Cinema 4D Training Introducing VRay: A short guide to the settings of VRayForC4D pptx

Next click on the tab that says IndirectIllumination figure 1.6 and check the first box that says Right click the light object in the Object Manager and scroll down to the VRay tags.. Fo

ZIP ARCHIVE PASSWORD

SherbetLemon

case sensitive

Project files can be found at http://www.thec4dvault.com/down/vrayguide.zip

Page 4 Page 5

TheC4DVault.com was created by Stuart Lynch and provides Cinema 4D training, resources and sample

files for the C4D community Stuart is a veteran Cinema 4D user with over 10 years of professional

expe-rience in the industry and several years providing instructional information to intermediate users

This first in a series of new short instructional pdf’s is aimed at providing the user with enough

informa-tion to digest without becoming overwhelmed It highlights the backend of VRay and relates the

impor-tance of understanding the settings that make VRayForC4D one of the best render engines around

Over the coming months, TheC4DVault pdf booklets will cover a wide variety of topics, providing a free

and reliable way to learn Cinema 4D

“I hope you enjoy this first addition and learn a thing or two about VRay in the process Thanks for reading.”

Stuart Lynch

http://www.project1media.com

Please make a minimal donation to keep the resource alive

*See Donate section at www.thec4dvault.com

You’ll notice now a new VRay Bridge Option

figure 1.5 has appeared for VRay on the left hand side

figure 1.5

Clicking the VRayBridge option will reveal

the many VRay bridge parameters geared

towards the final rendering setup

Next click on the tab that says IndirectIllumination figure 1.6

and check the first box that says

Right click the light object in the Object Manager and scroll down to the VRay tags From the list, you’ll want to add a VRayLight tag figure 1.8

figure 1.9

The information required to manipulate VRay lights specifically is now contained in the VRay light tag for future reference figure 1.9 You can double click this tag at any time

to retrieve this set of parameters

If you were to search images.google with

the term ‘VRay’, one of the first thing you’ll

notice is the abundant usage of the Physical

Sun/Sky setup

It’s a practical solution for lighting interiors

and will serve as our first demonstration of

the VRay engine at work

In figure 1.1 we have the Sun light parameters

which allow us to make changes to the Sun’s

intensity and it’s physical properties

For this example we’re simply going to

acti-vate the Sun/Sky and become familiar with

the lighting it provides for the scene

figure 1.1

In figure 1.2 I’ve created a primitive cube with

2 segments on each side The geometry has been edited and 2 simple openings have been extruded on the most distant polygons

The editor camera has been positioned in the center of the screen and I’ve exaggerated the field of view slightly to include both open-ings

Color mapping

To make a point stick with his students, my college instructor would occasionally use profanity

So let me say that using Color Mapping in your workflow will often save your ASS!

There are plenty of headaches to come, trust me, but with this simple workflow you’ll avoid blown out light sources that can make or break a render

It’s a straight forward concept to understand when illustrated with this amazing set of renders

In figure 1.13 The area light in this simple room appears to give off the correct light intensity and illuminates the portion of the room closest to camera But at the source, the light appears ‘nuked’

In figure 1.14 If we dial back the lighting intensity a little, the light at the source appears to have corrected itself, but at the expense of less light reaching the back of the room

figure 1.15

In figure 1.15 the light intensity is restored to the brightness of figure 1.13, color mapping is applied and the light adequately reaches the entire space Balance is accomplished

The VRayLight tag contains a lot of

impor-tant functionality and we’ll become familiar

with all of the options as we progress

For now, our intention is change the target

spotlight to become a Sun light with a

PhysicalSky

In the option Light type figure 1.4 Change from

a Spot light to an Infinite light

Also check Enable shadows

figure 1.10

The light now is as it suggests, an infinite

light and not yet a Sun Light In order for the

light to become a Sun Light we must click

the Sun Light tag and make a few

Focussing on the details of the render, it’s possible to notice that there’s a slight flaw in the accuracy of the lighting

The sunlight that enters the room looks overexposed, yet the subsequent light fall-off is too dramatic and occurs too quickly

Meaning, the rest of the room is too dark when compared to the intensity of the sun

Maybe it’s that the scene needs a fill light or perhaps we’ve missed the more GI ‘bounces’

option, after all it’s an interior scene

Not so fast!

The solution in Color Mapping

figure 1.12

Page 12 Page 13

It wouldn’t be fair to offer you this solution without some idea of what’s happening behind the scenes After all, you did buy the book from a desire to learn

So if we apply the ‘Indoor’ settings from the guidelines on the previous page, we can draw comparison to the original Sun/Sky render

figure 1.17 is the original without color ping applied (Gamma 1.0)

map-figure 1.18 in exponential mode with the line values (3.6/1.8/2.0) (Gamma 2.0)

The difference lies in the mode tial, this prevents colors from becoming too burned out and bright It’ll saturate the colors as they reach too high of an intensity The darkness multiplier value increase to 3.6 indicates that the original output of grey/blue will have be multiplied by 3.6, giving rise to the additional light and brightness in the scene

Exponen-The brightness multiplier is increased to 1.8Which indicates that although the exponen-tial value has reduced overall brightness, it can still be multiplied to compensate for the dip you see in figure 1.19

You can consider it a compromise

In figure figure 1.19 the brightness multiplier

is being left at its default value of 1.0 and the sunlight would be too low in its intensity

figure 1.19

FOR INDOOR SHOTS

Type: Exponential (some prefer HSV exponential)

If you’re not already aware of the implications of using a LWF - See the following PDF

http://www.pixsim.co.uk/downloads/The_Beginners_Explanation_of_Gamma_Correction_and_Linear_Workflow.pdf

By default VRay uses the ‘Color Mapping’ type (menu option Type) Linear Multiply with default Dark and

Bright Multiplier values set to 1

You will notice several other Types in this drop down menu

The most practical color mapping Type for interior rendering is Exponential mode (or HSV Exponential)

For exterior shots, it is recommended to use Reinhard

The following settings are a general guideline for what these numbers should look like

You can use find this function in VRay settings under Color Mapping figure 1.16

figure 1.16

Switch to the point tool and from the attributes manager uncheck ‘Only select Vis-ible Elements’

figure 2.5

In the Top viewport select the polygons that make up the front window and rear wall

Drag to the left as in figure 2.6

Repeat the process for the Right window as shown in figure 2.7

figure 2.6

In VRay it’s important to use as accurate a

measurements as possible

This basic cube is 500cm x 200cm x 800cm

and represents our small bedroom figure 2.1

Chapter 2: The test scene

In this chapter we’ll begin with a quick scene and will introduce the Physical Camera

figure 2.1

In figure 2.2 I’ve added 5 segments to the X

side of the model Edit the model (hotkey C)

and move the editor camera inside of the

model

figure 2.2

With the camera inside the model, select all polygons and reverse normals so that they’re facing inwards (hotkeys U~R)

figure 2.3

Selecting two faces as shown below figure 2.4

Use inner extrude (hotkey I) and extrude (hotkey D) to create two window openings

Here you could split the glass geometry for later (hotkeys U~S) or simply delete the two polygons

thin moulding feature figure 2.9

Select all ceiling polygons, perform a slight inner extrude and then a regular extrude upwards figure 2.10

figure 2.9

figure 2.10

Page 16 Page 17

figure 2.12

In figure 2.12, extrude the ceiling polygons to

match the moulding extrusion.

Perform a simple select loop command with

the polygon tool (hotkey U~L) figure 2.13 and

Extrude slightly at 91°

Don’t be concerned if your shot is nothing like mine If your scene has two openings I’ll forgive you

Moving on

figure 2.14 - ‘Test render for illustration purposes only’

Find a place where you’re happy with the viewport representation of your scene and add a regular Camera

figure 2.14

‘Roughing out’ the lights

Add a Target Light and position yourself in the top view, try to find a placement for the target light where you would like the Sun to exist figure 2.15

figure 2.15

Select VRay as your primary render engine

in RenderSettings and as in the earlier ample, check GI on in the IndirectIllumina-tion tab

ex-Add a VRay Light tag to your Target light, check enable shadows, change from spot light to infinite light In the SunLight tab activate the Physical sun and Physical Sky

Returning again to the editor view and ensuring you have you camera selected Run your first render figure 2.16

figure 2.16

Clearly not enough light entering the room

on this angle This issue of balance is more prominent than in my initial examples.It’s time to apply Color Mapping to bring the room to life figure 2.16

figure 2.17

REMINDER

FOR INDOOR SHOTS

Type: Exponential (some prefer HSV exponential)

Select the wall polygons in between the

win-dows and extrude figure 2.11

figure 2.11

figure 2.13

Bringing the shot to life.

From the Material Manager add a new VRay

Advanced Material to the scene figure 2.18

So far we haven’t included any materials in our scene VRay responds to color change quite dramatically,

as does color in the real world If we add a tiled wood texture to the floor, the bounced light from the

wood and its color values will have an affect on the scene Let’s explore further

figure 2.18

Assign the new VRay vanced Material to your cube object in the OM (object manager) figure 2.19

Ad-I’ve named my material ‘base’

figure 2.19

Double click the Material ‘Wood’

In the Diffuse Layer - Texture map slot

Load a wood tile texture figure 2.21

Create a second VRay Advanced Material

and change the name to ‘wood’

Select all floor polygons and drag the ‘wood’

material over the polygon selection figure 2.20

figure 2.20

figure 2.21

In the OM select (click) the wood material tag In the Attributes Manager, change the projection mode from Spherical to CubicUsing the texture axis tool and the scale tool

figure 2.22 Scale the wood texture to fit your scene

figure 2.22

Running a new test render, notice how the wooden floor is affecting the lighting in the rest of the scene figure 2.23

figure 2.23

Make a few adjustments to your camera sition and to the SunLight and try to balance both lighting direction and composition

po-figure 2.24

figure 2.24

In this next phase of scene creation, we’ll start blocking out some objects with simple geometry to get a better feel for composition

Place Cube primitive objects and position them where the bed, side tables and cabinet would potentially be

The presence of more white objects has changed the dynamic lighting of the scene once again figure 2.25

figure 2.25

Adding some additional materials with variations in color (preferably chosen to represent how you imagine your scene) we’re able to improve upon the overall feel of the shot figure 2.26

figure 2.26

With all of these geometry additions you may have noticed that the light in the room has taken a significant dip in brightness

On the next page we’ll discuss the Physical Camera and how to implement photographic values in order to restore the balance

Page 20 Page 21

Introduction to the Physical camera

The Physical Camera in VRay allows us to reproduce the parameters of a real world camera

To improve our scene without further adjusting the Color Mapping or the default lighting intensity of the

Sun, we’ll implement the Physical Camera and make use of its features such as ISO and FStop

From the OM select your scene Camera,

right click and apply a VRay Physical

Cam-era tag from the VRay Bridge Tags submenu

figure 2.27

figure 2.27

If you test a new render you would notice

that no significant changes have occurred

Though the default settings of the VRay

Physical camera provide a similar result,

they differ greatly from a camera that has no

Physical properties

The proof is in the numbers

In the VRay Light tag under the Sun light

parameters you’ll see two separate values

Intensity multiplier for Phys Cam : 1

Intensity multiplier for Std Cam: 0.037

The sun intensity is divided by exactly 270

to work at the default value of the standard

camera

It’s good practice to use the Physical camera

for all of your scenes

We’ll look more closely at comparisons in a later chapter, for now we’re occupied with increasing the exposure values of our scene without making any adjustments to light intensity

Click the VRay Physical Camera tagUnder the Tab - Lens parameters

Scroll down to Film ISO and change the value from 100 to 400

A test render produces this result figure 2.28

figure 2.28

figure 2.26 ‘Old result for comparison’

An increase in Film ISO values from 400 to

1600 figure 2.29 increases the film sensitivity further and subsequently alters the bright-ness of the scene

Changes in F-Stop values can also be made

to compensate for brightness

Shutter speed can typically be left alone, although it is an important value when intro-

ducing motionblur into your workflow

figure 2.29

If you’d like to continue with your scene and start placing actual geometry or re-working

the blocked out models to become realistic

fixtures in the room Feel free

figure 2.30

In figure 2.30 I’ve added a new bed and a some subtle blurry reflections (refereed to as Glossiness reflections in VRay) to the wood material

Although I’ll shortly be introducing more detailed information on the VRay Advanced Material, if you want to add this effect to your current scene, I’ll show you the quick way

Activate a Specular layer in your wood rial and input a Glossiness reflection value of 0.8

Chapter 3: Introducing VrayBridge render settings

Quality is very important when we think about final renders, but this shouldn’t always come at the cost

of lengthy render times Through exploring VRay’s render settings we’ll learn to tweak the parameters in

order to get the best results in an optimal amount of time

In this section we’ll optimize the VRay Render Settings and then explore the method behind the

mad-ness Open the example bedroom.c4d open from www.thec4dvault.com and lets begin

figure 3.1

From the Tab Indirect Illumination (GI)

select ‘01_IR-LC_very very fast’ from the GI

Preset list

IR refers to Irradiance Map

LC refers to Light Cache

In the Primary and Secondary bounces

you can see the two modes being selected

With this setup the Light Cache will be

cal-culated first figure 3.3

figure 3.2

After successful completion of Light ing, the Engine then begins computation the Irradiance Map

Cach-A simplified explanation of the Irradiance map is to imagine a point cloud that origi-nates from the Camera and projects out-wards into the scene

In figure 3.6 I’m using a VRay resource called IMapViewer.exe With this we’re able

to view the resulting IR map in its raw form

From the camera perspective in the bottom left corner, it’s quite easy to visualize the job

of the Irradiance map There are no samples outside of the Camera view and so we can conclude, it only samples what it needs to see

figure 3.4 ‘The familiar Light Cache calculation’

Although with the correct values, this

tech-nique is equal to Unbiased rendering

GI Preset 19_LC_LC_PPT_unbiased is set

up to render in this way

The Light Cache is an approximation of the global illumination in the scene figure 3.4

Objects such as walls and floors require very little IR sampling, but a detailed ornament

on a table needs more refined samples to reach the tight spaces

Providing adequate coverage is much less

of a struggle than it may seem VRay comes loaded with presets that provide an excellent starting point in most cases

figure 3.8

The DMC Sampler figure 3.8 is the VRay control Room These 4 values basically de-

termine how VRay uses its adaptive sampler,

which I’ll explain later

The engine itself uses deterministic Monte Carlo sampling (DMC), which is a high quality approximation GI algorithm The better the settings, the more reliable and ac-curate that approximation becomes

It’s not imperative that we learn exactly how the backend of VRay works What’s impor-tant is that we understand how these vast amounts of values will affect our end results We’ll need to break them down a little in order to make sense of them

Page 24 Page 25

Back to the DMC sampler, add the values

from figure 3.10

figure 3.10

Adaptive amount will be left alone

Minimum samples have been reduced from

8 to 4

Noise threshold was in an okay place with

0.01, so we won’t touch that either

Global Subdivs multiplier is set 0.5, down

from 1

In the tab Indirect Illumination (GI) click the Irradiance Map arrow to reveal more set-tings figure 3.11

change the Hemispheric SubD

Anti Aliasing (AA) is another important

fac-tor in determining final VRay render quality

The Adaptive DMC Anti Aliasing mode figure

3.9 and the DMC sampler figure 3.8 are tied

in very closely, especially when compared

to other Anti Aliasing techniques like the

Adaptive Subdivision sampler

For Interpolated samples enter a value of: 10

We can ignore the other settings for now

We’ve barely scratched the surface of this new information and I have you dialing in numbers that probably make no sense to you whatsoever

My instructional method is to teach through example Once you’re familiarized with the process it will start to become second nature Yet that is only part of the equation

It is also important to have a basic

under-standing of why each value works the way it does

In this next section we’ll adjust for a final render a along the way I’ll point out some of the method behind the madness

figure 3.16

If your render happens to look as it does in

figure 3.16, it would be because you haven’t yet applied Color Mapping Not mentioning it this was an intentional test

Creating a higher quality render

Reload the example bedroom.c4d

Attempt this without illustration (please)

#1 In the tab Indirect Illumination (GI)

se-lect ‘03_IR-LC_medium’ from the GI Preset

list and check the box GI On

#2 In the Irradiance Map sub menu

Min Rate: -3

Max rate: -1

Hemispheric SubDivision: 55

Interpolated samples: 15

Leave every other IR setting alone

#3 In the Light Cache submenu

Un-check Override LC Subdivision

Global Subdiv multiplier: 1

#4 In the Antialiasing tab

Antialiasing type: DMC sampler

Filter on: Check Filter type: Catmull-Rom

But was the result ready to send to a client?

Well, aside from the fact that the scene is pretty bad to begin with, I’d still have to go with “no”

There’s artifacting issues in a couple of areas, too much noise in the wood and black glossy metal and it needs a little more

It feels close but it could use a little more to sell it as final

Indirect Illumination - Irradiance map

Tempting as it may appear to ignore this section and rely upon VRay’s preset values, I would recommend

against it Presets only solve half of the equation, they have no affect on AA or the DMC sampler and they’re not intended to provide a complete setup for your scene

Instead see the presets as a good starting point The Medium quality preset is sufficient for most projects

and only requires a few minor adjustments to bring it up to speed

When rendering a GI shot in VRay, the engine requires two solutions to produce ac-curate GI results Those are the Primary and Secondary bounces

In figure 3.18 I’veturned off secondary bounces and left the Primary bounces as: Irradiance map

With only one GI engine actively calculating,

a correct GI solution has not been possible

This is visible as the dark areas in the image

Each of the approximation methods has its

own very specific parameters

The IR map (IR map - Irradiance map) has many

options for the management of correctly calculating samples, sample accuracy and de-tail enhancement It also has a section used exclusively for animation,

The first two parameters you come across are

Min rate and Max rate These settings relate

to the size of the IR map calculation

The Min value “determines the resolution for

the first GI pass” The Max value determines

the resolution for the last GI pass

A Max value of 0 would imply that the

resolution of the last GI pass is equal to the

dimension of your scene resolution

A value of -1 is equal to half the resolution of the dimensions of your scene resolution

A Min value of :-4 and a

A Max value of: 0, would create 5 passes during the caching process ( -4, -3 -2, -1 and 0)

The Max value 0 is typically not needed for less complex scenes

figure 3.18

So far we’ve used only two of the five possible GI engines in these slots The preset values highlight other combinations, though the IR and LC is our focus for the time be-ing

As Irradiance map is our primary bounce, we’ll introduce its settings first

Page 28 Page 29

You could ask why not just have a Min value

of -1 and a Max value of -1? Or a fixed value?

The smaller values in the Min setting (-4 for

example) are important for sampling the bigger

details (walls etc). The larger values in Max (-1

or 0) are for sampling the finer details

You would rarely encounter a need to use a

positive value like 1 in the Max setting If it

occurs to you to do so, it would be wise to

look elsewhere before concluding that this is

the solution to your rendering woes

Min and Max however, are only part of the

solution

In figure 3.19 the low settings I’ve deliberately

chosen are creating very obvious splotches

This suggests that are settings are way off the

mark and need to be re-worked

A Hemispheric SubDivision is part of the radiance map and is used to determine “the quality of individual GI samples” Relating only to the samples of IR map in this case

Ir-We can look here to improve upon our IR map caching process A good starting place when attempting a final render is around the

50 mark

Values upwards of 150 are not unheard of for problematic renders, though typically you’ll find that 50-80 is adequate

Interpolated samples refers to the tion of the IR map samples This calculation look towards the range of existing data and attempts to refine that process further

interpola-With this setting - Too high a value is likely

to blur the solution, Too low and the samples

will receive no interpolation

So we can determine that some Interpolated samples are necessary

To compliment the Hemispheric sion, a value of approximately 15-30 is usu-ally always sufficient

SubDivi-figure 3.19

In all of the above examples Hemispheric SubDivision has a value of 50

In figure 3.21 the Interpolated samples value is 1, which at close inspection is too few samples

In figure 3.22 the Interpolated samples value is 100 At this value the samples tend to become blurred

In figure 3.23 the Interpolated samples value is 20 This is practical amount for most renders

At this scale it is perhaps difficult to see a huge difference Also any remaining noise issues you’re seeing would then be an issue with the DMC sampler and not the IR Map

There are clearly other settings in the Irradiance map parameters, yet in the interest of not becoming bogged down with too much input, I’ll only briefly highlight their usage

To the left of the Min/Max rates lie the Intensity Threshold, Normal Threshold and Distance Threshold

These settings refer to individual sensitivities to light intensity changes, surface normal changes and changes in the distance between surfaces I would leave these values alone and let the provided preset

values guide you until you become more familiar

There are occasions with objects such as ‘thin blinds’ where adjustments to these settings will benefit your render

In the Detail enhancement section are a couple of additional parameters

This function as the name implies creates additional detail (to small detailed objects, not really useful in a typical scene).

Utilizing this process will force the GI algorithm to compute with Brute Force GI in places that are ficult for the IR map to reach

dif-When using this function it is often necessary to consider lower overall IR Map settings

With this option checked we’re dependant upon Detail enhancement to enhance our render, the ing settings would therefore only need to made be suitable for coverage of the larger objects

remain-Note: This is because large objects such walls/ceilings typically get adequate coverage from fewer samples

In my professional career I have used the Detail enhancement function no more than 3 times

Refining the IR map settings is how we

prevent splotches and other inaccuracies It

ensures that this approximation method (our

primary light bounce) will provide enough

samples to assist in a flawless final render

Note: Over estimating the values can

cre-ate unnecessarily long render times Under

estimating causes artifacting (Not to be

confused with noise)

figure 3.20

Look to the next page for some comparisons

Indirect Illumination - The Light Cache.

The light cache (once again) is the approximation of global illumination in the scene

Spot3D.com “The light cache is built by tracing many eye paths from the camera.”

figure 3.24

I have to admit the first time I saw this

menu, I felt a lot less threatened by it than I

did the Irradiance map options

In our current IR - LC setup we’re depending

on the Light Cache portion of this solution

to computer the secondary bounces

In figure 3.24 the Overide LC subdivision box is

checked by default

The value inside the samples ratio ( drop down

menu) and the Subdivision amount (greyed out

box) differ only in mathematical approach

They both determine how many paths are

traced from the camera

Un-checking this box and manually

insert-ing your own value is recommended,

espe-cially if the secondary bounces require more

attention

The number of paths traced from the camera

is the square of the subdivision value 1000² equals 1 million paths 2000² is 4 million and 10,000² is 100 million

The Samples Ratio values are based on the resolution of your scene In the case of the 1/1 setting in a scene that’s 1280x720, the Light Cache would produce 0.9 million paths

There are some instances where you’ll need

to go beyond the 1/1 Samples ratio amount and manually enter Subdivision values

Though 1/2 is quite sufficient in most cases

It is imperative to use a higher value during animation, but rare to exceed 4000

With scenes outside of the typical Physical Sun/Sky setup, where perhaps we’re using an abundance of light sources The subdivision values will benefit from a boost

figure 3.27

We’ve covered the essentials and I’ll provide

a very brief explanation for the remaining settings

Sample size determines the spacing of samples in the light cache (Leave alone for now)

Passes calculates the light cache separately

on each thread of your cpu (Useful for animation)

Store direct light This option stores and interpolates direct light This can be useful if you’re using and IR - LC setup

Use camera path This is for tracing the Light cache along a camera path and is beneficial for animation

Adaptive tracing Stores additional mation about incoming light for each light sample

infor-figure 3.28

Reconstruction parameters

These parameters control how the light cache

is used in the final rendering, after it has been calculated

Prefilter Checked on - Prefilter will be used Prefilter is applied once the cache is com-puter or loaded from disc

Prefilter samples More prefilter samples equals a more blurry and less noisy cache Filter type This is for tracing the Light cache along a camera path and is beneficial for animation

Use light cache for glossy rays If this option

is on, the Light Cache is used to compute glossy rays as well normal GI rays

This option can speed up rendering of scene that have blurry reflections by a significant amount

Enable LC retrace This option assists VRay

in preventing light leaks at the expense of slightly increased calculation time in the IR Map

Retrace threshold This value can be used in conjunction with ‘use light cache for glossy rays’ to help VRay dynamically decide when and where to use the Light Cache

Page 32 Page 33

The DMC Sampler

Having covered two of the possible GI engines and how their individual passes contribute towards the

quality of the final render, we need to talk about the big guns! The DMC Sampler

You can think of the DMC sampler as a global quality control

This group of 4 settings play a very important role in the rest of the workflow

figure 3.29

During render time, the DMC sampler is

looking out to your scene for additional

samples

These rays are computed in such a way that

if the values are too low, they will produce a

noisy result figure 3.30

figure 3.30

Values from almost every part of the render

engine are coming in through this channel

and being processed here

Whenever you create a light, a blurry

reflec-tion, use Depth of field or Motion Blur (to

name a few), you’ve actually been providing the

DMC sampler with more samples and

infor-mation about the scene

Looking at an Area light’s parameters (VRay Light tag) , it reveals that the DMC sampler will

be receiving 8 Subdivision samples from this light source figure 3.31

The actual amount of samples, is the square

of the subdivision value / 8² subdivisions = 64

samples

figure 3.31

The DMC Sampler can look at a series of

these samples and determine that it has enough information to not use them all.

This is the role of the Adaptive amount

Let’s imagine the DMC sampler sees 100 samples for the entire scene If we chose an adaptive amount of 0.8, the DMC sampler will always use 20 of those samples The other 80% will be considered internally by the engine, but not necessarily used

Control over this function is decided by the

“Early Termination Algorithm”

With an Adaptive amount of 1, we hand over

control of all of the samples (Fully adaptive)

With an Adaptive amount of 0, All 100

samples will be used without consideration

for the “Early Termination Algorithm” (Not adaptive)

In most cases, the default value of 0.85 for the Adaptive amount is more than sufficient

At 85% adaptive, this may seem like we’re

leaving too much to chance, giving too much control to the DMC sampler But VRay handles the process very smoothly and it’s not often necessary to go below 70% (0.7) Noise Threshold

The Adaptive process communicates directly

with the Noise Threshold amount This value determines the amount of noise that is ac-ceptable before it can tun its attention to the next pixel in the calculation

figure 3.35

figure 3.32 Noise threshold has a value of 0.01

figure 3.33 Noise threshold has a value of 0.1

figure 3.34 Noise threshold has a value of 1

A value between 0.01 and 0.004 is a good range for final renders

When the Noise Threshold is 1, the DMC sampler is satisfied that the desired amount

of samples has been met while it’s still ducing a very noisy value This amounts to

pro-an incredibly fast, but noisy render

The DMC sampler frequently needs more samples from the external elements that are difficult to compute

A metal material that has blurry values (Glossiness subdivs ) such as Brushed steel will

often require more samples to achieve noise reduction and a quality result

Instead of trying to compensate for this in the DMC sampler, we would simply increase the Glossiness Subdivs value in the specular channel of the Vray Advanced material to a value higher than the default 8 figure 3.35