Utilization of simulation of dynamic effects
to achieve artistic intention

SUMMARY

In my dissertation I deal with the topic of simulation of dynamic effects in the field of audio-vision.

The main goal is to increase the quality of achieving an artistic intention, to artistically control the outputs and to artistically style them. The aim is to examine the artist’s intention in comparison with physical reality (because the type of dynamic system may naturally exhibit chaotic behavior, or the artist’s intention may deny the laws of physics). The aim is to compare the possibilities of the artist in the realization of his intention, in comparison with the possibilities of computational models. The aim is also to examine the quality of available computational models in comparison with physical reality and to examine the layer of attributes of the resulting work of art that directly contributes to the perception of photorealism and stylization in visual effects (fluids, etc.) and further explore the use of these attributes to increase the quality of artistic output.

INTRODUCTION

As a creator of events and visual effects, I very often deal with situations in which accurate physical behavior and photorealistic outputs are not enough for the director or creators of audiovisual works, especially in the field of production and image post-production. For a more artistically interesting result, I have to bend the laws of nature that are transferred to the digital environment and stylize the result to some extent. Although I practically deal with the creation of these “stylized” visual effects, on a theoretical level, the possibility of simple stylization of complex effects is not reflected in the field of visual effects. For that reason, I would like to focus on it in detail in my dissertation.

1.          CURRENT STATE OF THE SOLVED PROBLEM

Currently, in the field of visual effects and specifically simulations, great emphasis is placed on artistically controlled outputs. Digital creation uses digitally created fluids, atmospheric phenomena, destruction and deformation of objects, or simulation of crowds of characters. Classical physical and mathematical computational models are used for control and their photoreal visuals, which can calculate each simulated particle as it would behave in reality. So flowing water is flowing water and not steam or honey. Fire is a blazing flame and not a fireball in a vacuum. The plate shatters and the shards scatter chaotically across the floor. However, reality is no longer enough for creators in modern films and series, the whole field is moving further and it is important for them to stylize photo-reality, but with the preservation of the original physically correct behavior. For example, the previously mentioned flowing water will no longer flow only along the base, it will break more dramatically on the stones, the drops will fly higher and further, or even flow uphill, it will levitate in the shape of the character and all this according to the wishes of the director / author.

1.1      Basic analysis

Defining both the initial and final state of the sequence, versus the natural course of this sequence.

1.1.1         Dynamic effect and artistic intent

The viewer is able to perceive the state of the dynamic system at time t0 as ordered, harmonious, as an artist’s work of art. If we let this system be subject to natural physical forces, a dynamic effect is created, i.e. other states of the system t1, t2, t3 up to ti as a sequence (animation, film). For the viewer, the time tx occurs during the sequence, when he already perceives the state of the system as chaos. The viewer could still perceive the resulting chaos as orderly, harmonious and as a work of art, but our artist’s intention is to insert into the sequence another state of the system at time ti, which will also be the author’s work of art, and the sequence will naturally flow into this state.

The artist’s intention is a dynamic effect, a resulting sequence (animation, film) of state t0, t1, t2, t3 to ti that meets the following conditions:

  • state t0 (initial conditions) and state ti (final conditions) are defined by the author
  • the author is not limited in any way when defining the state t0 and t1
  • the sequence t0, t1, t2, t3 up to ti acts on the viewer as if the system were subject to natural physical forces
  • the state in time ti can be the initial conditions for the next sequence
  • states t0 and ti can be static (high entropy, the system is at rest)
  • states t0 and ti need not be static (low entropy, the system has energy, speed, acceleration, deformations are taking place, etc.)
  • the entire sequence is computer generated, including t0 and ti

1.1.2         Artist’s intention versus physical reality

An artist may discover two types of conflicts when comparing his intention with reality:

A dynamic system type exhibits chaotic behavior

Thus, a small change in the initial conditions leads to a very different result over time. This behavior is difficult for artists to influence. Sometimes the artist’s intention can seem unrealistic. Example of a conflict:

  • t0: thrown ceramic plate on the ground
  • ti: shards from the plate formed a silhouette of a dancing girl holding a fan on the floor

 

The artist’s intention defies the laws of physics

Thus, there is no configuration of input conditions such that in reality the sequence reaches the point ti. Example of a conflict:

  • t0: a freely thrown soccer ball on the first floor
  • you: the balloon bounced up the stairs to the tenth floor, and the movement seemed natural and realistic, like on planet Earth

 

We want to closely resemble the laws of physics, match or agree with them almost exactly. At the same time, we inevitably want to violate them. The relevant laws of conservation, laws of classical mechanics, laws of gravitation, thermodynamics are:

  • Laws of reflection
  • Friction (Amonton’s three laws of dry friction)
  • Fluid friction
  • Newton’s law of viscosity
  • Hooke’s law (spring)
  • Euler’s laws of motion
  • Euler’s equations (rigid body dynamics)
  • Archimedes’ principle
  • Bernoulli’s principle
  • Poiseuille’s law
  • Stokes’ law
  • Navier–Stokes equations
  • Faxén’s law
  • Newton’s law of gravitation
  • First law of thermodynamics
  • Second law of thermodynamics

1.1.3         Physical reality versus computational model

The behavior of the computational model adds additional kinds and levels of unpredictability. The model appears random, chaotic, although it is deterministic in the sense that it is well defined and contains no random parameters. Two close trajectories in phase space diverge (exponentially) with increasing time. Even a dynamic system that appears to be predictable in reality (a pendulum on a material fiber in a liquid) can have a chaotic behavior in the computational model, an unpredictable position in time ti . Types of conflicts between reality and the computational model:

  • the model must “sample” when describing reality (i.e., for example, replace the mentioned material fiber with a limited set of points, replace the liquid with square voxels, or possibly flowing particles)
  • the model contains pseudo-physical attributes. For example, the Euler model and fluid pressure: not only is the pressure not the same everywhere, but it serves as an intermediate result of the compensation of non-divergent flow, which we can also influence local positive and negative “unreal” divergence
  • the model contains empirical rules that only apply in certain metrics (for example, surface tension, viscosity), and the model calculates with these metrics
  • the model contains hidden calculations whose attributes cannot be influenced and may be deliberately “blackboxed” by the manufacturer’s know-how
  • the model is too narrowly specialized (particle simulation where each particle is a character in a crowd, simulation of waves on the high seas)

Examples of conflicts:

 Simulation of liquids and gases:

  • viscosity problem in Eulerian simulation model (voxels)
  • problem with explosions in Lagrangian simulation model (particles)
  • gradual loss of volume during a sequence in the Fluid-Implicit Particles model
  • free slip boundary condition
  • no-slip boundary condition
  • fluid compressibility

Softbody simulation (plastic/elastic fiber/area/volume):

  • very different results at different degrees of resolution – FÉM (finite element method)
  • repulsion and at the same time contraction of individual points during collision and friction
  • tearing the object

Rigid point simulation (fixed objects):

  • relative velocities versus momentum and rotation (problem of small vs large particles)
  • explosion when required to comply with restrictions (enforce constraints)

1.1.4         Computational model versus artist’s intention

After choosing the models and trying to realize the plan, the artist faces the following problems:

  • inconsistent outputs at different sampling settings
  • illogical behavior of the system when trying to set attributes (including pseudo-physical ones)
  • lack of knowledge (or inability to deepen knowledge, thanks to blackbox solutions)
  • lack of resources (computational power, time)

1.2      Houdini Software Specifics

We assume that the artist will choose Houdini software to realize his intention. This sw meets the requirements for the creation of an author’s work as well as for the advancedness of computational models and the ability to influence model attributes and create other custom computational models and tools.

Typical challenges

From their point of view, the viewer or the author expects the plot to develop in a certain way:

 A challenge involving fluid simulations

  • the sugar falls into the milk and the splash of the crown forms a cow
  • the water flowing from the kettle creates the shape of a flower

A challenge involving fire / smoke / steam simulations

  • the steam rising from the coffee creates a loop shape
  • the flames of the fire form the shape of a heart
  • the smoke rising from the cigarette forms the shape of a scalpel

A challenge involving particle simulations

  • flour thrown into the air forms the shape of a deer
  • raindrops form the shape of a transport vehicle in the air

A challenge involving Rigid point simulations

  • the plate falls, breaks, and the shards form the silhouette of a knife
  • mountain faults, rocks slide and fall to form the shape of a human figure

A challenge involving Soft body simulations

  • the hair, after being cut with scissors, falls to the ground and forms letters
  • the balloon pops up the stairs, deflates and creates a shape

A challenge involving crowd simulations

  • flying birds form the silhouette of a hand with their wings
  • running rats, viewed from above, form the silhouette of a pipe (instrument)

A challenge involving growth simulations

  • tree branches grow to form the outline of the wolves head
  • the vine grows on the building and shape of the silhouette of the glass

 

1.3      The current state of the solved problem

1.3.1         The situation from the point of view of technology

Physical simulation of dynamic systems according to produces unpredictable behavior that is difficult for artists to control. Systems are sensitive to initial conditions. The behavior of these physical systems, exhibiting chaos, appears to be random, even though the system model is deterministic in the sense that it is well defined and contains no random parameters.

Due to the final configuration of the fluid, some methods proceed backwards in time. Creates a sequence that is visually similar to traditional forward simulations when played forward. This allows artists to create simulations with rapid iterations. They then correspond to the art form in the group of simulation moments.

The current methods of project implementation are as follows. While trying to realize the plan, the artist encounters the conflicts mentioned in the previous points. Based on experience and knowledge, he chooses his path and uses one or more methods:

  • trial and error method
  • brute force generating combinations and choosing the most watery one
  • approximation by combining multiple results into one another (in time and space)
  • restrictive limitation of attributes during the sequence (loss of natural appearance)
  • hand animation (high labor)
  • creation of own computational model
  • approximating physical behavior mathematically (for example, jumping a balloon abs(sin(t)))
  • correction in post-production

1.3.2         The situation in terms of research questions and hypotheses

Research, add specifications TODO.

The user would appreciate the editable layer, the application of which will create the desired result that has the original natural properties of the given element. A generalization of an artistically editable layer that will automatically predict ideal options for changing parameters from a physically accurate model to an artistic model. A possible way is, for example, to manually add a layer of turbulence or controllable airflow according to the curve, which after a specified time will gradually start to work and change the smoke into the desired shape.

Hypothesis:

  • Computational models can be supplemented with a user layer that allows artistic influence by simulating a dynamic effect.

 

Specific areas of research:

  • categorization of intention (according to t0 ti)
  • typification of methods for individual computational models and intentions
  • creation of tools for individual computational models
  • the examined sample will be the creation of a simulation
  • a tool for artistically influencing fluid simulation
  • a tool for artistically influencing gas simulation
  • a tool for artistically influencing a rigid point simulation
  • a tool for artistically influencing soft body simulation

 

Research methods – experiments:

  • Creation of source data and assignments for processing by entities that deal with simulations professionally.
  • assigning simulation creation to different entities (or their parts)
  • detailed description of the experiment approaches, advantages, disadvantages. Preparation of output for further processing.

 

Comparative analysis :

  • Comparison of simulation efficiency, using common methods and using a layer
  • Testing the outputs of the experiment
  • Questionnaire as part of tool testing – finding out ergonomics

 

Possible solutions:

  • Generation of input conditions t0
  • Modification of the physical properties of the environment in time and space

1.4      Examples from practice

The most common keywords and situations emerged from the initial questionnaire survey:

  • water flows into the triangle
  • RBD impact of piece of object on music
  • flames around the sword
  • crowd simulation
  • particles to an event depart to another event to arrive
  • a number of events synchronized to music
  • scale and scale the effect
  • patterns in nature (sunflower, seashell)
  • scale / observer-gun triangle
  • expansion, implosion
  • dividing, joining (gestalt)
  • PDG (procedural dependency graph) data generation
  • PDG concept generation (detail, anticipation)
  • energy transfer between types of movement (motion, volume, rotation, volume, deformation)
  • explosion story (time stretching, exaggeration)
  • the story of the dam breaking

 

Research , complete specifications TODO.

The simulation problem Probability
(see later in the text)
Flip Fluidsm, Mograph style setups, PDG – TOPsm, Procedural Animationm, Prodedural Modelingm, Pyro – Smokem, Pyro – trailsm, Pyroclastim, RBDm, Retiming simulationsm, Riggingm, Colonisation – Infection – growthm, Space colonisationm, Tornadom, Vector Velocity Fieldsm, Vellumm 10/10
Lightning FX 9/10
Differential Growth, Ocean – Terrains, Particles & Grains, Solver Sop, SOP Solver in DOP, Curly Abstract Geometry, Ray marching over particles 8/10
Gears, IceBreak, Maelstrom, Procedural Graffiti Art, Worm locomotion 7/10
Tying and Knot 6/10
Noises – Displacement, Bubbles, WIRING, Volumes – VDB 5/10
Organic Modeling 4/10
Rendering – Materials – uvs, Neurones, Rectangle divisions – Wall patterns 3/10
Vector displacement 2/10
Crowds, For each – copyToPoints, L-Systeme & Vegetation, Rotation matrix transform 1/10
COPs, Mathematics, Dart Throwing – Packing Circle, Galaxie et Nebuleuses, Axes Origins centroid, GRID FLOW 0/10

2.          Objectives of work

2.1      The main goal of the work

The main goal of the presented concept of the dissertation is to investigate the use of computational models to achieve artistic intent (visual rendering). Investigate the layer of attributes of the resulting work and physical phenomenon directly involved in the perception of photorealism in visual effects (fluids, etc.) and the possibility of using these attributes to achieve artistic intent (look). Search for combinations of computational models that preserve the attributes of photorealism even under unrealistic physical conditions and intentions.

The specific output is a digital tool for the Houdini program (Houdini Digital Asset, hereinafter HDA), which, thanks to my research and generalization of physical regularities and regularities in the perception and assessment of the quality of photorealism, would be able to very easily shape effects, place phenomena in shapes or trajectories, change their shape in time and space and stylize them.

2.2      Secondary objectives of work

The secondary goals of my work include:

  • motivation to contribute to the audiovisual industry, production and post-production studios, schools and students to a higher quality of presented works
  • building awareness of the use of physical laws and computational models for controlled visual rendering of visual effects
  • an attempt to design and test a methodical approach to individual types of assignment, including the creation of a metric formula
  • outputs of artistic as well as technical applications in the Houdini software

3.          Research questions & research methods

The topic of simulation of dynamic effects in the context of audiovisual creation is primarily viewed in this work through the following factors:

  • mind set
  • human resource set / human capacities
  • tool set / technological means
  • communication tools / communication

Using the proposed methods (see below), the following outputs are created:

  • research of the current state of use and the possibilities of using computational models and their combinations in the creation of visual effects
  • HDA digital tool for Houdini
  • the text of the theoretical part of the dissertation summarizing the research
  • compilation of audiovisual works in which I participate or will participate in the positions of post-production supervisor, vfx supervisor, production manager or technology consultant.

3.1      Methods used

During research, a new product is created that will be used by artists. This new product will need to be tested. There are several phases to this research and testing. For qualitative and quantitative research, the sample size is determined, where, when, how and why someone was approached and the key questions are evaluated. It is important later to specify the focus of the research (especially the primary one – who, when, how and why will we ask, how will it contribute to the fulfillment of the topic, etc.).

Research before developing a tool

At the beginning, the first step is to gather input on what this new tool is supposed to be able to do. One or two investigations can be done in this step. This will more precisely specify the target group of artists and possibly introduce a new target group for this tool.

  • qualitative research, interviews with potential users
  • in case it is unimaginable for the user (due to the innovativeness), this step is omitted
  • a questionnaire survey is being prepared for users who have already encountered a similar tool (those who can imagine a future extension), how much they would value the new tool

Investigation during tool development

In the second phase, the tool is developed. Partial functional units, prototypes and development versions (alpha and beta versions of the tool) are created. Previews of the user interface are also created, which can also be used to verify the direction of development. Here it joins as a method:

  • user testing (assigning specific tasks to artists, verifying whether and how they will complete the task using the tool)

 

Investigation after putting the tool into practice

Here, the success of achieving goals is already being verified:

– how the tool suits the artists

– what they use the tool for

– where he sees potential for further development

 

List of methods that are generally used for the processing of work:

– an interview to gather ideas and inspiration

– a questionnaire to verify whether the majority thinks so

KOZEL, Roman, Lenka MYNÁŘOVÁ and Hana SVOBODOVÁ. Modern methods and techniques of marketing research. 1st ed. Prague: Grada, 2011, 304 pp. Expert (Grada). ISBN 978-80-247-3527-6.

Methodical procedure

  • formulation and specification of working hypotheses and research questions
  • preparation, creation and compilation of materials for testing (Houdini source files from studios, from artists, from the manufacturer of the Houdini software, from its beta testers, in the Czech Republic and around the world, as well as own creation of other combinations of source files)
  • research on the use of computational models and their combinations (in the form of a questionnaire and a study of source files)
  • testing (resulting appearances of simulations from obtained source files and examination of the degree of achievement of the intention)
  • evaluation testing
  • research into the theoretical possibilities of computational models and their combinations (by studying the literature, using questionnaires with users, with the Houdini producer team and beta testers, and by interviewing artists)
  • creation of advanced combinations of computational models and own computational models
  • creation of a tool in the Houdini software in the form of HDA (described above as the primary goal)

3.2      Research questions

  • Can intentions (t0 – ti), computational models and methods be categorized?
  • Can tools be created for these categories that will create a layer over the computational model that will allow the definition of both the initial and final state of the sequence, and at the same time guarantee the natural appearance of this sequence?

3.3      Purpose and benefits of research

The resulting possibilities of automatic control of energy conservation during forced stylizations, and the possibility of automatic compensation by adding other phenomena that improve the quality of perceived photorealism. For example, automatic change of color, state or even ignition when the temperature is exceeded, automatic deformation and destruction when the momentum changes. Stylization options with an unrealistic degree of this automatic compensation (centrifugal force of a turning car offers a crushed flying road, or even flames from tires and felling of nearby trees, the loss of momentum of the hero offers the demolition of the wall behind him, with a visible pressure wave).

4.          Research part

4.1      Possibilities of asynchronous control of the event in time

The user often needs to arrange events that take place in time in his timeline, similar to how a conductor only plays instruments when he needs them. The user therefore uploads event samples, for example, collision of two objects, destruction, explosion, smoke, escape of residents. He has recorded all these events at time 0, but for the viewer, he releases them overtimed into his timeline according to his artistic intention.

4.1.1         Analysis of options

A suitable tool for processing this issue is the so-called spectrogram, which, in addition to the wavelength of any sound, can also display graphically processed melody and graphically processed frequency over time.

4.1.2         Houdini Software Specifics

Houdini allows all events to be normalized and retimed. So the artist simply lets go of each event using a normalized 0 to 1 chart.

4.2      Possibilities of orthonormal bases in space and time

The artistic intent and story often contains an illogical arrangement of physical forces and objects. When flying home, in one room gravity may act from above, in another room gravity may act from below, and in a third room from left to right. So in the first room it can snow from the ceiling, in the second room it rains from the floor up, and in the third room papers fly from left to right and levitate.

4.2.1         Analysis of options

For the artist, when realizing his intention, it seems to be the best possibility to cancel each of these local coordinate systems into the origin of the global coordinate system. Individual relationships, positions and rotations and scaling of all local coordinates must be easily editable and the overall work with them clear.

4.2.2         Houdini Software Specifics

In Houdini, it is possible to automate and monitor all connections between bases. To simplify the whole process and facilitate the artist’s work, all bases will be considered orthonormal, that is, three-dimensional, the base vectors will be perpendicular to each other and will all have the same normalized size. Therefore, it will not be possible to use Einstein’s theory of relativity, and when the bases are reduced and enlarged, their mutual distances will change.

4.3      Options for automatically defining camera positions in space and time

When working with space, with bases, the artist needs to define the positions of the camera and individual objects in the scene. Camera and objects are always bound to one and only base. When transforming the base, all child cameras and objects are transformed.

4.3.1         Analysis of options

The artist has two options to define the camera path. The first option is to directly define the position and properties of the camera in time. The second option is to define a route. If he chooses the second option, he must additionally define additional parameters – the parametrically generated target position and the parametrically generated position of the camera location in time.

4.3.2         Houdini Software Specifics

Houdini enables the automatic generation of editing and transformation handles for each custom tool, and therefore our tool, which is the result of a dissertation, will have these practical automatic handles.

4.4      Possibilities of simulating the mutual movements of objects and the camera

What the viewer sees and therefore what the artist intends is a mutual constellation of the object and camera parameters. During his work, the artist needs to move the camera in time and space. If he wants to preserve the speed of the camera movement and if he wants to preserve the time in which (for example, a static) object reaches the center of the frame, then the artist encounters a problem when he needs to move the static object. From the basic equation “path equals speed times time” it follows that the camera will arrive somewhere other than the displaced object.

4.4.1         Analysis of options

There are two possible solutions here. First: changing the object’s relative size. If the camera will travel at a speed of 100 km/h along a giant model of an ant (for example, with one antenna 20 m long), then the viewer gets the impression that the camera is traveling slowly. The second possible solution: keeping all mutual realistic sizes of all objects. In this case, it is necessary to supplement the artist’s user changes with an automatic layer that intelligently calculates paths, speeds and times so that the artistic intention is fulfilled.

4.4.2         Houdini Software Specifics

In Houdini, a so-called event handler can be linked to each user action, which starts an intelligent Python script. The dissertation is therefore based on the second option from the previous balance sheet, i.e. maintaining all mutual realistic sizes of all objects. The advantage of this path is also when used in virtual reality or 3D rendering.

4.5      Possibilities of a universal data format for representing an object in space

During the entire development of the industry, many formats for the representation of a 3D object were created. Below is a breakdown into the most common categories. Each of the formats has its justification for a certain kind of object and a certain kind of editing. The mutual compatibility of individual categories is described in the table. However , the artist for whom we are preparing a universal solution needs a format that meets compatibility across all categories.

  • Polygon and Quadric surfaces
  • Volumetric (Voxel, SDF Isosurface, Octree Encoding)
  • Point cloud
  • Spline surfaces and construction
  • Procedural methods (Fractals, Lsystem, CAD Sweep)
  • BSP (Binary Space Partitioning) Tree
  • Implicit (Algebraic)
  • Depth map (Z−Buffer)
  • Constructive Solid Geometry (Hierarchy of boolean set operations)

4.5.1         Analysis of options

A format that will allow the artist to perform universal transformations must meet the following criteria:

  • is universal for any object and any material substance (with different physical properties and therefore also optical) and/or its state (solid, liquid, gas)
  • can convey the impression of realism or stylization to the viewer
  • it allows the impression of an animated change between any material substance and or state

4.5.2         Houdini Software Specifics

Houdini software has a sophisticated data import and export system thanks to its long-standing leadership in processing simulations and animation effects in the film industry. Such a representation of a 3D object that would meet the criteria in the previous point, but does not yet exist and is the subject of the HDA tool proposal in this dissertation.

4.6      Possibilities of partial movements of parts of objects and their effect on stylization and realism

In order to enable the artist to fulfill his intention (in our case, primarily the transformation of one object/substance/group into another and the stylization of this transformation), we must add a user layer to the universal format (described in the previous chapter 4.5), in which we allow the artist these changes and stylizations implement. The goal of this dissertation and the proposed tool is to make these changes not only real, but also easy and understandable and memorable in everyday practice .

4.6.1         Analysis of options

Completing this subtask goes hand in hand with the previous and next step. The previous step 4.5 was to create a universal format. This step 4.6 is the animation of the transformations and the paths of these transformations towards the camera (object/substance/group). This step 4.6 is not yet the final product that the viewer will see. Only the following step 4.7 will be the visual side of this transformation.

This subtask 4.6 can therefore be described as the invisible backbone of the transformation (morphing), on which realism and stylization are “dressed” in the next step.

4.6.2         Houdini Software Specifics

Houdini contains a huge number of low-level tools, that is, building blocks that are close to the machine instructions of the processor and provide the artist with a low level of abstraction or almost no abstraction. That is, an abstraction from how the computer processor and other hw (gpu, ram) work. All of these low-level tools serve as building blocks for higher-level tools that are already close to how humans process problems with their thinking. The artist, who is the target group of this dissertation, is used to working with a high degree of abstraction. The tools created in this sub-step 4.6 enable him to:

When art directing simulations, he will be able to follow trajectories that are otherwise completely independent of each other, change and reverse the course of trajectories (for example, instead of simulating destruction, create a reversal, i.e. composition), detect and emphasize turbulence, add burning energy to selected places, create swirls in smoke and perform other various abstract actions at your discretion.

4.7      Possibilities of generating various substances along the paths of movement in order to create an overall impression of a natural phenomenon

This step is the third and final step in the overall task of “using computational models to achieve artistic intent”. Here, the basic invisible skeleton of the transformation (described in point 4.6) is complemented by the visual aspect, including the artistic intent of the stylization and/or realism of this transformation.

4.7.1         Analysis of options

Since we are dealing with the visual perception of the transformation here, we must return to the first step of the transformation and therefore the complete spectrum of all possible input data and their mutual combinations (from point 4.5). Technically, it is now necessary to allow the artist to link all forms of substances and states to the basic skeleton.

4.7.2         Houdini Software Specifics

Throughout his many years of development, Houdini perfects himself in an attempt to imitate the appearance of all things and give the viewer the correct impression according to the artist’s intention. It might seem easy to use existing display systems. However, these systems have their drawbacks:

  • Demand for calculation (rendering)
  • Specialization for a certain state and substance of the material, i.e. a narrow specialization for, for example
    • solid substance
    • smoke
    • fire
    • liquid
    • viscous liquid
    • sand
    • hair
    • textile
    • etc

This dissertation creates new display systems that allow artists to respond more quickly and iterate faster in their work.

4.8      Options for further processing of data, routes and substances in the Houdini program

Houdini contains a large number of tools from the complete pipeline of a large industrial studio.

4.8.1         Analysis of options

Since the conversion system from any other format to our new format was created in step 4.5, it is necessary to create a conversion back to as many formats as possible in the table.

All the steps and the artist’s user edits still take place in the software in Houdini, and the artist can at any time use any other part and tool (so-called Operator) of this extensive software:

  • SOP – Geo & Obj or Surface Operators
  • COP – Compositing Operators
  • ROP – Render Operator
  • GTC – Mat & VEX Operator
  • PDO – Channel Operator
  • DOP&POP – Dynamic Operator & Particle Operator
  • TOP – Task Operator
  • LOP – Lighting Operator
  • VEX – Vector Expression Language

4.9      Options for exporting data, routes and substances for processing or rendering in other software

The industry is still developing and there is an interest in the fact that the data created by the artist can be further processed in other software, used in other simulations, exported to virtual reality, used in game engines and the like.

4.9.1         Analysis of options

Due to its good position in the film industry, Houdini is also establishing itself in the gaming and entertainment industries. It works closely with these software and it is possible to create a so-called livelink with them:

  • Unreal
  • Unity
  • Maya
  • 3ds Max
  • Cinema 4D
  • Blender
  • ZBrush
  • Substance

5.          Practical part

To implement the plan, I choose Houdini software, which is a standard in the visual effects industry. This software meets the requirements both for the realization of the author’s intention and for the advancedness of computational models and the possibility to combine computational models, to influence them in any way, and also to create own computational models and tools. Houdini is a comprehensive software covering the entire pipeline from modeling, animation, shading, lighting, rendering, VFX and compositing. It is used by UPP, PFX, Kredenc, Blue Sky Studios, Framestore, The Mill studios to create visual effects.

5.1      Digital Instrument (HDA)

5.1.1         Basic description of the tool and its functions

5.2      Node Group A – xasync

def xasync init(OBJ)

def xasync array (SOP)

 

5.3      Node Group B – xbasis

def xbasis

 

5.4      Group C – xcam

def xcam view

def xcam waypoint

5.5      Node group D – xdriver

def xdriver

 

5.6      Node Group E – xessence

 

def0 null

def xe0 source bake

def xe0 source sdf copies

def xe0 source intersect planes

def xe0 source inflate from points

def xe0 source simulation

def xe1 seams from curvature

def xe1 seams from direction

def xe1 seams from simulation

def xe1 seams from files

def xe1 seams from folders

def xe1 seams from unshared edges

def xe1 seams ornaments

def xe1 seams growth

def xe2 extend & trim

def xe2 advect on surface

def xe2 relax on surface

def xe5 folders from spreading

def xe5 folders from pairing

5.7      F node group – xflow

 

def xf pin (by camera, by plane, by time, by distance)

def xf inject (from pin, by distance, by time)

def xfsmooth()

def xf retime (reverse)

def xf detect flames

def xf detect turbulence

def xf detect energy

def xf trim (from pin, by distance, by time)

def xf create connection

def xf create along path

def xf create swirl around

def xf create swirl

def xf create waypoint

def xf create follow

 

5.8      Node Group G – xgen

def xgen init (input geo cover)

def xgen boxel

def xgen metaball

def xgen grain

def xgen brush

def xgen instance

def xgen hair

def xgen crowd

def xgen dop I/O

def xgen fracture

5.9      Houdini TOP, DOP

5.10Houdini ROP & Houdini Engine Experiment

5.10.1    Research sample

5.10.2    Criteria

5.10.3    Description and interpretation of results

5.10.4    Proposals, conclusions and recommendations

5.11Conclusion

5.12References

Discrete and continuous simulation: theory and practice
Bandyopadhyay, Susmita, Bhattacharya, Ranjan
Boca Raton : CRC Press, [2014]
ISBN: 9781466596399


Discrete-event system simulation
Banks, Jerry
Upper Saddle River : Pearson Prentice Hall, [2005]
ISBN: 0131446797


Feedback control of dynamic systems
Franklin, Gene F., Powell, J. David, Emami-Naeini, Abbas
Upper Saddle River : Pearson Prentice Hall, [2005]
ISBN: 9781292068909


From chaos to order: methodologies, perspectives, and applications
Chen, G., Dong, Xiaoning
Singapore River Edge, NJ : World Scientific, [1998]
ISBN: 9810225695


Chaos and nonlinear dynamics: an introduction for scientists and engineers
Hilborn, Robert C.
Oxford : Oxford University Press, [2000]
ISBN: 0198507232


Modeling complex systems
Boccara, Nino,
New York, NY : Springer, [2010]
ISBN: 9781441965615


Modeling random systems
Cogdell, JR
Upper Saddle River, NJ : Prentice Hall, [2004]
ISBN: 0131414372


Modeling and simulation of complex systems: how to better understand the world
Pélanek, Radek,
Brno : Masaryk University, [2011]
ISBN: 9788021053182


Nonlinear systems
Chelikovsky, Sergej,
In Prague : Czech Technical University, [2006]
ISBN: 8001034356


Numerical methods for solving flow problems II
Kozel, Karel, Fort, Jaroslav
Prague: Czech Technical University Publishing House, [2003]
ISBN: 8001026752


Numerical simulations of dynamical systems: a supplementary study
Černá, Růžena, Peterka, František, Čipera, Stanislav
Prague: Czech Technical University Publishing House, [1995]
ISBN: 8001013707


Principles of modeling and simulation: a multidisciplinary approach
Sokolowski, John A., Banks, Catherine M.
Hoboken, NJ : John Wiley, [2009]
ISBN: 9780470289433


Randomnicity: rules and randomness in the realm of the infinite
Tsonis, Anastasios A.
London : Imperial College Press, [2008]
ISBN: 1848161972


Stabilization and motion control of unstable objects
Formal'sky, Aleksandr Moiseevich
Berlin Boston : De Gruyter, [2015]
ISBN: 9783110375824


Statistical thermodynamics
Boublík, Tomas
Prague : Academia, [1996]
ISBN: 8020005668


The art of modeling dynamic systems:
forecasting for chaos, randomness, and determinism
Morrison, Foster
Mineola, NY : Dover Publications, [2008]
ISBN: 9780486462950


Introduction to Nonlinear Physics
Fiala, Jiří, Skála, Lubomír
Prague : Matfyzpress, [2008]
ISBN: 9788073780517

Continuum mechanics: density, navier-stokes equations,
momentum, rheology, fluid, non-Newtonian fluid, Laminar
flow, surface tension, ductility, fluid statics, shear
stress, plasticity, metamaterial, finite strain theory,
acoustic metamaterials, viscosity Memphis : Books LLC, [2011]

ISBN: 9781157430056