Impact of Concept Art on the Character Creation Pipeline

Abstract

This thesis focuses on the character creation pipeline in the game industry. It aims to identify the impact of concept art completion on the efficiency and quality of character artists and provide suggestions for improving the pipeline. To narrow down and define the concept art stages, examples and descriptions are created and used.
By means of a practical test with character creation students, cross-referenced with data from interviews with artists and managers, conclusions were drawn. The data shows that industry professionals cannot distinguish sculpts created with a low amount of concept art from sculpts created with a high amount of concept art. This indicates that the stage of concept art reached does not impact the quality and efficiency of character artists in the first two hours of their sculpting process.
There is a disconnect between current pipelines in the games industry and the results of this research. With the data gathered, suggestions for alternate, less linear character creation pipelines can be created.

Full text

MASTER THESIS
Impact of Concept Art on the Character Creation
Pipeline
Author:
Jef Bernaers
Supervisors:
Thomas Buijtenweg MSc
Dries Deryckere MSc
This thesis is submitted in fulfilment of the requirements
for the degree of Master of Science in Game Technology
In the programme
Professional Master Game Technology
Academy for AI, Games and Media
Breda University of Applied Sciences
Date
25/06/2023

1
Declaration of Authorship
I, Jef Bernaers declare that this thesis titled, “Impact of Concept Art on the Character
Creation Pipeline” and the work presented are my own. I confirm that:
● This work was done wholly or mainly while in candidature for a research degree at this
University.
● Where any part of this thesis has previously been submitted for a degree or any other
qualification at this University or any other institution, this has been clearly stated.
● Where I have consulted the published work of others, this is always clearly attributed.
● Where I have quoted from the work of others, the source is always given. With the
exception of such quotations, this thesis is entirely my work.
● I have acknowledged all main sources of help.
● Where the thesis is based on work done by myself jointly with others, I have made clear
exactly what was done by others and what I have contributed myself.
Signed:
Date: 25/06/2023

2
Empty page

3
Breda University of Applied Sciences
Abstract
Academy for AI, Games and Media
Master of Game Technology
Influence of Concept Art on the Character Creation Pipeline
By Jef Bernaers
This thesis focuses on the character creation pipeline in the game industry. It aims to
identify the impact of concept art completion on the efficiency and quality of character
artists and provide suggestions for improving the pipeline. To narrow down and define
the concept art stages, examples and descriptions are created and used.
By means of a practical test with character creation students, cross-referenced
with data from interviews with artists and managers, conclusions were drawn. The data
shows that industry professionals cannot distinguish sculpts created with a low amount
of concept art from sculpts created with a high amount of concept art. This indicates that
the stage of concept art reached does not impact the quality and efficiency of character
artists in the first two hours of their sculpting process.
There is a disconnect between current industry pipelines in the games industry
and the results of this research. With the data gathered, suggestions for alternate, less
linear character creation pipelines can be created.
There is significant potential for future research, such as confirming (or
disputing) the findings with a larger sample size and looking into a correlation between
the definition of usability and the company size.

4
Empty page

5
Acknowledgements
I would like to express my deepest gratitude to my supervisors, Thomas Buijtenweg and
Dries Deryckere, for their time, wisdom, and insights.
I would also like to thank the incredible people advising me on this project. A special
thanks to Dries Deryckere, Linda Effinger, and Jits Habekotté for contributing so much of
their knowledge and time.
Additionally, I would like to thank the people who contributed to this research by either
partaking in the practical test or sitting down for an interview. Their contributions were
incredibly valuable.
Finally, I would like to point out the incredible support of my family. They supported me
throughout the entire process, and I cannot be more grateful for them.

6
Empty page

7
Contents
Declaration of Authorship 1
Abstract 3
Acknowledgements 5
Contents 7
List of Figures 10
List of Abbreviations 12
Chapter 1: Introduction 14
Chapter 2: Literature review 14
2.1 Definition and clarification of the character creation pipeline 15
2.2 Time Management Methodology in Games 17
Chapter 3: Methodology 19
3.1 Gather Concept Art 20
3.2 Quasi-Experiment 20
3.3 Semi-Structured interviews 22
3.4 Analyse results 23
3.5 Limitations 23
3.6 Ethical considerations 24
Chapter 4: Data and Results 24
4.1 Test Runs and Iterations 24
4.2 Final Results 26
4.2.1. Quasi-Experiment 26
4.2.2 Interviews 27
4.3.3 Q-Sort 28
Chapter 5: Discussion of Data and Results 30
5.4 Data and Hypotheses 30
5.5 Data and research questions 31
5.6 Methodology reflection 32
5.6.1 Quasi-Experiment 32
5.6.2 Semi-Structured Interview 33
5.6.3 Ethical Considerations 33
5.6.4 Future Directions 33
Chapter 6: Conclusion and Future Directions 34
6.1 Conclusion 34

8
6.2 Future Directions 34
Appendix A – Preliminary Interviews 37
Appendix B - Interviews 37
Appendix B.1: Interview Charles Zembillas 37
Appendix B.2: Interview Rafael Natali 46
Appendix B.3: Interview Jitske Habekotté 50
Appendix B.4: Interview Radu Ianescu 57
Appendix B.5 Interview Lothar Zhou 64
Appendix C – Raw Data 70
Appendix C.1 Questionnaire Questions 70
Appendix C.2 Interview Guide 71
Appendix C.3 Questionnaire Data 73
Appendix C.4 Q-Sort Data 74
Appendix C.5 Used Concept Art 77
Appendix C.6 Quasi Experiment results 79
Bibliography 81

9
Empty page

10
List of Figures
1. Chapter 2
2.1 Example of a character creation pipeline
2.2 Stages of concept art in order of creation, created by Alyssa Herman
2. Chapter 3
2.1. Methodology plan
2.2. Q-Sorting Roster
3. Chapter 4
3.1. Artefact of the quasi-experiment
3.2. Artefact of the quasi-experiment
3.3. Artefact of the quasi-experiment
3.4. Artefact of the quasi-experiment
3.5. Graph describing mentions of proportions in the questionnaire
3.6. Graph describing mentions of angles in the questionnaire
3.7. Graph describing mentions of the amount of concept art in the questionnaire
3.8. Graph describing mentions of readability in the questionnaire
3.9. Graph comparing concept readability to participant efficiency
3.10. Box Plot containing the results of the Q-Sort
3.11. Graph comparing concept readability to participant efficiency and Q-Sort results

11
empty page

12
List of Abbreviations
BUAS: Breda University of Applied Sciences
AAA: An informal classification used to categorise video games produced and
distributed by a mid-sized or major publisher, which typically have higher development
and marketing budgets than other tiers of games.
Indie: refers to games or studios created by independent developers

13
Empty page

14
Chapter 1: Introduction
Time management in the games industry has been a subject of discussion for quite some
time now. The volatile nature of the game development space forces companies to
constantly adapt and makes proper time management difficult. This is also true for
specific parts of the game development pipeline, like character creation. A well-designed
character is invaluable for most games, players relate to their virtual avatars. Traits like
personality, backstory, appearance, flaws, and weaknesses contribute to winning gamers’
hearts (room8studio, 2022). These are not only traits important to character
development, the visual design of a character, when handled with care, can also reflect
these attributes.
This paper aims to evaluate the impact of time spent on concept art on the quality
and efficiency of character artists. Preliminary interviews indicate that there are
significant differences in how different companies perceive the importance of concept art
(see Appendix A). The impact of the individual steps of the pipeline, current industry
methodologies, and practical data were investigated and compared to the insights of
industry experts. This way the impact of concept art on the character creation pipeline
was measured.
This research aims to raise awareness towards having substantiated reasoning
behind game development processes. The focus was on stylized 3D characters developed
for high-end console and PC games. The impact of time spent on concept art on the quality
and efficiency of character artists was measured specifically on the sculpting stage, which
leaves retopology, unwrapping, texturing, rigging, animation, and postprocessing out of
the research. These processes were left out because they go beyond the scope of the
research.
A research question and hypothesis were formulated accordingly.
Q: To what extent does the completion of concept art impact the efficiency of character
artists and the quality of their work during the sculpting stage of the stylised character
creation pipeline for video games?
H: Saving too much time on concept art leads to character artists having to spend more
time on their sculpts due to a lack of information. At the same time this reduces the
quality of their work, therefore leading to a longer and less efficient pipeline.
Chapter 2: Literature review
Literature on time management in the character creation pipeline is scarce. Therefore,
this literature review also contains information gained from preliminary interviews done
to gain a better grasp of the industry’s current knowledge surrounding this subject.
Parkinson’s law states: “Work expands so as to fill the time available for its
completion.” (Parkinson, 1958). It was first conceived of by C. Northcote Parkinson in a
1955 article in The Economist. With this statement, Parkinson describes how a person
always finds a way to fill their planned hours with work, even if the work itself occupies

15
less time. This is no different in game projects, the less time you give yourself on a project,
the less time you can procrastinate. This paper aims to improve time management in the
character creation pipeline. This is done by clarifying the relation between time spent on
concept art and the quality and time required for sculpting video game characters.
2.1 Definition and clarification of the character creation pipeline
The first step is to define a character creation pipeline and the individual steps it consists
of. “Much like a production line, the game development pipeline helps organise the flow
of work so that everyone knows what they need to deliver and when.” (Stefyn, 2022). The
same is true for the character creation pipeline. A pipeline is typically tailored towards
the expected input and output. a team of 30 artists may not use the same pipeline as a
team of two. Depending on the results you want, certain steps in your pipeline become
increasingly important while others might not even be required anymore. The dynamic
nature of character pipelines makes it hard to create a universal pipeline that applies to
every character. Not all steps of the pipeline will be discussed, concept art and sculpting
were selected. This excludes unwrapping, retopology, texturing, rigging, animation, and
post-processing. By excluding these parts of the pipeline, the scope of this research
narrows. This allows looking at the researched processes in more detail. Figure 2.1 shows
an example of how a character creation pipeline might look like.
Figure 2.1: Example of a character creation pipeline
“Concept art exists to save time, anything else is not acceptable and destroys the integral
meaning of the discipline.” (Rässa, 2018) “As a concept artist your goal, via exploration
and elimination, is to craft a singular, coherent vision.” (Richards, 2009). Concept art in
the context of characters is not just about creating a good-looking image. “Concept art
should emphasise clarity and functionality of assets.” (Singh, 2017). Concept artists need
to keep functionality, scale, animation restrictions, and many more things in mind while
creating a concept. In previous research, a sketch was identified as the earliest usable step
when creating concept art (Habekotté, 2019), respondents to their research indicated 30
minutes as the most picked option for the time required to make a character sketch. In
their research, Habekotté identified the stages of concept art in order of creation as:
sketch, value sketch, initial photobash, rough shading, edited photobash, and final mood
concept. For this research, these stages were reiterated to better fit specifically to the
character creation pipeline. The stages of concept art used in this research are in order or
creation: sketch, value sketch, colour study, character turnaround and final render (see
Figure 2.2). This research focuses on finding the point in the concept art pipeline where
the discipline starts losing its time-saving value. Let us elaborate on the specific steps of
concept art used in this research.

16
Figure 2.2: Stages of concept art in order of creation, created by Alyssa Herman
Sketching traditionally refers to a preliminary rough type of drawing that an artist
might make in preparation for their work, often in the form of lines. It is typically executed
rapidly, with little concern for accuracy. Sketching is about capturing a mood or key
feature of the subject. (visual-arts-cork, 2019)
A value sketch is a simplified drawing that focuses on rendering a simplified
version of the light and dark masses in a subject and how they relate to each other (Benve,
2022).
“A colour study is usually a small painting or drawing which focuses almost
entirely on colour arrangements. The goal of a colour study is not to create a beautiful
painting, but rather to explore and test different arrangements of colour. It is often done
as a lead-up to a more serious painting.” (Scott, 2018)
“Character Sheets, also referred to as character turnarounds, are a way for an
artist or animator to gain a feel for the character they are designing by sketching out
different positions and views of the body.” (Bruce, 2017)
A final (mood) concept is a grayscale or colour image that has the highest level of
detail. This image can also contain special lighting, visual effects, and any other additions
to create a visually pleasing image (Habekotté, 2019).
When a concept art piece is finished, it is passed on to a character artist who will
start sculpting the concept. Digital sculpting is done by using software to sculpt 3D objects
in a similar way to sculpting clay in real life, allowing artists to create organic shapes with
great levels of detail. It is done using a wide range of programmes, such as Zbrush
(Pixologic, 1999) and Blender (Roosendaal, 2002) (McKenzie, 2022). In the sculpting
phase, the character artist will create the character model in its highest level of detail, to
match as closely to the concept as possible. The result of the sculpting phase is a high-
resolution 3D model. Because of the high resolution of the model, the file is also large,
making it unsuitable for use. Because of this, a low-resolution version of the same model
must be created. This version will often be referred to as the low poly model, short for
low polygon model, referring to the model having a significantly lower number of
polygons than the sculpt, which is often referred to as the high poly model.
To summarise; the steps in a character pipeline are, in order: concept art,
sculpting or modelling, retopology, unwrapping, texturing, rigging, animation, and post-
processing. For this paper we exclude retopology, unwrapping, texturing, rigging,
animation, and post-processing. These processes were excluded to create a more realistic
scope, allowing for more accurate conclusions.

17
2.2 Time Management Methodology in Games
When talking about optimising the production pipeline for video games, it is
important to also understand the bigger picture. How is time generally managed during
video game production? What methods are used to do this? A broader understanding of
time management principles can help comprehend decisions made in the video game
production pipeline. We briefly highlight two common time management methodologies
used in game development; Waterfall planning and Agile planning. Waterfall is a more
traditional development methodology; however, Waterfall has slowly been replaced with
Agile workflows by game developers around the world for several years now. (Waterfall
VS Agile workflows, 2019)
“The Waterfall methodology follows a chronological process and works based on
fixed dates, requirements, and outcomes.” It consists of five stages: requirements, design,
implementation, verification or testing, and deployment and maintenance. One stage will
usually not begin until the previous stage has been finished. Waterfall depends on all
project requirements being known in advance (Gallagher et al., 2019; Herawati et al.,
2021; Adobe Communications Team, 2022). Schreier (2017) mentioned in their book that
this is not always true for game development. Multiple variables can change during a
project which would make Waterfall less than ideal. “It is typically considered an outdated
method.” (Humayun, 2021).
Over the last years, companies have started using Agile methods over traditional
development methods. According to a 2021 survey done by Digital.ai with over 4000
respondents, 86% of software development teams have adopted Agile methodology. The
most common reasons for adopting Agile within the surveyed teams was to enhance the
ability of managing changing priorities and to accelerate software delivery (15th state of
agile report, 2021). Rather than a chronological approach, like the Waterfall model, Agile
planning is a project management style with an iterative approach. Instead of following
an in-depth plan from the start of the project, Agile planning leaves room for requirement
changes throughout and relies on constant feedback. Agile projects start by identifying all
the product features and storing them in what is called the project backlog. In every
iteration, the team will choose features from the backlog to develop and then demonstrate
them in an improved version of the game (Keith, 2010). Agile focuses on the frequent
delivery of working software. However, Agile projects can quickly become chaotic if not
managed well or if the team does not engage in the practices. Documentation can quickly
become out of date, making it harder for new team members to get up to speed.
Additionally, the scope can forever increase with no defined measurement of success
when a clear end-goal is lacking. Moreover, a significant amount of time is spent in scrums,
retrospectives, sprint plannings, and stakeholder meetings, which can strain both time
and financial resources. (Gallagher et al., 2019).
Even though Agile planning is currently the dominant development methodology,
other sources suggest that depending on your development stage, a hybrid of Agile and
Waterfall methodologies might be more appropriate (Gallagher et al., 2019; Al-azawi et
al., 2014). Certain phases of development benefit more from a linear, pre-determined,
approach, while other phases will yield better results using an iterative workflow. What
method works best is entirely dependent on the project itself. If you can define all the
requirements and challenges in advance of a project with certainty, the Waterfall method

18
will work fine. But, as mentioned by Schreier, this is difficult for game projects. “Agile is
well prepared to handle the inconsistency and difficulty during the game development
projects participation.” (Humayun, 2021).
2.3 Interviews and Conclusions
Because the written literature on the subject is scarce, preliminary interviews with
industry professionals were conducted to get a better understanding of the current status
quo in industry knowledge regarding this subject.
The six preliminary interviews were conducted with people working in various
functions. This included a technical/pipeline artist named Andy Davis, three concept
artists named Charles Zembillas, Nicholas Cole, and Lothar Zhao, and two character artists
named Kevin Bayliss and Thijs de Rijcke.
Interviews with industry professionals point in three directions. Firstly, there
seem to be a lot of miscommunications between management and artists about time
management decisions. Interviewed artists often felt that the people responsible for
making decisions about the pipeline and time management have too little knowledge
about the process itself and therefore make poor decisions (see Appendix A).
Secondly, interviewees indicated that there can be no one general pipeline for all
game characters. Andy Davis, mentioned in a preliminary interview that a pipeline is
completely determined by the input and output it is based on (see Appendix A). This
means that character pipelines are influenced by many variables such as the art style, type
of character, size of the team, etc. Therefore, by specifying the style and platform of the
character and focusing solely on the relation between two steps in the pipeline, we can
more accurately make conclusions about the pipeline. Following this logic, this paper is
aimed specifically towards the relation of time spent on concept art versus the quality and
efficiency of character artists during the sculpting stage. Herein it will focus solely on 3D
stylized video game characters developed for high-end console or PC games.
‘Stylized’ is a broadly used term, it is often used to describe unrealistic aesthetics
in games. For this paper, stylised is defined as a cartoonish style. The Cambridge
dictionary (1995) defines cartoonish as simple, exaggerated and more colourful than its
real-life equivalent. Cartoonish characters will often have larger eyes, larger heads,
brighter colours, and exaggerated body proportions compared to realistic styles. For this
paper, this is the definition we will refer to when we use the words cartoonish or stylized.
The last thing that came to light in the interviews was how certain steps in the
pipeline differ between game studios and artists. The most mentioned step was concept
art. Some artists indicated that too much time is being spent on concept art, while others
said that more time should be spent. In preliminary interviews, Nicholas Cole mentioned
that a longer, more expansive pipeline with more people involved sometimes has adverse
effects on the quality of the final product.
It is important to aim this research at more than just artists. Pipeline and
management decisions are often done by technical artists, art directors and management
functions. As explained by Davis (2021), it is important for these functions to understand
the workflow they make decisions about. Davis mentions both on his blog (Davis, 2021)
and in a preliminary interview that by working with a pipeline, it can constantly be

19
improved through the process of iteration. A pipeline is not a set document, it is ever-
evolving depending on the changing companies' requirements and needs.
In conclusion, by critically evaluating the efficiency of commonly used pipelines,
this paper aims to improve time management in the character creation pipeline. Five
stages of concept art were defined for which to test the impact on the sculpting process.
The subsequent parts of the pipeline were excluded to create a more realistic scope,
allowing for more accurate conclusions. A broader perspective on time management in
video game development was explored by investigating common time management
methodologies used for game production. Preliminary interviews indicate that artists
might disagree with management decisions. Additionally, it is mentioned that there can
be no ‘one’ general pipeline for all productions. Pipelines are typically tailored toward the
expected input and output of the product. Lastly, there appear to be differences in how
time spent on concept art is valued between companies.
This leads to this paper aiming to effectively research the impact of individual
steps in the character creation pipeline. To achieve this, it narrows down the focus to
specific steps of the pipeline. These steps are applied to a specific style, type, and medium
in which the character is created. By examining initial planning methods, a better
understanding of potential problems in the character creation pipeline can be developed.
Additionally, emphasizing the importance of individual steps in the pipeline contributes
to this improved understanding. Highlighting the impact of these decisions might also
lead to more substantiated and evidence-based decision-making in artistic video game
pipelines.
Chapter 3: Methodology
Research question:
To what extent does the completion of concept art impact the efficiency of character artists
and the quality of their work during the sculpting stage of the stylised 3D character creation
pipeline for video games?
The methodology applied in this research consisted of three steps: A quasi-experiment,
semi-structured interviews with industry professionals, and evaluation and conclusions.
After the third and last step, an answer could be formulated to the research question.
Every step itself consisted of another three steps: preparation, execution, and evaluation.

20
The advantage of this linear methodology was that it was very easily scaled. From
the practical test on, sufficient data was instantly acquired to support research and
conclusions. The practical test was adaptable to fit more participants or less by changing
the amount researched art principles or the amount of used concept art. The interview
round served two purposes. Firstly, it removed the internal bias of the researcher by
reviewing the data from the practical test together with externals. Secondly, it revealed a
broader view on the subject from the eyes of industry professionals.
Figure 3.1: Overview of the methodology
3.1 Gather Concept Art
Concept art completion was measured in the reached stage rather than time spent.
This was done because the speed at which artists work is personal and can highly differ,
even between the most experienced professionals. Therefore, measuring completion
using the reached stage allows for more universally applicable results. For this first step
two pieces of similar finished concept art were acquired. These were found on Artstation
(Epic Games, 2014). It was important to ensure that the concept art would be unknown
to the participants. Therefore, used concept art consisted of cancelled concepts or fan art.
Concept art was selected according to how closely the style matches the described style
of the research. Additionally, it was judged by two active concept artists deciding whether
the concept art could be described as industry standard. Having multiple pieces ensured
the conclusions drawn were tied to the completion of concept art in general and not the
completion of a particular piece of concept art. A concept piece with intermediate stages
reached before the final render was necessary for this test. These stages were selected
according to data from the literature study. All artists of used art were notified in advance
that their art would be used for this research (see Appendix C.5).
3.2 Quasi-Experiment
The Quasi-Experiment aimed to test the impact of concept art completion on the first two
hours of sculpting. A group of character art students was tasked to work on the same
concept in different stages of its development. This way it was possible to draw a
connection between the reached stage of the concept art they were given and the quality
of the sculpts each character artist delivered.

21
The preparation for this test was done by creating a detailed project brief. This
ensured that everyone was working with the same amount of information. The students
all got the same reference board, only the concept art was different among them. Five
stages of concept art were handed out: a sketch, a value sketch, a coloured sketch, a
character sheet, and a final render. As seen in Figure 3.2, each stage of concept art also
included the previous levels. A character artist that got colour sketches also would have
had value sketches and sketches. For every stage, at least 4 persons had to do the test,
this ensured that the data was less likely to be influenced by individual skill or speed of
the artists.
The test was designed to allow for different numbers of participants. If fewer
participants than expected showed up, the number of variables could be decreased. They
could be increased if more showed up. Variables that could be increased or decreased
included the number of different pieces of concept art and the number of disciplines
tested. In this case, there were two pieces of concept art and its impact was only tested on
the sculpting stage to make the acquired data as reliable as possible.
At the start of the test a project brief was given, shortly summarizing the goal of
the assignment. The test took two hours, at which point the characters reached a medium
level of detail. A test of more than 2 hours could have led to a decrease in sample size as
the participants are unpaid and have little to gain from the test. After creating groups,
every character artist got a reference sheet that was the same for everyone, and a set of
concept art randomly assigned to the artists, ranging from only a sketch to a final render.
The reference sheet was there to mimic an industry environment. After two hours the
artists were asked to deliver their work by making screenshots and uploading them to a
folder. A frontal, side, and a ¾ perspective frontal shot were asked.
After the test, the participants were asked to fill in a brief survey where they could
describe if their experiences during the test. The questionnaire allowed for more
information to be gathered about the individual experiences of the test subjects. This
provided a better understanding of the results, avoiding a situation of only isolated
Figure 3.2: Overview of the quasi experiment

22
outcomes without any context. This data served to interpret the results regardless of the
different skillsets of the participants. For example, two participants might have had
similar results but one of them might have indicated that they found their efficiency and
quality to be lower than usual, which could have indicated a problem otherwise gone
unnoticed. The questionnaire questions can be found in the appendix (see Appendix C.1).
3.3 Semi-Structured interviews
The second part of the experiment entailed semi-structured interviews with industry
professionals. The interviews were divided in two parts; a semi-structured interview, and
an analysis of the sculpts created in the first test, in the form of a Q-sort (van Exel, 2005).
The pool of interviewees consisted of management functions, art leads and artists. Both
character and concept artists were allowed to partake as the research concerns both
principles of art. The variety of functions is important to see whether the people who
make the decisions about art share the same opinion as the people who make it.
Management functions only had one criterion for selection: their job had to include
working with the art department. Art leads and artists were first screened on their
portfolio websites, ensuring they worked with characters and had experience with
stylised art styles. Participants for this research were contacted through email, LinkedIn
(Microsoft corporation, 2003) or Artstation (Epic Games, 2014). Because most
interviewee participants were from other countries, the interviews and Q-sort were done
online.
Interviews were done first to get the opinion of current industry professionals
before they had been influenced by data from the practical test. The advantage of a semi-
structured interview was that it allowed following up on interesting topics or statements
the participant made. Three types of questions were asked to the interviewees in the
following order: warm-up questions, substantive questions, and closing questions. Warm-
up questions introduced the interviewees to the subject but mostly served to get the
conversation started in a natural way. Substantive questions dove deeper into the subject
discussed. They went over the interviewee's opinions and personal experiences regarding
the subject. Lastly, closing questions were meant to end the interview, they were lighter,
easier, and allowed the interviewee to get the last things off their mind while also
gathering some demographic data (Lankoski, 2015). The interview questions can be
found in Appendix C.2.
The second part of the interview was an analysis of the sculpts created in the
practical test. The analysis of these sculpts removed the internal bias of the researcher,
this allowed for more objective conclusions. The goal of the analysis was to see if the
interviewees could identify the difference in quality between artworks made with later
stages of concept art from those created with earlier stages.
The analysis was done using Q-sort methodology. Q-sorting allows subjective data
to be turned into more objective data (van Exel, 2005). The interviewees got images from
Figure 3.3: Q-Sorting roster

23
the practical test and had to rate them on a bell curve-shaped roster as seen in Figure 3.3.
The exercise was done four times, each based on different criteria. These criteria were:
silhouette, proportions, completion, and concept resemblance. Q-Sorting is ideal with
smaller sample sizes and by doing it live the researcher can pay attention to details. For
example, the researcher can spot if certain pictures take more time for the interviewees
to judge than others. An online Q-sort was set up by using a Miro board (Khusid, 2011).
This way interviewees were able to do the Q-sort live in a video chat with their actions
displayed in real-time on the board.
The interviews were then transcribed using Artificial Intelligence, specifically a
smart transcription software called otter.ai (Liang, 2016). Intelligent verbatim
transcription (Leano, 2023) was chosen as the transcription method to make the reading
of the transcriptions easier.
3.4 Analyse results
After the practical test and the interviews, all the gathered data was analysed. This was
done using thematic analysis. The goal of thematic analysis is to recognise recurring
patterns and themes to make generalised conclusions about them (Braun et. Al, 2013).
This analysis method might also yield additional data that falls outside of the scope of the
research question. Some of this data has been included in the paper as it is relevant
towards the research question. In the case of this research, to answer the research
question only a connection between the quality of the sculpts created in the quasi
experiment and the stage of concept art used to create those sculpts was required.
Thematic analysis allowed us to find such a connection by identifying both patterns in the
interview answers and the results of the practical test.
3.5 Limitations
Including a reference sheet impacts the data of the test as it might negate some of the
potential shortcomings of the concept art. However, it is an industry practice to have a
reference sheet while working and the test aims to mimic an industry environment as
closely as possible.
As this is a master’s thesis, no budget was provided. This limited the quality of the
practical test. Due to the lack of a budget, the test was done with students instead of
industry professionals. Ideally, the practical test and data analysis would be recreated on
a larger scale. This way, a larger sample consisting of industry professionals working for
a longer amount of time could be used as participants. This would further improve the
research by either confirming or denying the findings with new, more accurate data. On
the other hand, students may have state of the art knowledge.
Time is another limitation to this research. Given more time to conduct this
research, the influence of concept art could also be tested on other steps of the pipeline
and in other creative industries.
The results of this research were influenced by a culture bias occurring in the
sampling of the interviewees. All participants were part of western society countries and
most of them were white males. While this is not a bad thing, it might influence the way
they perceive art. The priorities in game development pipelines might be different across

24
cultures. Because of this, the applicability of the results can only be ensured for
development aimed towards such countries.
The last limitation to the research are NDA’s, short for non-disclosure agreements.
Because the interviews were done with artists often under NDA, not all questions could
be asked. This heavily limited the knowledge that artists bound by NDA’s could share
during interviews. While the interviews were constructed to avoid NDA problems, ideally,
they should not have been.
3.6 Ethical considerations
Because the research included interviews with industry professionals, NDA was a large
concern. To avoid violating NDA or artists’ privacy two measures were taken. Firstly, the
interview questions were constructed in a way as to not require interviewees to break
NDA to answer the questions. Secondly, all interviewees were given the option to be
anonymised. In this case, their name was not included in the research and any
recognisable traits are censored from the transcripts. These two measures fully protect
artists against accidentally violating NDA’s. However, none of the interviewees wished to
be anonymised.
Another potential ethical issue was the very nature of the research. This research
tried to identify at what point concept art loses its time saving value. Therefore, certain
conclusions could potentially lead to a loss of jobs. Any conclusions or statements had to
be made with the utmost care to not lead to an unsupported termination of workforce.
The last thing to address was the cultural bias of the interviewee sample. Most of
the interviewee pool consisted of white males from western society. While this is not a
problem perse, it did affect the data gathered from them. Various cultures have distinct
views on what is most important in game art. Therefore, only including western
participants heavily affected the outcome of the research. The research was therefore
mostly pointed towards a western audience, as results might not match with other
cultures.
Chapter 4: Data and Results
4.1 Test Runs and Iterations
Before the actual data gathering, tests were done with a smaller sample size. This ensured
that the data gathered from the final experiment would be as reliable as possible.
4.1.1 Test Run of the Quasi Experiment
The first part of the data gathering was the practical test or quasi-experiment with
students. Before doing it with a large sample, the experiment was first tested with a small
sample of three students. The students each got the same piece of concept art, but in a
different stage of completion. One student got value sketches, one had colour sketches,
and the last one got a final render.
Three things stood out after conducting the test. The initial observation during the
test run was that students achieved varying results. Interestingly, this variation was not
solely due to the amount of concept art they received, but rather related to their individual
skill levels. The second observation was that the answers of the questionnaire were not

25
what was expected. Lastly the third observation was that some of the students focussed
very much on one part of the concept and lost track of the big picture. This resulted in
work that could not be compared to each other.
These issues were solved as follows. The questionnaire was supposed to be a
solution to large differences in quality of the artefacts, therefore, lacking results indicated
that the questionnaire had to be expanded further. The questionnaire was expanded and
potential answers to the questions were tested using an AI chatbot called ChatGPT
(OpenAI, 2022). The AI chatbot allowed for quick tests to see what answers to the
questions could look like. Additionally, an expanded questionnaire allowed for more
information to be gathered about the individual experiences of the test subjects. This
provided a better understanding of the results, avoiding a situation of only isolated
outcomes without any context. Making the questionnaire more comprehensive would
solve the first and second problems encountered during the test run. The third problem
was solved by adding a few sentences to the project brief, instructing the participants to
make sure to focus on finishing as many parts of the concept art as possible. Additionally,
criteria were set up by the researcher to filter the results of the Quasi-Experiment. By only
including images of sculpts where most pieces of clothing were finished and that were
suitable for continued production, the images could be compared in a more accurate
manner.
4.1.2 Test Run of the Semi-Structured Interview
The interview questions were initially also tested using ChatGPT to generate answers to
the questions. The use of a chatbot helped better understand what kind of replies to
expect from the questions. The goal was to formulate the questions in a way that would
trigger conversation on the subjects relevant to the research. By not directly referring to
these subjects, leading questions could be avoided (Lankoski, 2015). Because the
interviews were used as data confirmation rather than data gathering, it was important
to keep them concise. This way, valuable time was not wasted on transcription that could
have been better used for additional data analysis. After reformulating the questions until
they resulted in the expected replies, the interview was tested with a lecturer from Breda
University of Applied Sciences. The interview yielded the expected results. Answers were
broad enough to give additional context but focussed enough to control the flow of the
interview.
The second part of the interview consisted of a Q-Sort (van Exel, 2005). The Q-
Sort went as expected, the only thing that was missing from the Q-Sort was context to
the results. To solve this issue, the Q-Sort was made more interactive. This meant that
the researcher would ask questions to reveal the reasoning for certain results of the
exercise. Questions would be asked on three occasions. Firstly, the researcher asked the
participants their reasoning for when results were placed in the outermost positions of
the roster. Secondly, whenever the interviewee would switch two images’ positions,
their motivation would be assessed. Lastly, if the interviewee had a tough time deciding
on a certain image, this would also be evaluated.

26
4.2 Final Results
With the methodology tried, tested, and adapted it was time to start on the actual data
gathering. This process consisted of a quasi-experiment, a questionnaire about the
students’ experiences, and semi-structured interviews with industry professionals
including a Q-Sorting exercise.
4.2.1. Quasi-Experiment
For the quasi-experiment, 19 students created character sculpts in two hours. These 19
students were spread over two pieces of concept art and then again over five stages of
completion withing these pieces as seen in Figure 3.2. Of the 19 students, 14 filled in the
questionnaire and 12 of their results passed the criteria to be used for the Q-Sort.
The results were filtered based on two criteria. Firstly, the images had to consist
of most pieces present in the concept art. Secondly, the sculpts had to be suitable for
continued production. The images below are results of the quasi-experiment. The two
images on the left did not pass the criteria. Figure 4.1 had different layers of clothing
created on one mesh, therefore making it unsuitable for continued production. Figure 4.2
did not create enough pieces of the concept to be passed on to the Q-Sorting exercise.
Figures 4.3 and 4.4 are examples of artefacts that passed the criteria to pass on to the Q-
Sorting exercise.
---------
After the experiment, all participants were given a questionnaire to fill in. The
questionnaire gave more context on the personal experiences of the students. Figures 4.5,
4.6, 4.7, and 4.8 show the mentions of certain subjects and whether these were positive
Figure 4. 1
Figure 4.2
Figure 4.3
Figure 4.4

27
or negative. In these graphs you can see that participants that got a character sheet or
final render had no negative statements on these subjects while students that got colour
sketches or less, had no positive statements on these subjects (see appendix C.1). T
Figure 4.9 shows how participants rated the readability of their concept compared to how
they rated their efficiency during the exercise based on previous similar experiences.
They rated themselves on a scale from one to seven. Readability steadily increases over
the course of concept art completion increasing. However, efficiency appears highly
irregular, the most plausible explanation for which could be the small sample size. This
could indicate that efficiency in the first two hours of sculpting might not be affected by
concept readability.
4.2.2 Interviews
After the experiment, interviews were held with industry professionals to compare the
gathered data to current industry trends. Additionally, by evaluating results from the
earlier gathered data with externals, the internal bias of the researcher is minimised.
Figure 4.5: Mentions of proportions
in the questionnaire
Figure 4.6: Mentions of angles in the
questionnaire
Figure 4.7: Mentions of amount of
concept art in the questionnaire
Figure 4.8: Mentions of readability in the
questionnaire
Figure 4.9: Comparison of concept readability and
character artist efficiency

28
Because the researcher was active as a game development artist, they automatically had
an internal bias when judging the artefacts of the quasi-experiment. By letting multiple
external experts judge the artefacts, the researcher did not influence the result with their
own bias. Five interviews were conducted. Three of them were done with concept artists,
one with an independent game developer, and the last with a company founder and Art
Director. Their years of experience ranged from 3 to 40 years. The interviews were
transcribed using an Artificial Intelligence transcription tool called otter.ai (Liang, 2016)
(see Appendix B).
In the evaluation of the transcripts, a few significant themes arose. Firstly, in
nearly every interview, a character sheet is mentioned either in a positive way or as
necessary. This is consistent with data from the practical test, indicating that character
sheets improve concept readability (see Figure 4.9).
Secondly, the amount of concept art required is often referred to as being
dependent on the team. Multiple interviewees indicated that the size of the team, the
company, and the individual preferences of team members influence the amount of
concept art that is required for continued production. Two of the three concept artists
mention how the less connected the team members are, the more important detailed
concept art becomes. Small teams understand each other’s vision and can discuss it with
one another easily. Large teams can be much more disconnected. They rely on concept art
to convey all information from the beginning, as direct communication between the teams
is more challenging.
Lastly, there are differences between how different artists believe a pipeline
should be constructed. One concept artist said that all the concept art must be finished
before you can continue in the pipeline, while another states that concept art does not
have to be finished to be delivered. The results of the experiment with students also
indicate that the blockout stage of a character sculpt can begin as soon as a design is
chosen, and a sketch is made. This can be done without losing efficiency or quality in the
character blockout. This is however countered by Ianescu (see Appendix B.4). While they
realised this could indeed be done, he preferred to avoid such practices. He mentioned
that he wanted to make sure not to put too much pressure on his employees, stating that
the first driving factor for pipeline decisions should be employee comfort over efficiency.
Another counterargument is the risk involved in starting before concept art is finished.
Natali mentions that it is harder to change a 3D model than a 2D drawing, therefore
implying that if a lack of concept art does lead to a mistake, it might eventually cost time
rather than save it (see Appendix B.2, 19:55).
4.3.3 Q-Sort
After the interviews, the participants were tasked with a sorting exercise based on the
results of the quasi-experiment. The interviewees were presented with images of the
experiment results and a bell-curve-shaped roster as seen in Figure 3.2. They would then
be tasked to sort the images from left to right, left being the worst and right being the best
images based on certain criteria. These criteria were: silhouette, proportions, concept
resemblance and completion. The results of the Q-Sort were then put into a data sheet for
a more global view of the data (see Appendix C.2).
Interviewees were not able to identify a difference in quality between sculpts
made with an early or later stage of concept art. The box plot below shows the general

29
results of the Q-Sort, every image was rated from one to five. There appears to be no
connection between the amount of concept art used to create the sculpts, and their
quality. The high variation between the results could be explained by the low sample size
of the project. A higher sample size would minimise the impact of individual artist skill on
the results. Furthermore, the Q-Sort results stay consistent across different interviewee
professions. Industry professionals are consistently not able to identify the works created
with high amounts of concept art. This further implies that individual artist skill has more
influence on their efficiency than the completion of concept art received for their work.
Figure 4.11 compares the Q-sort results to the questionnaire results. An
interesting observation about the graph is the V-Shape. It could imply one of two things.
On one hand, the data could suggest the following. Artists that get little information to
work with, like a sketch, will try to fill in the details themselves without necessarily losing
efficiency or quality. Once you start introducing more information like with value
sketches and colour sketches, the artist has more initial information to disseminate.
However, because the concept art is not in a final stage yet, some of this extra information
might be unclear. This might cause the artist to lose time by trying to understand
potentially unimportant details. Once you start adding even more detail, and push the
concept to a more final stage, such as with a character sheet or final render, this extra
information is much clearer. Therefore, the artist no longer needs to spend time figuring
out the details, as the concept artist has done it for them. Another explanation could be
the small sample size. Because of the small sample size of the practical test, it is possible
that faster working candidates were coincidentally placed at the outermost stages of
concept art, therefore resulting in this V-Shaped graph. The latter might be the more
plausible explanation.
One thing is clear, there appears to be no impactful connection between concept
art completion and character artists’ efficiency in the first two hour of the sculpting
process.
Figure 4.10: Q-Sort results

30
Chapter 5: Discussion of Data and Results
From the evaluation of the results above, interesting implications immediately
arise. A first observation is the difference in perceived quality between all results. There
appears to be no immediate connection between the quality of the sculpts and the quality
of concept art used to create them. This could imply that the individual skill of an artist
has more impact on the quality and speed of character artists than the completion of the
concept art they use. On the other hand, it could also imply that the completion of concept
art an artist needs is highly personal, some artists might work fine with just a sketch while
others might be very reliant on a character sheet.
Another interesting result is the connection between concept readability and the
work efficiency of the students. Figure 4.11 shows that there appears not to be any.
Concept readability increased whenever students got later stages of concept art.
However, their efficiency values do not differ much from each other. This might imply that
in the initial two hours of sculpting, concept readability has no impact on the quality and
efficiency of character artists. This is further confirmed by the results of the Q-Sort, again
showing a disconnect between the readability of concept art and character artist
efficiency and quality as seen in Figure 4.11.
The results of the Q-sort match the conclusions drawn from the experiment, that
there appears to be no observable connection between concept art completion and
character artists’ quality and efficiency in the blockout stage of the sculpting process.
5.4 Data and Hypotheses
The main thing that stands out in the data is the inconsistency of the Q-sorting results. If
industry professionals are not able to distinguish the quality of character blockout sculpts
created with a sketch from those created with a final render, then there would logically
be no reason to wait for the final render to start sculpting the character. This result is
Figure 4.11: Q-Sort results compared to
questionnaire results

31
consistent with previous research by Habekotté (2019), implying that a sketch is the
earliest usable step of concept art. Introducing character artists earlier in the pipeline
might also increase job satisfaction by granting more creative freedom to the artist. The
interview transcripts (see appendix B) imply that this is currently not the standard in the
industry. However, they also suggest that such practices could potentially increase stress
on employees. For example, by increasing artist responsibility on the final look.
Furthermore, collected data suggests that artists always prefer to have more concept art.
At no point did any of the artists feel they had too much concept art. Even if the concept
art is ultimately not used, character artists might still feel more secure with having more
of it. While there is a potential increase in efficiency, the risks of starting earlier might
outweigh the benefits, as changing a 3D model could take more time than changing a 2D
drawing.
Secondly, the interview transcripts (see appendix B) imply that the importance of
concept art depends on the scale of a company, the individual artists, and the context of a
project. Large-scale projects with large teams could need more concept art as
communication between the teams is harder, especially in the case of outsourcing.
Smaller teams might get away with less concept art as they are more aware of each other’s
work, and communication between team members is easier. In the same way, a very high,
strict amount of concept art ensures that character artists stick closer to the concept,
while a more loosely created concept allows for more creative freedom from the character
artist. More creative freedom could increase job satisfaction but might also cause the
artist to venture too far from the art style of the project. Both outcomes could have a
different impact on a project and should be kept in mind when reflecting on the pipeline.
Lastly, character sheets are consistently proven to be an important stage of
character concept art. Questionnaire data clearly shows high improvements in readability
from the character sheet stage on, which is consistent with interviewees who nearly
unanimously emphasized the importance of a character sheet for character concept art.
While a connection cannot be drawn between a character sheet and character artist
efficiency in their blockout stage, its impact on readability is clear in the data gathered.
Charles Zembillas even describes rotations as the foundation of character concept
development. (See Appendix B.1, 24:26).
Practically, this means that companies might want to investigate the effect that
time spent on concept art has on the efficiency and quality of their character creation
pipeline. In-company research by developers is the most suitable way to apply the
implications of this research to the game development industry.
5.5 Data and research questions
“To what extent does the completion of concept art impact the efficiency of character artists
and the quality of their work during the sculpting stage of the stylised character creation
pipeline for video games? “
Because of the small scale of the research, only the first two hours of the sculpting stage
were researched. However, enough data was gathered to make implications towards
possible answers. The answer to the research question is that it depends on the project,
team, and individual artists. Interview data implies that small teams might get away with

32
less concept art because they are more aware of their colleagues’ schedules and can
communicate easier. Teams in larger companies can be more distanced from one another,
therefore making communication between departments harder. Especially in the case of
companies that outsource their work, rigid concept art that relays more detail becomes
more important (see Appendix B.3; Appendix B.5). The outsourcing company might not
be completely aware of the style and context of their work. However, no connection could
be made between the completion of concept art and the efficiency and quality of character
artists using said concept art in the first two hours of the sculpting process. This implies
that, in theory, as soon as the concept is defined and an initial sketch has been drawn, a
character artist can start working on the blockout. In the interviews, concerns about such
practices were raised as they could increase stress on employees while also being risky
as 3D art might be harder to adapt than 2D art (Appendix B.2, 19:55). On the other hand,
there are mentions of how individual artists fill in missing details themselves. Extra
creative freedom for the character artist in their initial stages could prove beneficial to
the quality of the developed character. The argument could also be made that additional
creative freedom might increase job satisfaction for the character artists. Once again,
depending on the context one scenario might be more beneficial than another. When
working in a large company or outsourcing work, rigid concept art might ensure that the
developed character stays within the defined boundaries of the project. However, in
smaller companies, in-house artists might be more aware of the project and context. In
this case more creative freedom and starting the 3D process earlier could prove beneficial
to the character creation process.
5.6 Methodology reflection
As there is little previous research into character creation pipelines, there is a lack of
standardized methods for conducting research in this space. Therefore, a combination of
standardized methods was used to ensure all the goals of this research could be reached.
A critical evaluation of the used methodology allows for future research in this area to
improve their methods.
The focus of this study was to compare industry practices to practical data, testing
the validity of current industry practices, and preventing the dissemination of faulty
industry practices through this research. The combination of practical data and industry
experience allowed for the formulation of more coherent conclusions while also
suggesting potential improvements to current industry practices.
Due to the scale of this research, the samples were small. Ideally, the practical test
would be conducted on a larger scale with a consistent artist skill level. However,
achieving this without funding is challenging, which is important to keep in mind when
considering similar research.
5.6.1 Quasi-Experiment
The quasi-experiment was effective in developing artefacts suitable for this
research. Having practical artefacts that reflect actual professional performances in the
researched area greatly contributed to the quality of the data. The inclusion of a

33
questionnaire after the experiment was crucial for this research. As previously
established, there are significant variations in speed among individual artists. Therefore,
having more context behind the created artefacts allowed for better understanding of the
reasoning behind certain results. However, self-reporting is not the most reliable form of
data. The Q-sort and interview data were invaluable in ensuring accurate conclusions are
drawn from such data.
A notable adaptation was made by implementing criteria to determine which
artefacts would be used for the Q-sort. By establishing minimum criteria, the differences
between the quality of the artefacts were significantly reduced, thus improving the
reliability of the Q-sort data.
5.6.2 Semi-Structured Interview
The interviews proved to be effective at determining industry experiences and practices.
The Q-Sorting exercise provided worthwhile insights in how industry professionals
interpret the research data, while also mitigating the internal bias of the researcher.
AI tools were used to better formulate interview questions, which proved to be an
effective method for predicting potential answers to interview questions. This allowed for
the creation of highly effective interview questions, triggering the discussion of certain
topics without leading the interviewees towards conclusions.
It is important to note that the interview data was used for data confirmation
rather than as primary data. Keeping the interview concise and focused on the research
subject saved time while still allowing for the collection of necessary data when handled
appropriately.
5.6.3 Ethical Considerations
The conclusions of this research, when wrongly interpreted, could have negative societal
effects. This data should not lead to termination of workforce, as it does not suggest
concept artists work is not important, because an increase in readability is observed.
Alternatively, the data implies that there can be more and earlier cooperation between
concept artists and character artists. “Games reflect the lives of their collective creators”
(Potanin, 2010). By including more different artists in the initial process, a higher variety
in results can be achieved.
5.6.4 Future Directions
The methodology used in this study proved sufficient for answering the research question
and providing additional insights into the researched field. However, considerable
changes would have to be made were this method to be applied on a larger scale.
More research could be conducted to determine the most commonly used stages
of concept art. Additionally, the quasi-experiment could be conducted on a larger scale,
involving a larger sample working for a longer duration, while also using industry
professionals instead of students as participants.
The questionnaire could be expanded with questions on the participants’ own
industry experiences, such as their familiarity with the amount of concept art received
and whether it was less, or more than they normally work with.

34
Furthermore, increasing the interview sample to include individuals from diverse
genders and cultures would enhance the applicability of the study's results.
Lastly, the research could also be conducted in different media entertainment
fields, such as film or the animation industry.
Chapter 6: Conclusion and Future Directions
6.1 Conclusion
The goal of this research was to formulate an answer to the stated research question
concerning the impact of time spent on concept art on the sculpting stage of the character
creation pipeline.
The answer to the research question is that it depends on the context, company
size and the individual character artists. The data from this research suggests that as soon
as a sketch is defined, a character artist can already start blocking the character out
without any impact on their quality or efficiency. This finding is consistent with results
from previous research by Habekotté (2019), stating that the sketch is the first useable
step in concept art creation.
Interview data suggests that smaller companies could potentially get away with
less detailed concept art because the team is closer, and communication is smoother.
Larger companies, where departments are more distanced or outsourced, could benefit
from more defined, rigid concept art. This could ensure a unified vision across different
departments and teams. Less rigid concept art could allow for more creative freedom for
the artist. More creative freedom could potentially improve job satisfaction and therefore
increase productivity. On the other hand, it might cause artists to go in a different artistic
direction than required, a risk that is potentially higher when outsourcing.
These results are inconsistent with current industry practices. Interviewed
industry professionals stated that concept art must be finished before further
development can start. However, the results of the research suggest you can start as soon
as the concept is defined. This might indicate a lack of research behind current industry
practices.
Lastly, a character sheet was proven to improve concept readability and was
consistently defined as essential by interviewed industry experts. The data of this
research was gathered on a small scale, and we highly advise future studies to replicate
this with a larger sample size to validate the data.
These conclusions serve to present an initial view on the character creation
pipeline and should be treated as a notice to developers to reflect and consider the
consequences of unsubstantiated pipeline decisions and a foundation on which to base
additional research. They should in no way lead to the termination of workforce but
rather encourage more teamwork and variety within game development processes.
6.2 Future Directions
This thesis opens multiple new potential alleys for research. This research specifically
only focuses on the initial two hours of the character sculpting stage due to time and

35
resource constraints. Ideally, the same question would be raised in all steps of the
character creation pipeline subsequent to concept art to measure the influence of concept
art on texturing, rigging, animation, etc. This not only for characters but also for props
and environments. Similar research could also be conducted in other creative industries
such as, for example, film.
All the findings of this research should only be seen as implications as the scale is
too small for accurate conclusions. The data of this research implies quite a few
inconsistencies between current industry practices and results from the practical tests.
Further research of a larger scale could confirm or deny these implications and provide
reliable evidence on which industry practitioners can base their decisions. The research
also introduces the question of how less concept art gives the character artist more
freedom in the character creation pipeline. Additional research could be conducted to
investigate how this additional freedom influences the 3D character creation process and
job satisfaction for character artists.
Another interesting observation to keep in mind is the potential difference in the
importance of concept art depending on the scale of a company. Further research could
be done to determine exactly how the scale of a company influences the importance of
time spent on concept art.
As of the writing of this thesis, we are on the verge of a potentially massive change
to the game development space. The recent introduction of AI tools has left lots of artists
and game developers in awe of their potential. Tools such as DALL-E (OpenAI, 2021) or
Midjourney (Midjourney Inc., 2022) can generate highly detailed imagery in seconds. Five
years from the writing of this thesis, the game pipelines described in this research may
have been altered significantly. AI tools could dramatically increase the quality of the
initial stages of concept art, what is currently described as a sketch might soon be
converted to a high-quality render in a matter of seconds.
While this innovation in the industry changes the premise of this research, the
essence will remain. How does time spent on concept art influence the quality and
efficiency of character artists? Whether or not the concept art was made by an AI, the
question remains relevant. It is hard to say how this research will relate to the games
industry five years from now but staying critical and continuously evaluating the
efficiency of the methods developers use will always remain essential to the progress of
this and any industry.

36
Empty page

37
Appendix A – Preliminary Interviews
Preliminary interviews were done during the first stages of this research. Due to the
NDA sensitive nature of these interviews, they were not transcribed. The interview
recordings are available upon request and after signing an NDA.
Interviews were done with
Andy Davis, 30/09/2022
Lothar Zhao, 04/10/2022
Thijs De Rijcke, 10/10/2022
Charles Zembillas, 10/10/2022
Kevin Bayliss, 25/10/2022
Nicholas Kole, 26/10/2022
Appendix B - Interviews
Appendix B.1: Interview Charles Zembillas
Wed, Mar 22, 2023 10:02PM • 1:30:50
SUMMARY KEYWORDS
character, concept art, studio, people, artists, animation, concept, develop, production,
drawing, work, happening, project, design, problem, creative, big, doodle, personally,
decisions
SPEAKERS
Jef Bernaers, Charles Zembillas
Jef Bernaers 00:00
Okay, so you're ready to get started? I have to do this this level of formal thing at the start.
So I'm going to get that out of the way right away. First of all, thank you for the opportunity
to talk to you again it very much helps my research and it means a lot to me.
Charles 10:25
Certainly, you're more than welcome.
Jef Bernaers 10:28
That's good to hear. So my research has evolved quite a bit since the last time we spoke
because the last time we spoke I was still looking specifically at anthropomorphic
characters and I was looking at general pipeline improvements, but I have narrowed my
scope down quite a bit because of the scale of my research, as it is a masters research, I
don't have that much time. So I had to narrow the scope down quite a bit to make sure
that I could actually get decent results. Where I'm at right now is I'm specifically looking
at the impact of time spent on concept art and what impact it has on character artists in
the further stages of their pipeline. And for my research, I'm only looking at the sculpting
stage. And well, I'm just looking at if you save time on concept art, how does that affect

38
character artists down the line? Does it actually save time? Or do you end up spending
more time because your character artists don't get enough information afterwards?
That's like the short elevator pitch of what I'm researching right now.
Charles 11:44
All right, I would have to say it all depends really, on what the situation is. And like, is it
video games? Is it television? Is it feature films? Because for example, in feature films,
sometimes they spend years in concept development. And in television, for example, you
have to move very fast. You have to make decisions very quickly. You don't have the luxury
of just spending time going back and forth with the concept. And I think video games is
probably like right in between, you know, you need to move quickly, but your production
time isn't as long as it would be for feature films in general. But you don't have these super
tight deadlines like you do for television. And sometimes when you have too much time,
it can be detrimental. When I was working on projects that were related to feature films,
I found it frustrating because I wanted to have feedback quickly. I came mostly through
television, so I was used to working quickly to getting fast decisions. And it would help me
to make adjustments and go in directions that were meaningful, you know, but when I
was working with these long-term projects, it was kind of unnerving you know. It was
like, let's see, let's try this and we will have another meeting in two weeks or something.
And I just don't work that way. I like to keep moving. And I like to get input quickly and
make decisions and then we can go on to something else and then come back to a little bit
later on and just address any kind of issues that we might have. So I would say I think
that's one thing about working with video games I like because you have the flexibility of
time, extra time as opposed to TV, but they don't waste it like they do in motion pictures
because a lot of time is wasted. It really is and then it comes down to crunch time at the
end, you know, you have to put in all this overtime because the deadline is coming up. The
release date for the movie is coming up and you're like months behind. A lot of it has to
do with poor management and in animation, like in Hollywood for example, from my
experience and from talking with very high-level professionals that I know, whose
opinion I highly regard. The management is very poor. It's very, very poor. And it's simply
because these people are not creative individuals. They're not artists, they don't know.
They don't know, yet they sit there, and they make these decisions that are just very poor.
A lot of the creative decisions they make are stupid, just stupid. It doesn't necessarily have
anything to do with making the production more efficient or making let's say like the
characters more appealing and I can give you an example of that. I feel one of the greatest
films animated films that have been made in the in the modern era is a Klaus by what's
his name? Sergio Pablos was a studio artist; I think he was over at Disney. He went to Cal
Arts, the big animation School for the industry and he broke away. I'm not sure exactly
when he broke away, but his movie is a masterpiece. It's a masterpiece of design. It's a
masterpiece of storytelling, because he didn't have these people telling him to make it
stupid, or to dumb it down, just forcing him to make these outrageously bad decisions.
And another movie that I saw recently was Guillermo del Toro's, please excuse me
because I get names confused, but I saw Pinocchio. I thought Pinocchio was really very
unique. The look of it was unique because it appeared to me that it was very true to the
original story, much more so than like Walt Disney's Pinocchio, which I don't care what
anybody says. I still think that's one of the greatest films of all time. I'm a big, big fan of
Walt Disney's Pinocchio. But this one was really really different. And not only was

39
Pinocchio like a character carved out of wood. All the characters looked like they were
carved out of wood, and it was stop motion. How can human beings do this? I just don't
understand the patience and the effort that it took to do this. And so, you see these films
that are done with creative people in control, and how much better they are, in many ways
across the spectrum, than the corporate studio based stuff that can be really unnerving.
They're just really not as good as they could have been. They could have been really, really
great. Instead, it comes out to be good or really good, but it could have been really great.
If they would let people do their job. So, I hope I'm answering your question because I feel
like I'm kind of going off and on side topics or chant tangents. It can be problematic, in my
opinion, when you have too much time and you're not able to make quick decisions and
move things along. And do it in a way that makes sense and is responsible. So, yeah, how
was that?
Jef Bernaers 18:31
That was good. That was great. So before I ask you the next question, I'm going to quickly
mention, again, this was not a problem last time, but I'm just going to say, if at any point,
I don't know if you are under NDA, but if at any point you feel that you can't answer a
question, you can just tell me and say I can't answer this question due to NDA. And like I
said before, if you answer a question and you then realize I don't want this to be on the
record, you can also just tell me, and I'll note it down and remove it from all transcripts.
Charles 19:06
Okay. I'm not under any in NDA right now. So that's not a problem.
Jef Bernaers 19:13
Um, the interview itself is going to be 30 minutes, something like that of just some
questions about my research. And then after that, I'm going to send you I don't know if
you've used "miro" before.
Charles 21:21
I don't think I have.
Jef Bernaers 21:22
I'm going to send you a link to my report. It's some brainstorming slash teamwork
software. But on the Miro boards I've created a sorting exercise, with an exercise that I
did with students. Basically, I had a couple of students make some 3D art and I'm just
going to have you rate the 3D art based on the concepts and certain criteria. So, it's just
going to be a couple of images and then criteria like silhouettes and then just order them
from least good to best. It's going to be just a sorting exercise and then afterwards, I can
discuss why I'm doing it but I'm not going to do it before because I don't want to change
your opinions before we do it.
Charles 22:45
I don't want to hurt anybody's feelings. I don't want to beat on one's feelings. Because if I
choose one thing over another. My approach is always to look for the positive. Yeah, and
I don't really like being mean, that's why I'm a poor judge. I can do it, but I don't enjoy
doing it.

40
Jef Bernaers 23:18
Students are not going to see the results of the sort; they're just going to be disseminated
into data. So, the students are not going to know how you rated their work. So no problem
at all. Then I can get started on the questions?
Charles 23:49
Yes, go ahead.
Jef Bernaers 23:51
Alright, so the first question is something that you kind of mentioned already earlier about
how much time is spent on concept art and how it highly differs between different
companies or different media that you create them for? So when you work on concept art,
if you were to be in charge, how do you decide what is the optimal amount of time to
devote to a piece of concept art? What criteria do you use to decide: "This is how much
time I should be spending on it" or "this is done".
Charles 24:26
I look at the production schedule, and what the deadlines are, and that's a big part of it,
because you really want to be mindful of that you don't want to go beyond the deadline
because that causes lots of problems. So that's the first thing that I do. Yeah, and the
second thing that I do is think: "Is this realistic?". Can you realistically do this development
art in the time that's given. If I feel that you can, great, and if it's not then I just try to work
with whoever it is that I'm involved with, and see if there's any flexibility, and that's the
first thing if I'm doing the work. I work pretty fast, I like to give them lots of stuff at the
beginning and knowing that they're going to say no to about 90% of it. So I'd like to give
them a lot of options. And then once they start to focus on what they're partial to, then I
can start to home in on, coming up with a final design. It usually goes pretty quickly. If
there's any problems with it, it usually comes from their end because they're either having
committee meetings and are not able to come to a consensus. Or they're not sure which
way to go. And sometimes I just need more time. You know, sometimes it's not flowing,
maybe I have a creative block and sometimes I need a little bit of time to step away and
Figure out where to go with what I'm doing. So, I would say the biggest thing would be,
how much time do I have, and then I just kind of go from there. And then I usually beat
the schedule. You know, and then once they agree on something, I move it into defining
the character. Let's do the rotations for it. That's the foundation of it, it's the suited
rotations to make sure that we can see what the character looks like from at least a 180
degree turn and then start to explore the personality, do the facial expressions, the
attitude poses, action poses, maybe workout, some walk cycle, some animation, that kind
of thing. Just get it to the point where it can be handed over to the team. Okay, we've done
enough, at some point it's handed over to the animators who are usually very competent.
And then just take it from there, you know? The schedule is very important. You want to
be mindful of that, you want to be mindful of the budget. But on the flip side of that, you
want to have the time to really do good stuff. And if it's not there, and if you're under a lot
of pressure and you're dealing with like the poor managerial decisions, poor creative
decisions that I was referring to earlier, then morale is affected, and you don't come up
with things that are as good as they could have been.

41
Jef Bernaers 29:38
All right. So you mentioned collaboration with your team members and the points that
you that you have to give work to your team. When you work on concept art for a 3D
project. How do you collaborate with those team members? How are other team members
involved in your concept art progress?
Charles 30:22
Are you talking about the creative team? Or are you talking about like the managerial
team?
Jef Bernaers 30:26
Are you involved with creative team, with the 3D artists or is it mostly the managerial
team? What input do you get?
Charles 30:37
Okay, the creative team, what I mean by working with a creative team is that I might not
be the only one who's doing concepts for a specific aspect of the project. So in other words,
let's say you're doing a Christmas special, and you're going to be doing Santa’s helpers at
the at the North Pole, the toymakers or whatever. And there could be several people that
are doing concept designs for this particular part of the of the project. Normally we would
get to a point where we see what everybody's doing, and we can share that and you're
influenced by that. So you go: Oh, that's a great idea, like how you did the ears here, oh, I
like how you did their shoes and there's a back and forth with that. So you kind of become
inspired by each other. And it's healthy, it's a good thing, and it's not necessarily
competitive. I think there may be a competitive edge to it, but it isn't necessarily
competitive. It's more of a sharing, like you hang out with your friends and everybody
says, let's come up with new ideas for Jedi Knights and for Star Wars, and everybody does
his own thing, and then we just kind of look at it. That's always a lot of fun. And then you've
got the people who will follow up with what it is that you're doing, let's say the modelers,
and you want to be mindful of: is what you're doing realistic, will it function, can it be
animated? And that was a big part of Crash Bandicoot, when we were developing Crash
Bandicoot it was like, can that, can this work? You've got this this initial concept, which is
really very awkward. It was like a very awkward character, how can you make it work?
Can you make it animated? Crash Bandicoot looks great when he's smiling, but what
happens when he's upset? What happens when you do other things with his mouth?
Because his mouth is within his body. So you have to work out those kinds of things and
then later on, when the technology kept improving, there were other things that you had
to kind of keep in mind like, in the early PlayStation format, you had very low polygon
limits per character, so everything had to be very simple. But then later on, with the
advent of a PlayStation two girth and the thickness of things was an issue because you
had 20 times more resolution than you had with PlayStation One. So you adapt your
designs to what the technology you know, to max out the technology. And then you've got
the management team, as I said before, it could be a situation where these are corporate
people, they're not creative people, and they can come in and make a decision about how
characters should look and it'll set you back. It makes things more difficult because they
don't know what they're doing. It all depends on who you're working with. I personally

42
always design with production in mind. In other words, you know, I come up with a
character concept, I want to make it work, so I put effort into the rotations into exploring
the character, how the character would animate, how the character would act. And all of
that is very helpful as the art that you're doing goes down the production pipeline. And I
think that's priority number one, it would be come up with a with an appealing design.
Okay, and because I'm character centric, I'm talking about character. So you could do the
same thing with environments. You could do the same thing with vehicles you could do
the same thing with props, but I always look for appeal, design appeal. Then I move that
into functionality, can it function, would it be able to work, and then trying to work around
any kind of adjustments that you have to make that are a consequence of managerial
decisions. I think that's the best way to go.
Jef Bernaers 37:04
So, you mentioned that you have to keep the constraints of your character modelers and
the media that you're working for in mind. Can you describe a time that you've created
concept art that might have negatively impacted the quality or the efficiency of a character
artists further down the line,?
Charles 37:29
I can recall, definitely. I can recall back in the days when I was working in television, where
I had very poor art direction and the changes that we're making on the characters are just
not going to animate, you know? You're making it worse, and it was sensitive because I
really couldn't say anything because you're dealing with egos and hierarchy, all that kind
of stuff. Then you see the results when the footage comes back, and the project is in the
can, as we say, and they look terrible. They look terrible, and we weren't given enough
time to do appropriate models. We couldn't do the models. We could just maybe punch
out a rotation really quick, but we couldn't do the appropriate models showing how this
particular weird thing that the art director wanted me to do or the director or producer,
how we can make it work. And it showed. Since all this was shipped to a remote location,
it was your map. You're working in Burbank, but it's all being shipped to Korea. And the
Koreans couldn’t make heads or tails out of it and they're really under pressure. And it
just comes out looking very poor. It hurts the show, it was cancelled. It showed for one
season and it'd be cancelled. This studio was owned by somebody who didn't care, very
wealthy person whose parents were deep in Hollywood and it was terrible. You had big
name personalities doing the voices for these cartoons and it just looked like garbage. And
it was only because the head of the studio didn't care. And the people who I was getting
direction from were very incompetent, you know. So there's plenty of examples of that
and that personally led to a great deal of frustration with me. And it forced me say, Look,
I can't work like this, and it forced me to go and look for alternatives. That was difficult
because I didn't really want to do that. I wanted to do quality work, but I couldn't do
quality work and everything that was coming out was just looking like junk. I didn't want
my name to be associated with stuff that was so poorly done. So that's one example. Okay,
but in time I think things changed and the demand for quality really kind of took over in
time. And I think the technology that we have today also really helps. But in order to
survive today. You have to have quality but back then people didn't care. If you don't care
today, you're not going to be able to compete. I have lots of experience with where your

43
concept development/pre-production art adversely affects the production itself if it's not
handled properly. Yeah.
Jef Bernaers 44:23
So how did you did you find out specific ways or develop methods to circumvent this from
happening or to prevent these kinds of situations from arising?
Charles 44:36
I went my own way, I just went off and I did my own stuff. I did my own projects. I was
really kind of a renegade you know, I was a very outspoken critic in the animation
industry for years. I started a whole movement, I would organized walkouts of the
industry over. On April 1 of every year, we would walk out at lunchtime, the studios would
walk at lunchtime and not go back. Because the only animation union that I know of in the
world is in Los Angeles, and it's tied in with the studios and it's a whole different kind of
dynamic there that you don't have in other places. Yeah, I don't think you have that in
Canada. I don't think you have that in the UK, France. I just have a tremendous amount of
respect for French animators. There’s just tremendously talented artists there. Then you
have the Spaniards with Klaus as I mentioned before, because I think it was animated in
Madrid, but you have a different dynamic there. The Union at the time tended to blame
non-union artists for all their problems that they were having and the non-union slash
independent artists were saying we have nothing to do with what you guys are doing
there. It's not our fault that you guys have studios that are poorly managed. So I was trying
to unify artists saying, we're not the problem. The problem is the management, the
executives in the industry, we're working for people who are basically idiots, but they get
the financing and they match. You get people who started as office help and have no idea
of how to draw or how to write a script and they try to walk into the shoes of Walt Disney.
I spent many years making an issue out of this, and then the union turned on me you know.
I think one of the really great cartoons of the last few years was primal with Genndy
Tartakovski, there is a really great work that's being created but the people that are
working on it, I don't necessarily think that they're getting any equity from it. Meanwhile
the people who are working independently are finding a whole new way of doing things
and somebody who was like a director at a Disney movie, let's say 15 to 20 years ago, has
now created a personal franchise for himself. Everybody in the world knows who he is.
And I'm sure that he's doing extremely well, a lot better than what he'd be doing on a
union base salary and stuff. And this whole new world opened up and that is affecting
animation that is affecting concept art.
Jef Bernaers 51:03
So I asked before about what you do beforehand, when you start on concept art how you
determine how much time you would like to spend ideally on your concept art. What is
your decision making process for deciding that a concept art piece is ready, that you're
done?
Charles 51:38
Yeah, it's just kind of instinctive. Sometimes I'll do a little doodle. And I'll just say that's it.
I nailed it. That's it. It's instinctive and then it's modified. And the great example of that is
Spyro, the dragon. I did that character very quickly. I think we had him pretty much nailed

44
by the second day that I was working on it. I just knew it. But Spyro was all over the place,
Spyro was kind of crazy. He was like a big and massive dragon at one point, kind of mean
and grumpy at another point and that's what I do. They just wanted a dragon character
and I was just throwing all this stuff out. And then by the second day, we were kind of like
assessing the work that I had done. I just said, why don't we try something cute? Like a
cute little character, and they had their meetings. After them they said, we want to have a
character like a bratty eight-year-old. A little bratty kid, and I said okay. I just started
doing these little doodles and stuff like that. I just did one and I just said, That's it. I felt
very strongly about that. And I didn't really have to convince them of anything because
they could see it too. And then we just became a matter of, okay, this is the direction we're
going in. So I would do variations on that concept, that idea. So instead of throwing
everything out there, we had a path to go through and that just became a matter of refining
it. A lot of what I do too is I just ask people. I'll go to people in the classroom, I'll come up
and I asked them, which one do you like better? Which of these concepts do you like
better? And, and, you know, and that gives me a consensus. And sometimes I'll go against
that consensus. I'll say, I know why you guys are looking at this but I personally feel that
this is a better thing to do, a better direction to go in. Because it's something that I find
personally appealing about that. So you know, instinct first and then getting other
people's input on it. And then you get a situation where… this is from a studio that was an
Academy Award winning studio. They were really high-end. They were going to be doing
animated feature films to compete with Pixar. They wanted me to do a penguin character
and this was before the Penguins of Madagascar. This was before Surf's Up. This is years
before that. So I did 119 Different penguin designs all over the place. Maybe I should have
to dig up that work. I have to find to see how varied the designs were but I know that I did
119 individual penguin drawings. And the one that I disliked the most was the first one
that I did. And that's the one that they chose. I'm like what? You’re not serious are you?
And what I learned is never show the people that you're working with anything that you
don't personally like. So I just learned to edit my work. No, you're not gonna see this,
because I guarantee you, they're gonna choose the worst one. And the reason why I think
they do that is because the people that you're showing this to aren't necessarily artists
and they may be subconsciously intimidated by the fact that your work is strong, and so
they naturally gravitate towards the weakest drawing or the weakest design because they
feel more comfortable with that because that's intellectually more accessible to. If they
were to sit down and draw, it would be a poor drawing, professionally speaking, but it
would be closest to the worst drawing that you did. So that's, that's how I see it. This isn't
an isolated incident. This happened over and over and over again. So I learned not to show
them anything that I didn't like. And that's what I teach. Everybody, like, don't show them
anything that you don't like, show your client, only the stuff that you like, and hopefully
they'll like what they see. And if they don't, then you can maybe show them the lousy stuff,
but always edit your work and show them only what you want them to see.
Jef Bernaers 57:20
Such a great tip to work with. In general. I can very much understand like the frustration
of like doing all of these characters and just having one that you really dislike and then
they choose that one
Charles 57:35

45
It happens many times. So that at that point, I would just throw up my hands and said
okay, and then I became emotionally distanced from it. That's when it becomes a job. Yeah,
like you're doing it for the money, and I’ll just give you exactly what you want. Just don't
blame me if this thing doesn't go anywhere. There were projects I did where I was doing
great characters. Where I felt that nobody's done this before. Look at all the things you
can do with this character. And these people just didn't get it. They didn't get it and the
project went nowhere. It went nowhere. I said okay, but I had it written in my contract
that all the work that I did, I can retain it. And so, I still have the characters, even though
they really rejected, technically those characters are mine. I could I still plan on doing
something with it someday. But yeah, those are just some of the different scenarios that
come along, Jeff. It's not a like a cookie-cutter thing. It’s always different from one
situation to another.
Jef Bernaers 1:01:27
The last question is pretty straightforward. Can you describe the stages that you go
through when you're doing concept art, the stages being like a line-art, colour sketches,
turntable, what are the stages that you go through, like in order when you're creating a
character concept?
Charles 1:02:03
That’s no problem. I'm happy to answer that. A lot of times it just starts with a scribble.
just scribbling it just scribble I just scribbled. Yeah, it could be just a little doodle in the
corner of a page. I'm just trying to get a direction going. Then if I like it, I'll kind of build
off that, I screw with that. I was at a friend's house the other day and they got this really
cute cat and I liked the name of the cat, and the cat was acting, really crazy. Just dashing,
running really fast from one room in the house to the other, almost like it was trying to
impress me. I just liked the name for the cat. I just came up with a complete character.
And I thought that the name for the cat was really suitable. I actually sat down and I
actually did rotations on the character and everything. It was just a lot of fun. It just started
with a doodle but I knew that what I was doing, I was not happy with, but it was the
beginning. Sometimes I'll step aside, I'll get away from it. It could be like an hour or two,
it could be a couple of days and I'll go back into it and I'll just see it with a fresh vision and
I can see okay, I like this, I don't like this and sometimes I just completely start over. I,
personally, like rough sketches, which are a little bit more developed than doodles, you
know, just little freeform sketches. I like getting in, and then getting to the point where
you can draw it out roughly where the construction is there. I like showing my work at
that phase but then you want to get ready for publication or whatever or production and
then you do a clean-up on it. The clean-up line and colour and all that other kinds of stuff,
you know, so it just starts from there. And a lot of times what I'll do is I'll have a really
rough sketch. Very rough, and very loose, but there's something in it that I like that I can't
quite duplicate when I tried to redraw it. So I'll scan that. I scan the little doodle, I bring it
into Photoshop, then I'll just expand the drawing and I just draw right on top of that. So I
can keep what it is that I was able to come up with and it won't be distorted. I can capture
what I originally did. That's usually how I work. Start with a little doodle, it could be the
eyes. I could just start drawing eyes and then build a face around that. I usually start with
the head and then I work my way into the body. I personally like doing things that are
very cute. Really cute appealing characters. I like doing villains, I like doing monsters, I

46
like doing, I don't like doing generic things. Know like a person walking down the street
or crowds of people that kind of thing. I've had to do that lots of times in my career, and I
think it's good practice. It’s great for your imagination and stuff, but I like doing marquee
characters, you know the characters that the spotlight is going to be, like marquee
characters, boss characters that kind of thing. And it almost always starts with doodles.
Jef Bernaers 1:06:13
All right. So you would say that doodling and sketching are the most important parts of
the process.
Charles 1:06:20
Yeah, because you're thinking. A lot of times I like to do it all with a non-erasable medium,
like a ballpoint pen or a black line marker. It's healthy because it develops confidence in
your drawing. You can't you know, you stop worrying about well what if it comes out
wrong and what if I make a mistake, whatever, so what? Yeah, just do it again, nobody has
to see that. That drawing, you know, you're not committed. I have tons and tons, probably
1000s of pages of drawings I've done that I'm not happy with that I would’ve never shown
anybody you know, and that's what the advice I give to people when they develop a
portfolio is only put in your best stuff. If you create something you don’t like, then don’t
put it in there.
Appendix B.2: Interview Rafael Natali
Fri, Mar 24, 2023 4:03PM • 25:30
SUMMARY KEYWORDS
character, concept art, sketches, artist, art, research, concept, work, game, 3d, pose,
sculpts, days, material, deadline, spend, needed, create, guess
SPEAKERS
Jef Bernaers, Rafael Natali
Jef Bernaers 03:11
All right. Hello, nice to meet you. How are you doing?
Rafael 03:21
Sorry. I'm doing fine. Nice to meet you as well.
Jef Bernaers 03:30
So, let me just sum it up a little bit. So first I'm going to ask a few questions. minutes. It's
not going to be a lot of questions, about 20 minutes, something like that. And then after
that, it's, we're going to do a little sorting exercise. It's basically just you'll get the concept
art and five sculpts that was made from the concept art and you just have to sort them

47
from good to bet based on four criteria. And then we will be done. Let me just give you a
little bit of backstory first, as to why I'm doing this. So I'm a master student as well apart
from a freelance 3d artist. And for my master's research, I am researching the impact of
time spent on concept art compared to the quality and efficiency of character artists. So
basically, I'm looking at, if you save time on concept art or if you make less concept arts,
how exactly does that impact the character artists speed and quality. So that's what I'm
researching. And I'm doing these interviews to first of all gain a better impression of how
it currently is in the industry and how people think about it. And secondly, I'm doing like
the sorting exercise to see if people can spot the difference between character art made
with only line art or a sketch and character art made with a final render and stuff like that.
So that's the short summary of what I'm doing right now. I have to do this little formal
thing first, where I have to say that you're completely protected for the research. I don't
know if you are under any NDA for companies, but any question I ask if you think I can't
answer this question due to NDA, you can just say that. If you ever answered something,
and then you realize I don't want this to be written down or in the research, you can also
just tell me, don't write that down and I'll scrap it. And then I'm also going to send you a
consent form. Basically it's a little form you have to sign and it basically just says that I
can use what you said in this interview for my research. It also says you can opt out of my
research at any time and then everything you have said to me for the research will be
scrapped. All right. Perfect. So how have you been doing because I think you have your
own studio, right.
Rafael 06:22
First of all, I'm not a native English speaker. My English is not that great. So sorry if I say
something wrong. Right now I'm working basically in three different companies. I have
my own indie studio that I have with some friends that it that are working with SSG as
well. They're working on one of their games. I work with SSG, obviously. And I work with
an outsource company, a game development outsource company here in Brazil. They
make work for a lot of companies for a lot of games, even triple A and stuff.
Jef Bernaers 07:15
So, how long have you been a concept artists now?
Rafael 07:24
I think for about three years, I guess.
Jef Bernaers 07:31
And do you do just to all kinds of concept art or do you mostly focus on like environments
or characters or
Rafael 07:41
As a concept artist, most of my time, I am doing environments, key art, props, basically
most of the time I'm doing props. Characters is definitely the last thing that I do. I
personally like to do characters but as a concept artist I do a few characters but not a lot.
Jef Bernaers 08:19

48
Yeah. All right. So then I just have a few questions. So when you are grading concept art,
how do you determine the optimal amount of time to spend on it? How do you determine
how much time you shouldn't be spending on a piece of concept art?
Rafael 08:38
It depends a lot on the deadline that I have. Because this way I need to try to create a
different pipeline. Like in SSG. I had one, I guess three or four days for one character. So I
didn't work on it full-time every day. So I needed to do it way more quicker. But I had
another job where I had like one month to do a character so you know a big difference.
Yeah, but overall, I give myself like one week to the character. So basically, I separate by
days, the first day I spend all day searching for references, what is my goal, who exactly is
the character, and to create a little bit of his backstory and stuff. The two days after I spent
mostly doing sketches. I spend the two days after doing concepts. Then the day after I
refine one of these concepts to make a more polished lineart and stuff. And the two days
after I tend to make a turnaround or a back pose. So you can see both of the sides that the
character has. And if needed, I make a rendered pose as well. I pick one pose and start
rendering to describe the material and stuff.
Jef Bernaers 09:02
So when you work on a character, how do you collaborate with other team members, for
example, the character modellers and the character artists? How do you work together
with them when you're concepting?
Rafael 11:33
Okay. Basically, first of all before starting my concept, I need to know exactly what
purpose I need this character for. So I need to know what complexity of polygons will be
available for work, if I needed to make our character a little bit more low poly or if I can
go crazy with the details and all the stuff. I need to talk with the animators as well so I can
be sure what can I do for the accessories and cloth and stuff and then I would probably
need to talk with the game designer. To get an idea of the characters attacks and
movements and how this character would be used in the game. Like is it going to be a
minion for the boss? Or would be the boss or would it be a playable character or any of
these things.
Jef Bernaers 13:06
Can you recall a time or a moment that you have created concept art and that it negatively
impacted like the quality or the efficiency of a character artist further down the line?
Rafael 13:20
Yeah, yeah, definitely. I guess. When you don't have a lot of time to do the research. I
always like to have at least one day or two to do research and at least two days to do
sketches. Because if you don't do that, I guess you end up with a generic character that
doesn't have like a unique feeling for it or have like a cool motif. I can do a character in
two days, or three days but the character would definitely be a lot more generic and beat
down and wouldn't have something cool you know?
Jef Bernaers 14:13

49
Can you describe how you decide when your concept art is ready? If you didn't have a
deadline? How do you decide that a piece of concept art is done, it's finished?
Rafael 14:27
I think that concept art doesn't have to be finished to be delivered. So yeah, I guess if I
don't have a deadline, the thing is that I need to be talking to my co-workers to see the
character is at the point that it should be done. If you need any changes or need
improvements for the 3d artist or the animator. So yeah, talk to your peers, your co-
workers.
Jef Bernaers 14:34
And do you think this changes depending on the pipeline, do you think for example, in
some cases can get away with less concept art and in some cases you need more?
Rafael 15:25
There is a lot of difference depending on the pipeline. Sometimes if you have a clear path
that you need to take to do this character, so the character is well defined for like the lore
or where it is going to be used. The games sometimes already have their art direction well
established. In these situations it's way easier. Sometimes I don't need to do a lot of
sketches. I can be more spot on, more direct. But sometimes, especially when you don't
have a clear direction or you don't have like a good briefing for what the character is, You
will probably need to do a lot more sketches, a lot more research, and all these things.
Jef Bernaers 16:31
Can you describe all the stages you go through when you're creating concept arts and the
stages being, for example line-art, colour sketches, turnarounds? Yes.
Rafael 16:45
First of all, I do like some very small sketches like so I can just get a good idea of the
silhouette of the character. I don't know I make like 10, 15 sometimes 20 of these. I then
pick about eight of these tiny small sketches and I develop them a little bit further, so I
upscale them and then start refining the silhouette, refining some values on the character.
After this if I have like these six rough sketches, I can do two things. If I like a lot of one
sketch, I can just take it and develop based on this. If I'm zoomed out, I can take like two
or three sketches, make some sketches, and make more refined ones. When I have these
more refined sketches, I pick one and make some small variations in this character. Like,
I have this sketch that is working, but what would he look like if I changed his hairs, or if
I change this cloth here, or if I change this little part here on his weapon. So I make these
little tiny variations of the character until I think okay, this is okay, this is great. So I start
making the line art, I don't make a very refined and pretty line-art. I do more rough one.
And then I place the flat colours, I start doing some occlusions, and then I put some light
and shadow, some reflective light, then I define a little bit more of the material. So what
materials could be more glossy or shiny, what's going to be more dirty? And then if
needed, I start I removed the line art or start rendering the pose.
Jef Bernaers 19:40

50
Which of these steps do you think are the most important to a character artist when
sculpting.
Rafael 19:55
When I'm handling the file for the character artists, okay? You have like, solid concept. So
you have a concept that is easy to translate to 3D. Sometimes you have a concept that's
too painterly rough and you when you give it to the 3D artist, they don't know exactly
what is happening here. Then they might have some trouble translating it to 3D And then
you might have some problems. It's way harder to turn to change a 3D model than a 2D
drawing. So it's good to have a well-defined solid drawing for 3D. Especially if you can do
some poses so you can see what is happening. So if you have for example a shoulder pad,
but you don't know exactly what is happening there because, in the view that you draw,
it is ambiguous. So you make some other views for the front of these shoulder pads so the
3D artists see what is happening. And another thing that I think is important, depending
on the project, is to do a material guideline. So check, what exactly are the materials of the
character, especially if the character has some crazy materials, like some magic bloody
liquid?
Jef Bernaers 21:55
Yeah. All right. So that were the questions already.
Appendix B.3: Interview Jitske Habekotté
Tue, Mar 14, 2023 11:03AM
SUMMARY KEYWORDS
concept art, working, question, sketch, artists, pipeline, interview, research, art, 3d,
company, stages, people, person, impact, team, steps, ai, character
SPEAKERS
Jef Bernaers, Jitske Habekotté
Jef Bernaers 00:13
Let's do it. So first of all, thank you for the opportunity to talk to you again I know my
research is pretty close to your research. So you're very much familiar with the subject
itself. That is a very much the reason why I wanted to talk to you as well. I first want to
ask them questions about your current job as well. Because now you are working as a
freelance concept artist, but specifically for one company if that is correct?
Jits 01:54
I am currently working as an independent game developer. And I don't take freelance
work anymore. I did one freelance project and I fundamentally hated working for another
client because in the end, that still means that you're doing work for someone else. So I'm
currently working on my own game titles.

51
Jef Bernaers 02:17
That's cool and by yourself or do you have a team backing you up?
Jits 02:22
By myself for now the intention is to eventually expand a little bit. I want to keep it small
but get one or two more people to join in will be great, because now I have to learn how
to code so well, we'll burn that bridge when we get to it, but it means I do a little bit of
everything, which is what I enjoy doing. But of course also encompasses some of the jobs
that I'm less familiar with. So yeah, we'll see.
Jef Bernaers 02:51
If at any point you need a character artist, you have my email.
Jits 02:54
Hey, see, this is how the Masters just connects people.
Jef Bernaers 03:01
Yeah, so I have to do this little formal thing at the start that you're very much familiar
with. So at any point in the interview, if you are not comfortable with answering the
question, you can just mention this you don't need to answer any of the questions if you
don't want to. I cannot answer this question due to NDA is also a valid response to any
question. But this should not be a problem as most questions are aimed to your personal
experience and not to company experiences. But I also wanted to mention this. After the
interview, I am also going to email you a consent form that basically just says I can use
this interview for my research and I can transcribe it. And the consent form also
mentioned that at any point in time, you're allowed to withdraw from the research and
this will also withdraw any previous statements or interviews that you have done from
my research to protect you at any point in your career if you want to withdraw so that if
you want to the interview can be anonymized which means that in the transcriptions,
every identifying feature, statement, and anything that identifies you as a person will be
censored. If that is something you want you can tell me now.
Jits 04:30
I don't think that's going to be necessary, but we'll see maybe I'll say something super
controversial *laughs*.
Jef Bernaers 04:38
After the interview. When I get to a publishing stage and a more final stage of my research.
I will also send out emails to everyone that I did interviews with, containing a small
summary of any direct quotes that I may have used so you can confirm whether or not
you agree with the stuff that I quote you on, because it will be a problem if I quote you and
you don't agree
Jits 05:03
That it's very fair. Yep.
Jef Bernaers 05:08

52
So yeah. That's that, I want to be as open as possible to the people that I'm working with
about that. And then we can go a little more in depth, I guess. So I think the last time we
talked we discussed, because you were already aware that I that I switched from
anthropomorphic to specifically only the impact of counterpart on the character creation
pipeline, I believe. The last time we talked I was going to research the impact of concept
art on both the texturing and sculpting stages. But this has changed. I have decided to only
focus on sculpting to reduce the scale or scope of my research and to improve the quality
of the data that I get from my tests. So currently, very short, I'm researching the influence
of the individual steps of the concept art pipeline, the individual steps being lineart,
values, sketches, colour sketches, you know them very well, and how specifically those
steps impacts the quality and efficiency of sculptors in their first stages of their work. So
if you were to give a sculptor only a line art, does that really make him sculpt that much
lower or does it really impact his quality that much? That's basically what I'm what I'm
researching right now. I did a practical test with students for this, well to test this
practically because I didn't feel like I would be able to get this all from interviews. The
flow of this interview is going to be first I'm going to ask you a few questions is going to
be very short, something like 20 minutes about my research to get a sense of how you
think about how you think about this topic and what your thoughts are and then we're
going to do a little Q sorting exercise where we get a bunch of images that are all created
by students. You sort them on which one you find is the best and which one you find is the
worst based on the criteria that is being investigated. Do you have any questions about
this?
Jits 07:34
No, sounds entirely clear. I am going to quickly turn on my noise cancelling so I'm not
constantly distracted by sounds in the background.
Jef Bernaers 07:48
All right, perfect. So first, a very broad question that I just wanted to get a sense of how
this is impacting concept artists and artists in general, have the recent breakthroughs in
AI tools, specifically AI concept art. Tools, have they impacted your workflow in any way?
Jits 08:08
No, I don't personally use AI tools. I find the discussion around them very interesting. You
can stop me at any time if I'm going too far off track. I find the discussion very interesting.
Mostly the ethical question with AI. It's difficult terrain, I truly believe that AI tools will
become a very standard practice into especially into our industry. Let's face it, and I think
that being able to write comprehensive AI prompts will be a job on its own. The way it's
currently being done, however, is unethical. It is theft from individual artists. And I think
initially in the beginning stages which we're entering now. It is going to take away from
concept artists, eventually people will realize that there are jobs that AI can do really well
and jobs that individual artists can do really well and there will be sort of a harmony like
a sort of symbiotic relationship but we're not there yet.
Jef Bernaers 09:12
It will very much improve the speed at which we can work which would be very good for
the games industry to just be able to make better games much faster.

53
Jits 09:22
Yeah, I would think so. Yes.
Jef Bernaers 09:25
Perfect. So when you do concept art, because now you are your own boss, so you have to
plan everything in very well. Especially making a games project is all about planning. It's
very difficult. So when you are doing concept art for your game, how do you decide or
determine the optimal amount of time and resources to devote to concept arts? For
characters? How do you decide this is how much am I going to spend and I'm not going to
spend more time on this.
Jits 09:59
Um, you're gonna hate my answer. I'm a one person company. I can I can just pick up tasks
like Oh, today I feel like drawing so I'm going to do some concept art or today I feel like
doing something more technical. So I'm going to work in Blueprints the whole day. And
the truth is, the answer is until I feel they're done, which is a luxury that most companies
cannot afford. But I am very mindfully setting my company up in a way where it'll be a
healthy work environment. For me that works with me and the way I think and the way I
work.
Jef Bernaers 10:39
Yeah. So it's very important to keep that your yourself in mind when working and to make
sure that you're unhappy with the way you're working. It's not that efficiency is not the
most important factor.
Jits 10:54
No, I set broad goals, sort of where I want to be in a month's time in two months time, you
know, so to keep myself engaged and to make sure that I stick to somewhat of a schedule,
but I very purposefully do not impose hard deadlines.
Jef Bernaers 11:11
Alright, that's fair. So the next question I can't really ask, but it's a funny one to ask
*laughs*. How do you collaborate with your other team members?
Jits 11:23
Oh yeah, I hate my other team members they suck *laughs*. Yeah, it's really weird when
you're working by yourself on something. Because you're looking at stuff like coding you
did the day before and you're like Jesus, our coder really fucked up here, and then you're
like oh that's me *laughs*. In general, looking back, I've really enjoyed working in teams,
especially in sort of a production type of role. But the experience now in a one person
team is very different.
Jef Bernaers 12:02
So what I gather from this working from a one person team, the one thing that you lack
and that you might miss yourself sometimes is this confirmation step. This step that
people look at your work and say, yes, this is good, or people or another set of eyes looks

54
at your work and says hmm maybe this can be a bit better and then you see it as well. Is
there a way you work around this? Or do you just trust your instincts and just keep it
going?
Jits 12:31
I mostly trust my instincts. I'm very close with my brother. My brother works in the UK.
At rebellion. I talk to him a lot. What I miss the most from working in a regular team is the
ability to not just confirm each other's work or to help each other Figure out problems but
it's also just the bouncing ideas of each other, the rubber duckie effect. So my brother and
I do that when we're in chat. And then you know, he talks to me about his stuff. I talk to
him about my stuff, and then we just sort of bounce ideas off each other so
Jef Bernaers 13:12
Alright, so the next question I think is might be particularly interesting to you because it's
it's once again actually in the two persons but now, it is aimed at you from two different
perspectives. So can you describe a time when your concept art negatively impacted your
own efficiency in further steps and that you had to go back to the concept art to change
things because something was too difficult.
Jits 13:35
Oh, that's an interesting one. Yeah, I do. I do sometimes run into the issue where I think
of something. And then when I start to execute it, I'm like, Oh, wow, I really took the most
difficult route first. And I don't necessarily go back to the concept art and change that, but
I just adjust on the fly. But that's also the luxury of not having to re-evaluate with a team
to see if everyone is on board with the changes you're about to make. In general, I do a lot
of things on the fly and not so much going back into the concept art.
Jef Bernaers 14:14
Yeah. That makes a lot of sense. So, do you create 2d games, 3d games? How exactly is
your pipeline currently?
Jits 14:31
I have an idea for a bigger game. That's going to be sort of two and a half D. So like 3d
characters, mostly 2d elements. I what I'm currently working on is an entirely 2d game,
mostly to sort of cut out the 3d aspects and be able to do something a little bit faster, just
to get back into the workflow. So that's the first project and then later on, I'm fully
intending to also go 3D.
Jef Bernaers 14:58
Yeah. So while you're creating concept arts and well, in your case, the further steps in the
pipeline for creating concept art would be starting animating these concepts and stuff like
that. So, when you are creating concept art, can you describe your decision making
process for determining when a piece of concept art is ready and you can move on?
Jits 15:27

55
It's another one that's mostly based in feeling, it's mostly that feeling of when I look at the
character doesn't match what I felt about them when I was thinking about them. Initially,
like the idea that I have about them, does the visual match that vibe?
Jef Bernaers 15:46
That's very interesting, right? It's a very big difference. And it's very interesting to look at
how exactly this would influence your workflow and quality in the end is to compare to
other companies that have like this extensive list of criteria to look at. How important is
that list actually?
Jits 16:07
Yeah, I think I think that's, that's, that's the interesting part when you start comparing,
you know, in these and very large scale companies, because I personally really liked the
most indies work, where it's a lot more freeform a lot more following your feeling, but at
the same time, could a bigger company function in that sort of system? Or do you need a
more hierarchical, structured setting of people imposing rules and guidelines to make
sure that everyone works in the same mindset.
Jef Bernaers 16:47
Do you think the in depth the impact of concept art is true universally for the next steps
in the pipeline? Or do you think that certain steps of your concept art have a bigger impact
on different steps in the next part of the pipeline? I think I phrased that very confusingly
*laughs*
Jits 17:14
I think I get what you mean. So, for me, I'm a one person company. So I am both the vision
holder and the person executing it. While I'm executing it. I know exactly what I was
thinking and what sort of the idea was that we're working towards. When you're with a
small indie team, you usually unless you have a pretty shit team, but you usually have
really good communication. Everyone just sort of knows what everyone is doing. There's
a lot more direct collaboration. So, my assumption is that if you scale up the team, that
becomes more impactful because what we saw, for example, Someone I know went to a
large studio, and it turned out that none of the different parts of the company were really
in touch with each other. Like they weren't really aware of each other's working
processes, what they were working on, how they work on certain things. So in that case,
if you have concept art that goes on to the next department, that concept art needs to relay
everything. Whereas in a smaller team, or even in a one person team, if you make the
concept art, even if half the details are missing, people know what you're working
towards. So even with fragments of it missing everyone has the same idea of where you're
headed. So I think you can omit more without it becoming a direct issue. Yeah, I think
that's a fair assumption. So following the logic, off set assumption, I'm going to make an
assumption of my own and say that you don't finish concept art for yourself to the
completion that you would see in other larger companies, you would not probably go to
the ends of making a final render or turntable for your concept art, I'm assuming. Yep.
Yeah, that's right.
Jef Bernaers 17:51

56
So what would be the stage that you often stop at and say, I'm going to go further.
Jits 18:12
It depends a little bit on whether or not I'm working in 2d or in 3d because if I do 2d, then
eventually the concept art that I made is gonna fall into the final product. If I'm working
to 3d, I usually stop at the sketch phase. Sometimes I make a value sketch. If I really liked
the art I'm working on I'll make a fully rendered one, but not so much to improve the
pipeline further on, but just to be like, I really like working on this character. So I'm just
gonna make a full render. So, that's more for own enjoyment than it is for the pipeline's
sake. But sketch I would say for me sketch or value sketch are the most valuable ones that
I use.
Jef Bernaers 20:23
All right.
Jits 20:25
Which internally I hate because it feels I don't know it feels like you're omitting so much.
And as an artist, I want to say like now everything needs to be fully rendered and look
awesome, but that's just me wanting everything to look awesome and not so much that
you actually need it for the pipeline, at least not for me.
Jef Bernaers 20:45
Yeah, definitely. So can you describe to me the stages you go through when creating
concept art and when it would stages being a line art, value sketches and stuff like that?
Jits 21:00
Um, we're talking specifically characters right. I usually sit down, I have a portable tablet,
so I usually sit down somewhere and I have a nice cup of tea and then I start just drawing.
One of the things I always do is gather reference images, even if it's just a Pinterest style
board, just to get an idea of what is the aesthetic of the person. Which for me, always helps
very much to put that visual together. For the project that I'm working on now, I have a
group of people are from a different time setting in the Netherlands. So I'm looking up like
what do people wear? What did they look like? How did the ladies wear their hair? How
did the gentleman wear their hair? All of that stuff. And then I just start sketching. I will
say I'm the type of artists that if I don't like what I'm working on, I just delete it and start
over. So I don't have a ton of different sketches. Usually by the time I'm done, I have one
but that one is then the one that I'm going to be using further down the pipeline. From
sketch I work into value sketch if I feel like needed or if I just enjoy working on the project.
Jef Bernaers 22:22
Yeah, and how does this relate to the 3D stage do you take additional steps working
towards 3d compared to when you're just working on 2D work? Or, is this just the same
pipeline?
Jits 22:38
Um, I think for 3d, especially if you work with other people, you know, like if you work
with other people, I feel a value sketch. is so much more valuable than a regular line art

57
sketch. Simply because you can convey the depth you can convey the roundness of
something you can convey texture, even in very broad strokes. So I feel like that adds a lot
of visual information. Where if you want to move to 3D, I would not recommend skipping
the value sketch phase.
Jef Bernaers 23:10
Yeah. All right. So I just have some demographic questions before we continue to the to
the test. So what is your current job title?
Jits 23:27
Indie Game Developer.
Jef Bernaers 23:30
How many years of experience do you have creating Game Art?
Jits 23:40
We are now in 2023. Do we count my bachelor's?
Jef Bernaers 23:43
Let's say job experience.
Jits 23:48
Oh geez. When did I graduate? Let's say four years,
Jef Bernaers 23:54
Four years of work experience and do in these four years also get experience creating
Character Art and how much would that be? Or have you just always been involved with
character art as well?
Jits 24:07
Yeah. Usually when I create art, it's a little bit of everything so character art is usually in
there. I also prefer to do characters over like environments and stuff.
Jef Bernaers 24:34
Alright, so that's so that was the interview. Like I said, it's pretty short. Now we can do the
sorting exercise and after the sorting exercise, we can go over the details a little bit more
and then I can also mention some of my current conclusions. And discuss them with you.
But I don't want to do that before doing the exercise because I do not want to taint your
opinion.
Appendix B.4: Interview Radu Ianescu
Thu, Mar 09, 2023 6:29PM
SUMMARY KEYWORDS
concept art, artists, character, work, concept, questions, details, deliveries, milestone, art,
production, sculpting, interview, lacking, terms, shading, feedback, ai, important, artwork

58
SPEAKERS
Jef Bernaers, Radu Ianescu
Jef Bernaers 00:00
So, this interview, I'm going to start off by asking you a few questions. None of the
questions should cross any NDA boundaries. But if at any point you feel that you can't
answer the question, because of NDA restrictions, you can just tell me "I can't answer this
because of NDA" and that's fine. But normally, that should not be a problem. After the
interview, I'm going to send you a consent form which is basically just a little form that
says I can use this interview for my research. It will be very short you can quickly read
through and see what it says and then if you just sign it and send it back to me then I can
use what we get from the interview. It basically just says, I can use whatever you said in
this interview for my research, and I can transcribe it. So I'm going to ask you a few short
questions. It should only take about 20 minutes. And then we're going to do a sorting
exercise kind of, I'll explain that later. But in short, the exercise will be a few character
sculpts and you will have to sort them from which one you think are good and which one
you think are bad. Okay, but that's after. So, first of all, how are you doing? I heard it's
pretty busy right now for you.
Radu 01:40
Yeah, we're trying to get the candy wrapper design for the whole project, because we are
at that point where we need to start bringing in additional funding like you know, and I'm
trying to do that because there's some roles lacking from our team one would be
something I would offer you for sure. Like something like in terms of visuals and overall
it's not necessarily art direction because it's more like overseeing stuff, you know. Yeah,
similar to something like the whole structure in the company is not very traditional in the
approach itself and everything but when we have documentation ready, I'm gonna send
this to you because I've noticed a lot of the people from the team and our collaborators,
they're not aware of what we do or what we try to do. They just know they are working
for a game but it's not clearly is either from us providing with too few information or is
either from the lack of interest coming from their side because some people just want to
work in a project and do the job and when the timer runs out the off schedule, you know,
but yeah, it's different from us from within the core team because we're here for like 1214
hours. You know, what, you have an exercise like this because I'm taking a break from the
stress and, and, yeah, anxiety.
Jef Bernaers 03:39
Yeah, that sounds like a plan. Okay, great. I hope you can also get something out of the
interview, and maybe I don't,
Radu 03:47
I don't know what I've agreed to is gonna be it's gonna be a very interesting thing. I don't
think you share the details. But I may be jumped over the message.

59
Jef Bernaers 04:01
I did it in the past, but I think it's changed a bit. So let me just give you a little bit of a
rundown of what I'm currently researching. why I'm doing it. So I'm researching right
now I am researching the impact of time spent on concept arts versus the efficiency of
character artists in different stages. So basically, I'm trying to determine if you save time
on concept arts, or if you spend more time on concept art does it make the character
artists work slower or work faster, and maybe come to a conclusion that we are spending
too much time or too little time on concept art. So that is what I'm looking at, in a broad
sense.
Radu 04:45
Okay, it's good. Because I also read in the meantime, your message, I don't have
experience. I mean, I've done hobby work, but not in the in the production pipeline like a
professional. But overseeing everything that involves character creation is or was mostly
me.
Jef Bernaers 05:09
So that's very important to me as well too because you are responsible for it in a sense.
Radu 05:15
I can give you like the upper management. Sir, go ahead.
Jef Bernaers 05:26
Alright, perfect. So what I've done so far from the research is what I did first is I did a test
with some students. We're going to discuss the results from my test later in the interview.
I hope they will be interesting to you. But first, yeah, just some general questions. One I
know that is very close to your heart in a way is AI and AI art. Have the recent
breakthroughs in AI art. Tools, have they impacted your workflow or the way to advice,
workflows in any way?
Radu 06:06
Yes, yes. They have in a good way.
Jef Bernaers 06:09
Yeah. So do you implement them do you inspire character artistry to use
Radu 06:18
I myself when I do artwork or when I'm building scenes or building a banner, or graphical
or something like that? I do use AI like a cheat code or shortcut to getting some elements
that you would need. You could get something for a backplate in in a scene, which we have
done and is working out quite nice. I know a lot of artists are against it because they're
like AI sucks. I do use it. I do find it something you would implement and I do think is
going to be an efficient tool in the future. The thing is, there's a lot of hype on it right now,
which was in the previous until like two months ago, and now it's slowly died. The hype
itself, but the tools are still there and they're growing and I think they're going to be very
useful someday.

60
Jef Bernaers 07:21
Yeah, they are already quite useful now.
Radu 07:27
Yeah, okay. Someday they're gonna be a little more useful then. They are still limited in
terms of consistency, repetitiveness, and stuff like this. So when they do that, when they
bring that to more professional level, I think that's when it's going to be very important
to start using them.
Jef Bernaers 07:54
Yeah. So you mentioned earlier that you are in charge, in a sense of the character artists
and what comes out of what comes out of the character art by blank. So how do you
determine the optimal amount of time that is supposed to be spent on concept art for the
character? What are the criteria that you look at and how do you determine which is a
good amount of time to spend on concept art?
Radu 08:26
Previous work deliveries done by other artists. We've worked with not only one we had
we had many people to compare a few people to compare deliveries. So you kind of guess
how much a person is going to spend based on other people or interactions with other
people, which you find that bring value to your project. So in a sense, you do want to aim
for those professionals who have done a better job. And you are always aware that some
people are going to be faster some will be slower but you do understand like, what's the
minimum and maximum they're allowed to spend based on previous experience with
other artists?
Jef Bernaers 09:14
Yeah, so what you're saying is you actually adapt it to the artists based on how fast you
know that they normally work.
Radu 09:21
Yeah, it's based on previous work
Jef Bernaers 09:26
Yeah. So you have like a benchmark, like they have to work at least this fast to work for
us.
Radu 09:34
Usually people say, when do you need this? The answer is always, either it's something
super important, we need like tomorrow because I need something done very fast. Can
you do this by three days? But I know like they're gonna spend like two or three days to
get the concept of art started to a point that we can have a first feedback round, then I am
aware it's going to take the same amount of time. So I decide basically on what I need, but
if time is not pressuring us, I will always say we need it in X amount of time based on what
other people did.
Jef Bernaers 10:26

61
So when you when you work with concept artists and give them feedback, how, how
exactly do you collaborate with them? Is it just in terms of feedback rounds? Or is it more
how do you communicate with concept artists while they're creating concept arts?
Radu 10:53
I'm trying to Figure out the question it's how do it technically, do I provide like feedback
on their drawings or can you like try to rephrase the question?
Jef Bernaers 11:07
How do you how do you work together? Like what is what does the concept artists expect
from you when they're when they're working with you? What do you give to the concept
artists? What is your role in their process?
Radu 11:25
Oka, I understand. It depends on my abilities on what concept art we have, in general, for
example, I would use other references if it's something I can do myself, I could say. So we
would either provide feedback based on reference, or I can jump in and provide a little
more guidance in terms of what I can put over the actual person. If it's something like
some effects, some post processing type, changing the colours, whatever, but if it's in
terms of the overall way it looks, I always refer to other references. Yeah. So it's either one
of these or both of them.
Jef Bernaers 12:24
And are the character artists somehow involved in the in the concept art process or do
they only see the concept art when it is finished?
Radu 12:35
We have both the 3d modelers watch the concept art channel. And we also have the
animation rigger watch it because in the pipeline for the production of characters, you
might wind up with some issues which we've come across. The rigger might say this 3d
character cannot be rigged accordingly and he's going to have issues.
Jef Bernaers 13:00
Yeah. So they always check if it's possible to create
Radu 13:04
they always have a look on the concept art for a character from the start. Yeah.
Jef Bernaers 13:12
That's good. So like you said you, you've had some problems with concept art in the past,
which is which is normal. Can you like describe a time when concept art has negatively
impacted the efficiency or the speed or the quality of one of your character artists? Or has
this not occurred?
Radu 13:37
I think I think this is very subjective, from artist to artist. And I think it has to be a good
synergy between the concept art and artists. Because we have the issue of some character

62
artists saying, this concept art is lacking in details, but we had other that did understand
our request and they were more autonomous and said, okay, I can do this myself and they
just had that self drive which led them to just do something. Some of the character artists
will say, this is not clear for me before doing it. Others will just do what they understand
from it, and say, I just understood this guy has like, these leaves on his back so is this okay
what I have done. So it's a matter of like stepping up. Yeah, there's good concept art and
bad concept art. So it can it can you have to have the right artist in the in both jobs, to have
good character designs.
Jef Bernaers 15:06
How do you decide when a concept art piece is done. What things do you mostly look at
to say yes, this concept per piece is ready, we can start working on the characters.
Radu 15:25
At first, so there's two levels to this or two milestones. One is when you start a project and
the other is when you are already in the production phase. So when you do concept pre
production, you will never have the right answer to that. You have to wait until you
accumulate stuff. So it's all about experiencing but when you're in the production pipeline,
you always put that artwork next to other ones that were validated. Then you can see if
the detail level is there, if the synergy with the other concept pieces are there, if they
follow the same art style and I'm not talking about the tiny details, because they need to
provide something with quality. So it like I said, it depends from where you get if you're
measuring the production pipeline is like I said, when you compare with other artworks
they need to be in the same harmony you know, on the same level.
Jef Bernaers 16:34
So as a concept artists when he when he does create concept art, he always goes through
a lot of steps. He starts out with some sketches and creates a line art, some colour
sketches. What do you think are the most important steps that a character concept artists
always have? Or always go through?
Radu 16:54
Okay, as an art director as the manager for this I'm always like super okay with the first
milestone being in two types of deliveries because I've seen this done in flat colour and
no shading, or just sketch, so I'm okay with whichever that's going to be the first
milestone. Yeah, so that's the most the most important step is give me a volume. I have to
see the proportions. I don't need details, I don't need shading. I just need either the colours
or either the lineart sketch. So that will be the first step you want to upload that I spaced
out on the question. Now, all of the steps are just the first one.
Jef Bernaers 17:55
The most important ones to you, the steps that you find most important to have in the
final result.
Radu 18:00
So first, the first important step is either flat colour volumes or lineart volumes. The next
one would be shading and details and the last one would be, if possible, either elements

63
that are attached to the character, let's say armour, sword shield, whatever, or something
the character is in relation to like a scene, let's say a flat ground, it just has to be a little
integrated. So even if you have like a disc, like we've done with the little elements that
brings that character into a location.
Jef Bernaers 18:49
Yeah, to bring a bit more backstory.
Radu 18:52
And more specifically, the last one should also involve like a side view front view of back
view of the graphics.
Jef Bernaers 19:01
Like a turntable
Radu 19:01
Exactly a turntable three view piece. So the 3d character, artist can work with a good a
good amount of information.
Jef Bernaers 19:17
We just talked about all the stages you go through. Do you think everything has to be
finished before you can continue and start sculpting or do you think maybe you could stop
earlier and start sculpting with the with an earlier version already? Or do you think all of
the concept art needs to be finished first and only then you can start sculpting
Radu 19:39
depends on the artist like I said before, but I think you can start off the volumes like you
could start the blockout for the character when you have like step one finished. You could
start that or you could start at step two, and just you could have them working in in a
synchronized manner but I don't know if it's worth the hassle because you're just saving
one or two days, and the risk of adding frustration to the characters artists. You could
work efficiently, but I don't know if it's that important to be under pressure like this. I
wouldn't do it. I wouldn't go, only if I'm time pressured. And I would like to be prepared
ahead of time.
Jef Bernaers 20:41
But otherwise, you would just wait for it all to be finished and then yeah, for
Radu 20:45
Yeah, for sure.
Jef Bernaers 20:47
So those were all of my questions, it was very interesting. Now I'm going to send you a
link to a Miro board. And then I'll explain the sorting exercise.

64
Appendix B.5 Interview Lothar Zhou
Wed, Mar 08, 2023 10:06 AM
SUMMARY KEYWORDS
Character, concept art, artists, ai, sculpting, game, concept, defined, impact, studios,
sketches, specifically, working, stage, uniform, people, image, influence, environment,
questions
SPEAKERS
Jef Bernaers, Lothar Zhou
Jef Bernaers 0:09
either Hello good morning. Good morning. Me Well yes I do. Yes. All right perfect. I just set
something up real quick
Lothar 0:26
Oh are you doing I was actually almost missed your call because they can't remember they
didn't I was walking that in the labs and they were the students you know talking to them
but just came back in time. Perfect
Jef Bernaers 0:44
here
Lothar 0:53
and there we go. Okay. All right. Perfect.
Jef Bernaers 0:56
First of all, thank you for sitting down with me and I appreciate it. So let me give you a
little rundown of what I've been doing so far, and where I am right now, and then we can
have a little chat. So the last time we talked, I was still planning on researching
anthropomorphic characters. So my research has changed quite a bit in the sense that it
is still focused towards characters but I focused more specifically on a single step in the
pipeline. Currently I am researching specifically the impacts of concept art, specifically
time spent on concept art, the quality and efficiency of character artists in the further
steps Yes. So I'm trying to determine how a concept artist how we influences, the
performance of character artists further down the line. Yeah. This is what I'm looking at
right now.
Lothar 2:00
in a broader sense, or in a narrower sense. I mean,
Jef Bernaers 2:06
in a narrower sense, I think, well, for the concept art, specifically, I'm looking quite
broadly, I'm looking at the individual principles and steps and how they influence but then
the further steps I'm currently only looking at sculpting because I don't have time to look

65
at every step unfortunately. So very specifically, I'm looking at how concept artists
influence the quality of the sculpting stage for character artists.
Lothar 2:39
Um, I would say is that question for me now? Yeah.
Jef Bernaers 2:47
Well, you can ask if you already have something to say about it you can.
Lothar 2:52
If you're talking about sculpt of the concept art, how much is going to be impacted by the
initial you know, concept in has a huge influence? I mean, I would say up to almost 85%
of the final quality of the sculpt will be, you know, defined by the initial concept. So the
concept of you know, would totally define the form, the silhouette, the colour, the
material, the level of detail. For characters it influences what kind of costume you got,
what kind of hairstyle you got, the makeup also has a huge influence, and what kind of
accessory the character carries. Yeah, and plus the weapons and everything extra. So it is
a huge influence. There's no doubt about it. Yeah.
Jef Bernaers 3:46
Yeah, so I have some questions about this. First I have some more general questions about
concept art and about your workflow, and then I have some more questions aimed
towards my subject. So one of the first questions is one that you also mentioned earlier, is
that if the recent breakthroughs in AI tools have affected your workflow in any way? Yeah.
So how have the AI tools impacted the way you currently work or if you start a new piece,
do you look at it differently because of AI? This way?
Jef Bernaers 4:30
This is a vast topic, you know, we could cover for a long time from now on and it's still in
a very early stage. We could evaluate the impact and the potential impact of AI image
generator versus you know, traditional concept art. And for me personally, because I've
seen a lot of people especially within the concept community, the impact for concept
artists or character concept artist, by AI is on an unimaginable level. The thing is, you
know, a lot of people have psychologically been impacted by the speed and the quality
and iteration of AI, no actual human artist could even match to that quality. Yeah, but as
far as I can see, there's something we can do. You know, I know a lot of artists, a huge
percentage of artists been put off by this AI and they're naturally trying to resist or refuse
or to deny the actual efficiency of AI and the quality of AI but there is no way we can we
can deny it or refuse it. Instead in the future a lot of big studios like Tencent, or some other
studios are already implementing AI into the workflow. And there's something we can do
is to use AI as a foundation. So your starting point, is already higher than traditionally let's
say you start from scratch you start something from value, and compose your image and
try to study light, this is not the case anymore. So you do already have a really high
standard as a bar for foundation, and then based on that you could make a lot more fine
tunes and fine adjustment. So basically make the AI image more personalized and even
more suited for, you know, personal taste or the specific requirement because AI still

66
haven't bit of trouble to deal with very specific requirement and details for now. Yeah, we
don't know about future.
Lothar 4:30
I also think that it will have a big impact. Yeah. It is hard to it's hard to determine right
now. What its place exactly will be.
Jef Bernaers 7:03
Yeah, you're right. Yeah.
But it is very interesting to see it evolve in real-time.
Lothar 7:15
As much as I hated it, I know I can't avoid it. I have to use it for my benefit. Exactly. Yeah.
Jef Bernaers 7:22
So like you said the AI, it potentially saves a lot of time for artists. as well. This is also
something I'm looking at, what would be the optimal amount of time to spend on a
concept art and at what point might you be wasting time on a concept art? So when you're
working on concept art, how do you determine the optimal amount of time and resources
to spend on the concept art?
Lothar 7:51
Yes, good question. But the question varies the answer varies, it all depends on the
experience level and how long you have been working in the relevant area. So a junior
concept art , when they look at things, or they judge or evaluate how much time they need,
even with the help of AI, it could take longer. For much more experienced artists from the
past or from the past experience from all the success and failure. You can probably much
quicker evaluate how much time you need to successfully build a concept or a serious
concept of a certain project. So, that varies a lot for experienced artist with the help of AI
or without or with numerous amount of resource you see the resource for senior artists
are also much more complete than their junior artists. Let's say regardless how hard you
try online. I know these days people got almost the equal amount of connection right
online. It's not like 20 years ago, you have the library and someone else doesn't. Still the
evaluation process of how this concept will be successful varies widely from junior to
senior artist or from the student point of view. It's difficult to say.
Jef Bernaers 9:18
So like you said, you also mentioned resources in there.
Lothar 9:23
Resource yeah, the senior artists have probably been working in industry in different
studios for a couple of years. So they naturally have or can obtain all sorts of a company
or studios documentation or from other concept artists who have been working with you
and show you new things. There is lots of documentation and material which the
companies could provide you while the junior artist or students don't have the privilege.

67
Jef Bernaers 9:56
Yeah. So about these resources, I also look at a large team or team members, like resources
you can use. So in the concept art stage, how do you collaborate with the main team
members, how are they involved in your concept process and which team members
specifically are involved? If there are any.
Lothar 10:20
Yeah, so character concept art is a slightly more independent job. Because you solely focus
on character in this case?
Jef Bernaers 10:30
Oh, yeah, I am, specifically.
Lothar 10:32
For character art you're gonna work with the art director and you're gonna work with the
Narrative Designer. Let's say the lore designer, and to find out the personality, the history,
the universe, the environment, the character is going to be settled in. So it's a more
independent job. Depending on first person or third person or other type of games, the
requirement is different, of course, but what I'm talking about is the first person because
that requires the most extensive development for characters, right. So, you know, and
then you need to really brief a lot, you know to define the personality and flavour for the
character. The environment is a different case. So you're gonna be working with, as I said,
Narrative Designer, lore designer, scripter, and your art director. That's about it, and also
don't forget a character also needs to go through in many companies, if it's specially for
MMO game or whatever you also need to go through a survey period. So basically, you
send out a few prototype characters for the same game with vastly different art style for
the community, so the community can vote what's their favourite style, from one to 10,
for example? Very stylized to very realistic, so that also has an impact on how character
artists could potentially define the looks.
Jef Bernaers 12:13
Do you do you also collaborate with the character artists themselves? Or do you think this
is the job of the art director to communicate with you and then communicate to separate
character artists?
Lothar 12:27
To be honest, not so much? Yeah, not so much, because normally it has to be done this
way. The character artist gets the approved concept from the art director and will start
working on it. So they don't really interfere about initial design process.
Jef Bernaers 12:53
I haven't asked but have you been involved in character art in the past?
Lothar 12:57
No I've been involved a lot in designing characters.
Jef Bernaers 13:02

68
Perfect. Then I can ask you this question as well. Can you can you describe or recall a time
when your concept art has ever negatively impacted the character artists?
Lothar 13:13
Yeah, there was of disagreement and even debate within the concept character stage of
the concept. Yes, I mean, again, depending on different types of games. So for example,
one example is a space shooting game which I happen to work with one of the UK team
and a team in Macau and some Japanese artists. So the character concept artists provide
initially a few characters is based on the suitability and adaptability and how they should
move in zero gravity space in the spaceship. So, which means tighter outfits and very
different material for different ranks and etc. So the character concept artists from the
third party they gave us a lot more flamboyant, crazy, martial art type dresses with crazy
headwear and crazy hairstyles. So we directly tell them it's not suitable for this kind of
environment and it's not possible to animate them in a realistic way, in a spaceship with
zero gravity. But the thing is, they could argue okay, we did a survey, and we know what
our players want. We know what our audience wants, they like this kind of character. So
we have to do it our way. So when you see these things, they clash. So in reality you're
trying to make something which is adapted for the gameplay and for the environment.
The concept artists for characters, they don't necessarily agree with this because they
look at things from an aesthetic point of view, instead of a practical point of view? So
things like that happen.
Jef Bernaers 15:14
Yeah. And when this happens, how do you resolve this issue? How do you go further once
you raise this.
Lothar 15:22
No, we all resigned *laughs*. No, seriously. So the whole team just said, there was no way
we can reach even a possible agreement because we have our character design and our
partner company also provides their character design, based on their fan base. So they
know what their fans want, but we think we don't really consider it from that point of
view, we want it to look more practical and so in the end, no agreement could be reached,
and no one can convince each other. So that was the case. Yeah. Also there was another
case in my second job almost 20 years ago, 19 years ago. So there was a big debate about
which part is more important for concept art: environment or character. So basically,
there was a dinosaur hunting child, so some ancient tribe tries to survive in a vast pre
historical landscape where it hunted dinosaurs and the game has got a lot of tribes from
different continents, and a lot of dinosaurs and obviously the environment is also equally
interesting. So there was also a disagreement, and naturally, there was a lot of clash
sometimes in the studio about which is more important. And yeah, the producer finally
resolved this by saying the character should be more dominant in that particular game.
And the result was a fight, the two guys fought in the studio, and then one strangled one's
neck, so that was settled. *laughs*
Jef Bernaers 17:12
Very professional *laughs*

69
Lothar 17:09
Yeah. Many years later. I still remember seeing that. But having said that, these things
happen, but also some other companies, for example, Creative Assembly. So, the one that
worked for 3 titles. I worked on a lot of characters, the Japanese warriors, and the
Napoleon's armies and stuff like that. We didn't have any problems because it's a
historical game. It's a game based on character and formations. So we don't have any
problem, but instead, we need to provide a lot of historical accurate parts and the
uniforms and the weapons and dresses, you know, for the outsourcing company to work
with. So, the more precise we provide them they can do a better job. I found out if we give
them slightly more freedom they come up with all this fantasy that never existed.
Patterns, sashes, or trims on the French or British uniforms, which was not right. So stuff
like this is sometimes you have to keep an eye on them to make sure the guidance is clear.
Jef Bernaers 18:34
Yes, so you're saying for this game specifically. It was very advantageous for you to be
very strict in what they had to do.
Lothar 18:40
Yeah, it has to be a case by case you know, you can't say all games has to be so strict, right?
Jef Bernaers 18:50
So, well I'm very interested in seeing how people perceive a concept art to be ready or at
least ready for production that is. So when you're working on concept art, how do you
determine for yourself that the concept art would be ready to ship or to start the next
stages of production with.
Lothar 19:14
Yeah, it should contain the shape, silhouette of the character, the main body type and the
profession of the character, very defined materials and textures and what the skin looks
like, what the uniform looks like with the colour palette, the model sheet and the character
sheet showing you side view, top view, front view, and, and possibly a binding pose. So
you can you can use that for character artists to sculpt. And ideally it should also contain
the lighting for the character in different scenarios within the level. So the lighting artist
could also have a rough idea of how the character should look like in a corridor or on a
desert and stuff like that. And now you also need to nail down the facial features, right?
Yeah, so people know what kind of facial type they are and etc. Yeah, so this means a
combination of fully rendered, with photo bashing or with AI in this case, or anything else,
you know. It should not be just some sketches, you know, with black and white lines.
Jef Bernaers 20:36
Do you think this is universally true for like the entire pipeline for example, from
sculpting, to the end result? Or do you think that for, for example, the sculpting stage the
first stage after concepting that they can maybe start earlier with earlier concept arts, or
do you think all of the stages have to be completed before you can continue?
Lothar 20:59

70
The sculpting should always wait until the concept is considered done. Otherwise, you're
risking wasting your time. All right. Regardless, if you're going to use sculpting or use
photogrammetry these days or use meta-human, or even using a character generator,
right. You always have to wait until it's finalized.
Jef Bernaers 21:27
So can you can you describe to me, all the stages you go through when you create concept
art for character. Stages being, line art, colour sketches, etc.?
Lothar 21:37
So first of all, you get a brief, you get a brief defined by the game. What sort of a universe
or environment are they going to live in, what sort of character he's like, a military retired
person or a banker or whatever. Everything is defined by the story of the game. So you
get a brief and after the brief you start to gather some reference, right, you know, from
movies from the existing games or from any other images you can find or even take your
own image reference and then you start to put these images together to try to find a mood
board. And from the mood board you start you know, to sketch out or to doodle the basic
shape, form, and the silhouette for the character. And then once you have defined the
overall shape and silhouette, you work into more defined body volumes. You know, and
then more details like dresses or hats or boots or whatever. Then you further define the
colour and material. Then you give it a more refined look by either photo bashing or paint
or whatever, and then further define smaller accessories right, smaller patterns or sash
or stitches or whatever. Then give it final lighting and well the functionality of course,
forgot about that. How does this character function in the game? Yeah.
Jef Bernaers 23:25
Yeah. Okay, I think those were most of my questions.
Appendix C – Raw Data
Appendix C.1 Questionnaire Questions
o What stages of concept art did you receive?
o What software did you use for the sculpt?
o How readable was the concept art on a scale of one to seven. One means it was
very hard to read even general shapes and seven means everything from the large
shapes to the details was clear.
o What were some of the biggest challenges you faced during the speed sculpting
session, and how did you overcome them?
o Was there anything missing from the concept art that you wish had been
included? If so, what was it and why do you feel it would have been helpful?
o Which aspect of the concept art did you find most useful or valuable during the
sculpting session? Why was this particular piece important to you?
o Were there any parts of the concept art that you felt were unnecessary or could
have been left out? If so, which parts and why?

71
o Do you feel that you were able to achieve the same level of quality in your
sculpting work during this session as you would on other projects with a similar
time constraint?
o How would you rate your overall efficiency in translating this concept to 3D
compared to your previous character creation experiences on a scale of one to
seven. One means I worked much slower than usual and seven means I worked
much faster than usual.
o What did you like or dislike about the concept art received? Were there any details
or aspects that you found particularly helpful or challenging?
o Do you have any other thoughts or remarks?
Appendix C.2 Interview Guide
Introduction:
Hi, thank you for the opportunity to talk to you.
Describe shortly what your research is about and how it has evolved.
Inform the user of NDA agreements. The interview can be anonymised, and I can’t
answer this question due to NDA is a valid answer.
Interview questions: (Concept artists)
General questions:
• Have the recent breakthroughs in AI tools impacted your workflow in any way?
• How do you determine the optimal amount of time and resources to devote to
concept art during the character creation pipeline?
• How do you collaborate with other team members, such as character modellers,
during the concept art stage?
Substantive research aimed questions
• Can you describe a time when your concept art negatively impacted the
efficiency of character artists. How did you resolve this issue?
• Can you describe your decision-making process for determining when a piece of
concept art is ready to be used for the next steps in the character creation
pipeline for video games?
• Do you think this is true universally for the entire pipeline? Or specific towards
individual steps?
• Can you describe to me all the stages that you go through when creating concept
art for a 3D character, stages being a lineart, colour sketches, etc. Follow up on
answers to see how they relate towards the sculpting stage.
• Has this discussion in any way influenced the way you look at the importance of
concept art?
Demographic questions:
• What is your job title?
• How many years of experience do you have creating character art?
Interview questions: (Managers)

72
General questions:
• Have the recent breakthroughs in AI tools impacted your workflow in any way?
• How do you determine the optimal amount of time and resources to devote to
concept art during the character creation pipeline?
• How do you collaborate with concept artists during their creation process?
Substantive research aimed questions
• Can you describe a time when concept art negatively impacted the efficiency of
your character artists. How did you resolve this issue?
• Can you describe your decision-making process for determining when a piece of
concept art is ready to be used for the next steps in the character creation
pipeline for video games?
• Do you think this is true universally for the entire pipeline? Or specific towards
individual steps?
• Can you describe to me the stages you think a concept artist should go through
when creating concept art for a 3D character, stages being a lineart, colour
sketches, etc?
• Which stages do you think are most important?
• Has this discussion in any way influenced the way you look at the importance of
concept art?
Demographic questions:
• What is your job title?
• How many years of experience do you have creating character art?
Interview questions: (3D character artists)
General questions:
• Have the recent breakthroughs in AI tools impacted your workflow in any way?
• What are, according to you, the absolute most important parts a character
concept art needs to have before you start to work on it?
• How do you collaborate with your concept artists and art director during the
concept art stage?
o Does this collaboration carry on during the following stages of the
pipeline?
Substantive research aimed questions
• Can you describe a time when a piece of concept art negatively impacted the
quality of your work? How did you resolve this issue?
• Can you describe your decision-making process for determining when a piece of
concept art is ready to be used for the next steps in the character creation
pipeline for video games?
• Do you think this is true universally for the entire pipeline? Or specific towards
individual steps?
• Can you describe to me all the stages that you think are crucial to have before
you start the sculpting process? Stages being lineart, colour sketches, etc.

73
• How do these individual steps relate to the sculpting stage of the character
creation pipeline?
• Has this discussion in any way influenced the way you look at the importance of
concept art?
• Demographic questions:
• What is your current job title
• How many years of experience do you have In the video games industry?
Appendix C.3 Questionnaire Data
This table shows the amount of mentions of certain topics in the questionnaire.
Concept 1
Lineart
Value sketches
Colour sketches
Character
sheet
Final
render
Readability rating
5+4+5 = 4.66
5+5+4 = 4.66
6+4 = 5
7
Efficiency rating
3+6+3 = 4
4+4+4 = 4
5+5 = 5
2
Proportions negative
1
2
Proportions positive
Angles negative
2
2
1
Angles positive
Readability negative
4
1
Readability positive
1
1
Unnecessary concept art
2
Not enough concept art
2
2
Concept difficulty
1
1
Concept without clothing
positive
1
Character sheet positive
Concept 2
Lineart
Value sketches
Colour sketches
Character
sheet
Final
render
Readability numeric
3
6
7
6+7 =
6.5
Efficiency numeric
6
6
6
6+6 = 6
Proportions negative
1
Proportions positive

74
Angles negative
1
Angles positive
Readability negative
1
Readability positive
Unnecessary concept art
1
3
Not enough concept art
1
1
Concept difficulty
Concept without clothing
positive
1
Character sheet positive
3
5
Appendix C.4 Q-Sort Data
Jitske Habekotté

75
Lothar Zhao
Radu Ianescu

76
Concept resemblance
proportions
Rafael Natali
Charles Zembillas

77
Processed Q-Sort Results
Appendix C.5 Used Concept Art
Concept one:
Created by Kelly Daley

78
Concept two:
Created by Alyssa Herman

79
Appendix C.6 Quasi Experiment results
Sketch
Value
Sketch
Colour Sketch
Character
Sheet
Final Render

80

81
Bibliography
Adobe Communications Team. (2022, March 18). Waterfall Methodology: Project Management
| Adobe Workfront. https://business.adobe.com/blog/basics/waterfall
Al-azawi, R., Ayesh, A., & Obaidy, M. Al. (2014). Towards Agent-based Agile approach for
Game Development Methodology. 2014 World Congress on Computer Applications and
Information Systems (WCCAIS), 1–6. https://doi.org/10.1109/WCCAIS.2014.6916626
Alias Systems Corporation. (1998). Autodesk Maya [C++, MEL, Python, C#]. Autodesk.
https://www.autodesk.com/products/maya/overview?term=1-
YEAR&tab=subscription
Benve, R. (2022, September 29). How Making Value Sketches Improves Your Painting Skills.
FeltMagnet. https://feltmagnet.com/painting/Thumbnails-Value-Study-for-Painting-
Composition
Braun, V., & Clarke, V. (2013). Successful Qualitative Research. SAGE Publications.
Braun, V., Clarke, V., Hayfield, N., & Terry, G. (n.d.). Thematic Analysis. Thematic Analysis.
https://www.thematicanalysis.net
Bruce, S. (2017, November 16). Character Turnarounds – Media Project 1.
https://samuelbrucemed3004m.blogs.lincoln.ac.uk/2017/11/16/character-
turnarounds/
ChatGPT. (2022). OpenAI. https://chat.openai.com/
DALL-E 2. (2021). OpenAI. https://openai.com/dall-e-2
Davis, A. (2021, November 26). 3D Character Pipeline Part II. Robodojo.
https://robonobodojo.wordpress.com/2021/11/26/3d-character-pipeline-part-ii/
Digital.ai. (2021). 15th state of agile report (Survey No. 15; State of Agile Report, p. 23).
https://digital.ai/resource-center/analyst-reports/state-of-agile-report/
Dingsøyr, T., & Moe, N. B. (2014). Towards Principles of Large-Scale Agile Development: A
Summary of the Workshop at XP2014 and a Revised Research Agenda. In T. Dingsøyr, N.
B. Moe, R. Tonelli, S. Counsell, C. Gencel, & K. Petersen (Eds.), Agile Methods. Large-Scale

82
Development, Refactoring, Testing, and Estimation (Vol. 199, pp. 1–8). Springer
International Publishing. https://doi.org/10.1007/978-3-319-14358-3_1
Gallagher, A., Dunleavy, J., & Reeves, P. (2019a, April 22). How to use Waterfall and Agile
practices on your next project—IBM Developer. IBM Developer.
https://developer.ibm.com/articles/using-waterfall-and-agile-
practices/?mhsrc=ibmsearch_a&mhq=%20waterfall
Gallagher, A., Dunleavy, J., & Reeves, P. (2019b, April 22). The Agile Method: Everything you
need to know—IBM Developer. IBM Developer.
https://developer.ibm.com/articles/agile-method-everything-you-need-to-
know/?mhsrc=ibmsearch_a&mhq=%20waterfall
Gallagher, A., Dunleavy, J., & Reeves, P. (2019c, April 22). The Waterfall Model: Advantages,
disadvantages, and when you should use it—IBM Developer. IBM Developer.
https://developer.ibm.com/articles/waterfall-model-advantages-
disadvantages/?mhsrc=ibmsearch_a&mhq=%20waterfall
Habekotté, J. (2019). Framing the Value of Concept Art [breda university of applied sciences].
https://www.researchgate.net/publication/335160773_Framing_the_Value_of_Concept
_Art_An_insight_into_the_applicability_and_development_of_concept_art_in_the_game_in
dustry
Herawati, S., Negara, Y. D. P., Febriansyah, H. F., & Fatah, D. A. (2021). Application of the
Waterfall Method on a Web-Based Job Training Management Information System at
Trunojoyo University Madura. E3S Web of Conferences, 328, 04026.
https://doi.org/10.1051/e3sconf/202132804026
Humayun, W. (2021, October 26). Research Methodology During Game Development.
ITChronicles. https://itchronicles.com/agile/research-methodology-during-game-
development/
Keith, C. (2010). Agile game development with Scrum. Addison-Wesley.
Khusid, A. (2011). Miro. https://miro.com/

83
Lankoski, P. (2015). Game Research Methods.
Leano, M. (2023, February 1). 4 Types Of Transcription: Definition and Examples | New Media
Services. NMS. https://newmediaservices.com.au/4-types-of-transcription-definition-
and-examples/
Liang, S. (2016). Otter.ai. Otter.ai. Otter.ai
LinkedIn. (2003). Microsoft Corporation. https://www.linkedin.com/
McKenzie, T. (2022, September 16). 80 Level Digest: Introduction to 3D Sculpting Tools &
Techniques. 80.Lvl. https://80.lv/articles/80-level-digest-introduction-to-3d-sculpting-
tools-techniques/
MidJourney. (2022). Midjourney Inc. https://www.midjourney.com/
Parkinson, C. N. (1958). Parkinson’s law, or, The pursuit of progress. J. Murray.
Retopology—Blender Manual. (2022, October 26). Blender 3.3 Manual.
https://docs.blender.org/manual/en/latest/modeling/meshes/retopology.html
Roosendaal, T. (2002). Blender. https://www.blender.org/
Schreier, J. (2017). Blood, Sweat, and Pixels.
Scott, D. (2018, December 3). Using Color Studies To Explore And Test Different Color
Arrangements. Draw Paint Academy. https://drawpaintacademy.com/color-studies-
guide/
Singh, B. (2017, August 31). What makes concept art useful from a 3D Character Artist’s point
of view. Linkedin.Com. https://www.linkedin.com/pulse/what-makes-concept-art-
useful-from-3d-character-artists-baj-singh/
Slick, J. (2020, September 24). The Definition of 3D Modeling. Lifewire.
https://www.lifewire.com/what-is-3d-modeling-2164
Spencer, S. (2011). ZBrush Character Creation.
Stefyn, N. (2022, September 5). How Video Games Are Made | The Game Development Process |
CG Spectrum. Cgspectrum. https://www.cgspectrum.com/blog/game-development-
process

84
Teo, L. (2014). ArtStation. Epic Games. https://www.artstation.com/
van Exel, J. (2005). Q methodology—A sneak preview.
Visual-Arts-Cork. (2019). http://www.visual-arts-cork.com/sketching.htm
Waterfall VS Agile Workflows. (2019, September 24). KolibriGames.
https://www.kolibrigames.com/blog/waterfall-vs-agile-workflows/
Whooper, A. (2021, January 19). Exploring Character Art Workflows: Skin Texturing Tips.
80.Lvl. https://80.lv/articles/exploring-character-art-workflows-skin-texturing-tips/
Wojno, R. (2018, April 22). Agile planning: A step-by-step guide and template. Monday.Com
Blog. https://monday.com/blog/project-management/agile-planning/
This template is based on a template by:
Steve Gunn (http://users.ecs.soton.ac.uk/srg/softwaretools/document/templates/)
Sunil Patel (http://www.sunilpatel.co.uk/thesis-template/)
Template license:
CC BY-NC-SA 3.0 (http://creativecommons.org/licenses/by-nc-sa/3.0/)