ajkelly-dissertation-blog

[ajkelly-dissertation-blog]

Learning experience

In completing my first dissertation, I have learned valuable new ways in which I can bring together and coherently structure a wide scope of research. I had the tendency over the first two years to think too far ahead and get anxious. This new approach took things step by step and made the whole process a lot more manageable.

In accumulating my first set of qualitative data on the scale that I managed to, I also learned a great deal. It was far more difficult and time consuming than I’d imagined but it was very enjoyable. In the future, I would definitely look much more in to best practices in analysing this kind of data. I felt a bit out of my depth and wondered whether I was doing things right or wrong at times.

I also learned that it is ok to not have any answers in the end. I spent some time trying to find perfect answers before fully starting part B but  soon realised that it would probably take a wider scope of primary research and much more of a literary review and that may not even result in any answers.

[ajkelly-dissertation-blog]

Tag Clouds and interpreting qualitative data

I went through each genre’s responses to each question on dialogue, music, sfx and what is expected from audio in guiding progression. I copied and pasted all responses for each of those questions and ran the text through a tag cloud app online to see what words were cropping up the most.

This was a very useful way of evaluating any arising themes that could then be further honed in on by analysing answer by answer.

I picked out quotes that were most represented by similar responses and analysed them against each genre to paint a better picture. It was then much clearer to see what patterns and differences were arising.

Lastly, it was a case of analysing the results against the findings from the literary review and formulating a coherent structure to best discuss this.

[ajkelly-dissertation-blog]

Analysing the data

I made a few executive decisions on how to approach picking apart both the qualitative and quantitative data collected. I used a tally and percentage system to find the base level of who my respondents were and what aspects of game audio they favoured most on a basic level.

It was interesting to find that the majority of gamers play every day and that quite a few play 4+ hours every day. This made me intrigued as to what might be said about music in games due to the sheer length of time of hearing the same game’s score.

Further, I saw that I could download the raw data so I did that and then proceeded to organise the data in to separate book sheets based on things like genre, time played, days played to see if I could find differences in opinion based on those factors.

It was found that this was useful in sorting comments by genre as this would really help in looking in to what players views on music, sfx and dialogue were genre by genre.

[ajkelly-dissertation-blog]

The results are in

Within the week my form was shared I managed to accumulate 93 responses. I was correct in assuming a heavier weighting towards more popular genres but I feel that there were more than enough respondents to find usable data among the genre by genre results. The one worry may be in the survival horror genre where only 17% of the respondents stated that they played it.

Genres in order are: action/shooter, RPG, simulation, strategy, sports, survival horror

[ajkelly-dissertation-blog]

Questionnaire dissemination

It was decided that social media would be the best way to get as many respondents as possible. I have tried to join over 30 groups of equal genre weighting but it might be difficult to get an equal outcome of players of different genres as it is clear that certain genres are much more popular.

[ajkelly-dissertation-blog]

Planning for part B

The research and writing of part A brought up questions that needed further clarification from the player’s perspective to see if their thoughts aligned with the thoughts of academics.

I opted against seeking industry insight as I specifically wanted to see what players thought at the genre level. Questions brought up from the first part were:

What roles do each aspect of game audio have in guiding progression on a genre by genre basis?

Do sound effects, dialogue and music play individual roles depending on genre?

What might be the best practices for each, if at all?

In figuring out these questions, I devised a questionnaire which can be seen below:

https://docs.google.com/forms/d/e/1FAIpQLSfzk8nNC7NhU2GARj8Z_wNz2hYNjA3qWoJMaAeQqj3JTEIyvQ/viewform?usp=sf_link

[ajkelly-dissertation-blog]

Organisation techniques

I tried a new way of preparing and organising the structure and content of a written piece of work. I first figured out what structure I wanted with headings then I made notes under each heading from my reading and research. I then used scissors to cut up and order my notes in to the most coherent layout. I then made piles of each heading, bulleted notes and worked through writing them up.

It proved really useful and is now my preferred method of working.

[ajkelly-dissertation-blog]

Books I may have overlooked

I picked up a book based on basic computer sounds for interfaces and it turned out to be of a lot of use. The author, Gaver, speaks of just how much information can be given by the smallest of sounds e.g. weight, position in space, volume, textures and probably much more [1].

It reminded me of the game Overwatch as games utilise this to help convey off-screen info when things get mad... ie. using weight, clothing, position in space etc to tell a player what enemy is approaching.

[ajkelly-dissertation-blog]

Dialogue may not be necessary

I have read that dialogue could arguably only be a need-to-know form of game audio, even though most games seem to be creating more and more complex dialogue trees and systems. Skolnick considers that giving the bare minimum adds intrigue and, arguably, more depth from a subjective perspective [1]. 

Journey is a fascinating game that uses absolutely no dialogue for the whole game. I’m not sure if this is a sandbox game or whether it can be classed as an RPG. This could be an interesting development in to whether the role of audio in games can be defined by genre or not.

1) Skolnick, E. (2006)  ...And Make It Snappy! Crafting Concise, Effective Dialogue for your Game. GDC Vault. Available from: <http://www.gdcvault.com>

[ajkelly-dissertation-blog]

Subjectivity in game audio

A study on the subjectivity of influence in game audio looked at how music and sound effects affect a user’s experience when either one is off, both are on or both are off. It shows music creates tension and sound effects create flow [1].

I’m unsure whether it depends on genre / mood and whether this holds much of an overall significance. It could be just as possible, in my opinion, to create high tension with sound effects - e.g. in a horror game.

Nevertheless, it is an area I am enjoying reading in to as I develop more breadth of understanding in game audio and genres.

1) Grimshaw, M., Nacke, L. and Lindley, C. (2010) More than a feeling: Measurement of sonic user experience and psychophysiology in a first-person shooter game. Interacting with Computers, 22 April, pp. 336-343.

[ajkelly-dissertation-blog]

Emotional appeal of games

Reading the book Media entertainment: The Psychology of its Appeal by Zillmann and Vorderer (2000) has pointed out how big a role emotion plays across all aspects of games. The link between film and games is explored and differentiated as passive and not passive in the sense of involvement. This has led me to thinking what aspects of game audio may produce the biggest emotional response and whether it directly links with immersion.

Much more reading is needed.

[ajkelly-dissertation-blog]

Game research and structure

A few games have been very interesting when looking at how sound affects guidance…

Alien: Isolation (2014) – creepy and tense music keeps you on the edge of your seat when the main alien is near

Thief (2014) – SFX is key as footsteps are of high importance in sneaking through the levels

Witcher 3 (2015) – a lot of minor detail can be useful in gauging what is happening as a reminder when hearing public chit-chatter.

It is interesting to compare this to games only 30 years ago like Tetris (1984) where only feedback bleeps existed.

The idea now is to play a lot more games and do lots of reading in to studies and published books around different areas of game audio: dialogue, SFX and music. I can then focus entirely on each area to dive deeper.

[ajkelly-dissertation-blog]

Changing strategy

I will now be undertaking a dissertation in the exact same area of interest, but in an academic sense, rather than a project. Initial idea of title:

“Investigating the role of audio as a form of guidance in video games”

I can then fully explore the academic side of the use of game audio in general. This could further expand my wider reading in to music and SFX as well as dialogue. Further, I can look at differences depending on genre and look to film and film sound as a comparison.

[ajkelly-dissertation-blog]

Facing problems

The feasibility of the project may be hindered by a few things. Life outside of Uni might get in the way more than would be necessary to undertake such a project as I would need:

1)     Many, many lines of dialogue (preferably by voice actors)

2)     Access to the labs to use UE4 for many, many hours

I am currently in talks with John Flattery over the possibility of switching to a dissertation which can be completed well from anywhere I need to be.

[ajkelly-dissertation-blog]

Planning for proposal

After tutor feedback, I learned that my project proposal hand-in could be a rough “in the works” proposal that did not have to be strictly adhered to. After realising the possible restrictions/lack of reading in the specific areas I was originally looking in to, it might be a good idea to continue forward with the aim of using dialogue as the main form of guidance.

New idea:

Solving object-based puzzles using VR ArchVis and Rift controllers in the living room level provided. Have the TV on radio channels that you can flick between in real-time. The radio channels are each different and indirectly interact and advise.

e.g. Set up a dialogue system where text clips are randomly selected top form conversations. I can probably hook up the game states to the system and trigger specific dialogue in aiding progression depending on the state of the player in the level.

Psychonauts has a similar system that isn’t attached to game state [1]. This idea might also help reduce the possibility of a user getting lost whilst trying to keep them guided [2].

1) https://www.youtube.com/watch?v=Xgp1e0JQyGI

2) https://www.youtube.com/watch?v=UBbI_JsZc-k

[ajkelly-dissertation-blog]

Possible solutions

After first hand playing of GTA and Witcher, it dawned on me that dialogue may be the most appropriate way of guiding players in a VR environment without a need for much movement.

Looking back on a talk referenced in my Game Audio module [1], I thought about the possibility of an adaptive dialogue system. I came across a paid module for Unreal Engine called VR ArchVis [2] that allows interaction with objects including full grab/throw/explore capabilities with Oculus controllers that feeds back to the engine in blueprints. It also includes a ready-made room. The idea could be that dialogue tells the player what to do with objects in front of them, but in a not so obvious way.

1) https://www.gdcvault.com/play/1021930/Dialog-Systems-in-Double-Fine

2) https://www.unrealengine.com/marketplace/archvis-vr

[ajkelly-dissertation-blog]

Brainstorming problems

Motion sickness in VR appears to be a hurdle unsolved [1]. I thought it would be good for a player to solve a succession of puzzles in different rooms where the user is static with a point-and-click type puzzle solving system. Although:

- how many ways in this set-up can you direct attention?

- really not much reading material directly addressing or looking at this specifically that isn’t vague and warranting further study [2].

Originally wanted to direct attention to very specific things like objects but I have since realised that this is a level of control you can’t expect to always have in VR. Instead, it might be best to look at directing attention towards an area in general and let the player do what they want on the way. This also maintains that agency of control and allows for better immersion. [3]

1) https://filmora.wondershare.com/virtual-reality/10-ways-to-prevent-oculus-rift-vr-motion-sickness.html

2) https://link-springer-com.ezproxy.leedsbeckett.ac.uk/article/10.1007%2Fs00371-016-1294-0

3) https://medium.com/journalism360/one-on-one-with-vr-storytellers-92d502f7780f