape setting

we can change the shape of the surface

and make it either radial or arbitrarily rectangular

we can change the direction of the rays

with the spread parameter

from parallel to all directions

this allows us to create a very soft and diffused light

moreover we have the possibility

to create lighting with the environment

its settings are in the tab in the world properties

or by switching the shader editor window

from object to world

as you can see

node's background is already connected there

in which we can set the color

and intensity of the lighting

this method of lighting is applicable

when we need to create a filling light

such as from the sky

this method is most often used as an additional method

after all

this method of lighting looks very unexpressive

if we create

nodes texture sky and plug it into the color parameter

we get a more complex type of lighting

nodes texture sky simulates

celestial solar illumination

and doesn't require additional light sources for this

we can use the sun size parameter

to create lighting both in clear and sunny weather

and in cloudy weather

when the light passes through the clouds

and becomes very diffuse

the sun intensity parameter

can adjust the overall effect of the sun

on the lighting model

with sun elevation we can adjust

the height of the sun in the sky

according to the height of the sun

the sky will also change its hue

ranging from clear blue to sunset orange

also

we use sun rotation to rotate our celestial sphere

this allows us to adjust the angle of the shadow

we set the altitude parameter to the altitude above sea

level

and with the right settings

we can achieve lighting like in a mountainous area

meanwhile with the air

dust and ozone options

we can create more specific scene conditions

besides we can use HDR images to illuminate our scene

we'll use the nodes environment texture

for this purpose

and load a suitable image into it

this way we can create a very accurate environment

lighting model

often only one HDR map is

enough to illuminate 1 3 dimensional scene

we can rotate the HDR image

if necessary to find the best angle of illumination

now let's go to our scene and

do the lighting for our character

I've already picked up

one of the animation frames

that looks quite expressive to my taste

and based on it I made a scene with lighting

so let's look at the final result and analyze it

it all started with a camera setup

I found the right angle

and set the focal length on the camera to 80mm

this is a typical focal length for portraits

which both leaves a sense of perspective

and doesn't distort

the proportions of the subjects too much

I also made a background on which our character

will be placed

I made it rounded

in

order to get rid of the flat lines in the background

for the background

I used as simple of a material as possible

matte without reflections

its light is quite dark with a slight blue tint

typically

it's completely unnoticable in the final rendering

this is because the horse itself is block

thus I have to raise the brightness

of the light sources rather high

and in order to keep our picture

within the correct exposure

I had to make a rather dark background

for convenience

I divided our light sources into two collections

the light foreground collection

contains the light sources that work

directly with a horse

the 1st and main source of light is the fill light

this is a large area light

that softly illuminates the horse and the entire area

around it

I made the light source a neutral warm shade

which almost doesn't brighten up the scene itself

the intensity of the light source isn't very high

so that there is a range in brightness for

other light sources

the next light source is key light

this is used to create the main character

of the fluck

this is also area light

but a little smaller

also it creates the basic falling shadow of the objects

at the same time

I changed the value of spread for this light source

I made it a little smaller

to focus only on the character

and to minimize the effect on the background

the next group of sources works to create

outline lighting

these sources are called rim lights

as always these are area lights

one at the top and one at the back pretty wide

its main task is to create a beautiful light

on the main its color is neutrally cool

so that it slightly balances out

the fill light source

and it's also slightly reduced in spread

to increase the effect

of the slight source on the main character

there are also two other sources of backlighting

on the sides

on the left hand right

which also highlight the horse's outline

on the left is a blue light source

which solves the problem of

excessive blockness of the horse

due to the nature of the fluck

on the right side it's a fairly small area light

a cool shade of blue

its placement and size

made it possible to create

quite an interesting fleck on the glasses

both of these light sources are appointed upwards

so as not to give excessive light on the floor

that's how we work out

to the two backlights that work for the silhouette

was the 3rd one we light the hair

fill light is for general lighting

and key light makes the volume more expressive

now

let's see how the light sources work in the background

they're organized in a much simpler way

there are only two light sources

one serves as a large fill light source

and the other is a local key light

they create a Halo of light around the horse

this is the source of light

that I did with a normal point light

when we include all sources together

we get quite a volumetric picture

it's also worth mentioning

that I do all the render tests

with a reduced number of stripes on the horse

therefore in the final render

the brightness value

on different elements will be slightly different

now let's talk about the render set up itself

for this project we're going to use the cycles render

which is the main one for blender

let's take a look

at the settings that might be useful in this project

1st of all

let's select the device

that will be rendering the picture

I'll enable GPU compute mode

it'll allow me to render the picture

not by the central processor

but by the video card

let's move on

the sempling tab is the main one

where we adjust the quality of the final image

the preview tab is

where we adjust the preview render settings

and in the render tab we adjust the final render

the main image quality settings are noise threshold

and max samples value

if the noise threshold parameter is activated

it allows you to force render to minimum sample values

and the

noise threshold parameter defines the level of noise

at which the rendering will stop

in other words

this parameter works as a switch

which determines whether to use

the maximum number of samples

or the minimum will be enough

moreover in this tab

we have the possibility to activate the denois mode

but since we're working with fur and hair

the denois algorithm will just turn our picture

into a mess

so they cannot be useful to us in this case

if we want to render an animation

we need to enable animate seed

such an icon with a clock

this means that for each frame

a new pattern for the emission of rays will be created

if this option isn't enabled

the noise on a picture will be static

and it'll look very unnatural in the animation

in the light paths settings

we have

just a number of reflections for each type of ray

the more re reflections

the more accurately

the light in the scene is calculated

and therefore the quality of the picture increases

however we have to pay for this with rendering time

so I'd recommend you to leave these

parameters by default

or try to use presets to increase the

quality of rendering

the hair tab contains the parameters responsible

for rendering hair

in the shape parameter we specify the type of geometry

which will act as hair

now we use rounded ribbons as the shape

and it's the optimal one for our situation

the curved subdivision's parameter is responsible

for additional smoothing of the hair

during rendering

the world properties

tab is used to set the output of the final image

you can set the frame size in the format tab

and also a convenience manipulator

of the frame size as a percentage

you can use it to

increase or decrease the final frame size in multiples

in the output window

we can specify the path

in which the image will be saved

and the format of the file itself

in our case I use open EXR

this format is a standard

when working with two dimensional images

and its subsequent processing

it's possible to write the image with a wide

dynamic range of colors and related channels

also this format has a minimum compression value

that is

all information about the picture is saved

in principle

even at this stage we can click on the render button

and get the final image

but in this case

we'll get a pure beauty pass

in which it'll be very difficult to

edit individual values of brightness

color or saturation of individual elements

therefore there is a practice in computer graphics to

split the render

into separate layers

this allows you to work

more freely during the color correction

and global compositing phase

let's check how this can be done in blender by the way

the work with layers in blender is

organized quite clearly

the active layer is marked in an upper right corner

and we can create additional layers there as well

we can save a specifically configured view

of the scene on each layer

in order to do this

I need to regroup my light sources a little bit

I've already rearranged them into three groups

this is the filler the backlight and the background

I've created three layers with the same names

now I'll keep the collection of the same name

active in each of the layers

thus only key layer will have key light active

and only the background layer will have the background

light active

there is a similar situation with the backlight

I also want to point out one more aspect

after we created several layers

in the render tab

it's now

possible to change the sample value for each layer

if it's set to zero

the number of samples will be taken automatically

this is very useful

when you need to optimize a scene

for a better performance

I'll do a small test before rendering the final image

but since this is a test

we only need a small number of samples

now let's click on render and wait for the final result

unfortunately

this way we're forced to render every single layer

thus the rendering time of the final image

increases in proportion

to the number of layers we've made

nevertheless

this is all compensated at the compositing stage

when the render is finished

we can check each of the layers

in the frame buffer window by selecting them

in the top menu

now you can move to compositing layout

compositing

works on the same principle as the shader editor

therefore you shouldn't have too much trouble

though there are a few small nuances

1st of all we need to activate the use nodes checkbox

and reconnect

the node render layers

to update the rendering information

you can select any of the layers

in the bottom list of this node

to see the result of the compositing

let's create a node viewer and connect our images to it

after copying the node render layers several times

we select the desired layer in each one

to assemble

the two layers we need node mix

let's connect our images to it

and choose add in the mixing method

in the same way we connect the last image

if we compare the clean render and the assembled image

we can see that we got an identical result

we spend a lot less time on rendering

for the beauty frame

instead

in a stacked image we have huge editing possibilities

and we can literally change the lighting of the whole

scene

also it's necessary to mention the creation of masks

often

we have to edit individual elements during compositing

in order to do this we have to create masks

in order not to do it manually

we can use the cryptomat module

which will create

masks for the whole scene automatically

besides to activate this function

you have to activate all the layers

in the cryptomat parometers

and it also supports more fine tuned customization

but it's not relevant to us now

another thing worth mentioning is that in our case

I activate cryptomat

only on one layer

since we have the same set of elements on all layers

I activated cryptomat on the key light layer

and having switched to the compositing mode

so we can see

that now the key light layer has additional outputs

these outputs are exactly the information for cryptomat

currently they're empty

and we need to render the image again

in order to write the information there

I don't need to render all layers

but only the layer that has cryptomat on it

I'll use the render single layer function to do this

it'll allow rendering only the selected layer

when the render is complete

we can create a node cryptomat

which automatically works with this information

from the list of layers

we'll choose the one on which we have cryptomat stored

for clarity we plug it into the viewer

to see what's going on in it

and as you can see

we have some kind of color image here

this is technical information

so we don't care exactly how it looks

as long as the information is present

to create a mask for a specific object

click on the plus sign and select the desired object

then again

for clarity let's connect the mask in the viewer

this is what the mask of the background geometry

looks like

we can add and remove elements from the

mask with the plus and minus buttons

this allows us to make a mask of several elements

we'll consider further work with

masks at the compositing stage

but now let's talk about how to save all this treasure

the point is that all of the layers we rendered

except the beauty layer

are now stored in the blender buffer

and this buffer is cleared when you close the program

accordingly we'll lose all our renders

this is why we need to save each layer separately

this could be done manually through the viewer window

but it's not a productive idea

it'd be much more correct

to create unnote file outputs of each of the elements

and the separate output for the cryptomat layer

the output settings are

identical to the ones we have in the render window

all we need to create is the correct file path

but the cryptomat layer should be set up separately

to save it

we need to connect all kryptomat elements in note

file output

we'll do this by creating

an additional number of inputs

that corresponds to the number of kryptomat elements

let's connect all these channels one by one

and make correct names for each channel

that is the way we're going to render our final image

all we need to do is enable automatic samples

for each of the layers

and make all the scene elements active

by the way

if we aren't gonna use the beauty layer

we can leave it with a single sample

so as not to waste time on it

finally

let's click on the render image in the render tab

when the renderer of all the elements is finished

I'll create a new scene

exclusively under compositing

next I'll transfer all the rendered layers into it

I'll also create node kryptomat

switch it to image mode and select our file with masks

now let's create a node viewer

to display the information on the screen

and check the quality of all the layers

if the displayed picture doesn't fit on the screen

we can adjust the image size in the side menu

or use the handy handlers

now let's take a look at all the elements one by one

we have the background

the backlight

and the fill key light

and this is exactly how node kryptomat looks

with all the elements

let's immediately select the background mask

click on a plus sign

turn on the idropper and select the background element

if we switch to mask mode

we can see how our final mask looks

and now let's put the whole picture together

1st I'll create an RGB node

which will be the source for our setup

I make it black and create a node mix

I set the blending mode to add and merge the two layers

I can use the node exposure

to edit the brightness of the picture

this is a better approach

than using the node mix channel

furthermore I can use node exposure

to increase the brightness of the picture

to the values I want

in the same way I'll connect the other layers

if we put it all together correctly

our picture will look exactly like the final render

however I guess I'm not happy with this result

the color and contrast of the image is lacking for me

to fix it

we can use the color profiles

there is a color

management tab at the very bottom

of the render settings

by default it already has filmic mode

and all we have to do

is to choose the preset that suits us

let's try high or very high contrast

this way

you get a very vibrant and bright image

with a minimum of effort

the next step

I want to take is to make the background darker

and more saturated

but I'd

also like to leave a spot of color in the center

I'll try to solve this with node gamma

let's make the gamma values 2.7

in this way our picture became more voluminous

I should say though

that this method is a bad solution

because it distorts the gamma of the light stream

that was recorded in the layer

most likely it can cause a lot of artifacts in more

complex compositions

but as we are doing this

as a part of a simple color correction

such dirty tricks can work

and of course how could I forget about our cryptomat

let's create another node mix

we'll make the mixing method multiple

and connect cryptomat in mask mode

now by changing the color in a 2nd node mix channel

we can send our background in any color

so half of the job is done

we lack special effects for the final image

therefore let's create a node glow

we connect it to our setup

and switch it to the fog glow mode

and following the traditional way

will add chromatic aberrations

we'll do this with node lens distortion

but as you can see

this node has created not only chromatic aberration

but also global lens distortion

probably we don't need such imperfection of the image

that's why we activate the projector function

and increase the dispersion value

to show chromatic aberrations a bit more

was the m hotkey

we can deactivate the selected node

and compare the result

with and without the effect

perhaps further action in compositing makes no sense

so the color correction work can be considered finished

we only have to export our result

I'll create a node output file

where I'll specify the path to the folder

in which the image

and final format of the image will be rendered

in this case I'll use the PNG format which is

able to keep a good quality picture

with minimal file size

but in order to see the final image during rendering

which is helpful if we're rendering animation

we need to connect to our setup node compositing

finally

let's render the image

well this is probably the end of our course

thank you for your attention