– ofBuffer: http://www.openframeworks.cc/documentation/utils/ofBuffer.html#getText
– c++ vector: http://www.cplusplus.com/reference/vector/vector/
– multidimensional vectors: http://www.cplusplus.com/forum/general/833/
Thesis II – Colleen Macklin & Barbara Morris
Production/Testing Status Update – 3/14/13
Prototyping and Production
I like to think of my project as consisting of 4 primary categories of work: Narrative, Art, Coding, and Logistics. For the sake of clarity I’ll discuss my production in terms of these classifications of work. So far most of my work this semester has been geared towards Logistics and Narrative. Logistically, I’ve been done a good deal of project planning and research into where rob3115 will end up at the end of the semester. I also recently created a thesis blog where I have started to document the background work, research, production, and testing of my thesis to date.
Because my story consists of an interactive narrative with branching story paths, it is necessary for me to, for the most part, finalize it before I can begin to finalize the back-end algorithms, or Coding, of the system, as well as the complimentary art. With that said, I’ve done some foundational coding that will serve as a blank template to insert the interactive narrative once it has been finalized. As for the art, it will be the last thing to add to the work. The story, plot, and structure are of a higher priority to the project than the quality of the artwork. Therefore, most of the Art that I have done so far has been conceptual – character design, and aesthetic style research. Below I’ve detailed more specifically the production I have done for the various categories of work.
– Thesis Website: https://rob3115.wordpress.com/
– Production Schedule
– Reached out to voice actor, James Bennett, II: If I have time at the end of the process, I’m going to have James do voice recordings for the protagonist in the story.
– Acquired 3 different Neurosky EEG devices: these will be used for testing and final presentation of project. I will test to see which device is most conducive for the experience of the project, but make the application work with all 3 devices so that the final project is more easily accessible for anyone with a NeuroSky device. I will then put the application on the NeuroSky store.
– Interactive narrative structure design: I’ve spent a good deal of time designing the system of the interactive narrative and how it will relate to the content of the story.
– Wrote the story: Chapters 1, 2a, 2b, 2c, and summaries of chapter 3
– Interactive Narrative Prototype in Twine: Twine is a free application used for easily creating and testing interactive narratives. I put the draft of my story into Twine and then exported to a web application so that I could have people experience and test the story with different pages, with a similar flow to the final work.
– Working with writer, Morgan Fletcher: we’ve met periodically and we have an ongoing Google doc with the story and our edits/comments.
- To refine plot arch of story
- Language of the character, mannerisms of a robot vs. a human
– System Architecture Plan: before I began coding I spent a good deal of time drafting the system architecture on paper to ensure that the application would be as malleable as possible, allowing for easy fixes if the story were to change last minute. I also want the code to be easily understandable and reusable for future projects by me or other artists.
– Coded a foundational template application in openFrameworks. What it has currently:
- EEG connection to all 3 NeuroSky devices working.
- Simple display for visualizing the attention level of the user
- Empty classes for the Story which will contain and present the content of the work, and a Class to control the relationship between the Story and the user’s brain activity.
– Character Design: I’ve done some character concept design sessions. I’m trying to work out the final aesthetic of rob3115 before I begin doing any of the final art. I’d like to have this worked out by spring break.
– Aesthetic Style: Additionally, I’ve been researching and testing what I want the final visual aesthetic of the illustration in the piece to be. Consistency and originality are prime considerations. Additionally, I want the aesthetic to add to the experience in an elegant and subtle way.
Testing, Feedback, and Lessons Learned
Once again I think it’s best to split the testing into the various categories of work. However, for testing it’s best to look at my playtests in the following categories: Narrative, Code, Art, and Overall Experience and Goal.
– Early on before I had the Twine web app set up, I had people read the paper version of the story to get feedback about the monologue style of the piece and whether or not the first chapter was catchy. I really want to the first chapter to hook people, driving their attention upwards. Most of the feedback I got was that they liked the story but wanted me to play around more with the language and mannerisms of the robot, making him seem more robotic.
– Once I got the Twine web app up and running I had put the next draft of the story online as a web application and had a few people test it. From these tests I got mixed feedback. Almost everyone said that I should put let text on each slide, because they immediately lost focus and desire to continue reading when they saw a big block of text. Additionally, I have received feedback about the arch of the story and how to make the variation between the different story paths more meaningful. Colleen and others have suggested that I need to make the story lines more variable to show greater feedback. Additionally, I’ve been told from numerous testers that the influence of the brain activity should be apparent. In other words I need to create a better feedback system.
– I’ve also been working with a writer, Morgan Fletcher, whom I consider to be one of my best play testers for the narrative. I’ve worked with her before and she gives great feedback about my narrative choices: plot, language, nuances, etc.
– Before jumping into the development of my system, I brought in the draft of the layout of the system for the application and talked about it with Ramiro and Jane, two people whose coding skills I respect. They gave me good feedback about the structure of the system and advice as to how to make the system as adaptable as possible for a piece with variable story and art moving forward.
– I’ve shown concept art of the character, rob3115, to other students as well as talked about my ideas for the visual aesthetic. Most of the time the conversation evolves into, “how would a robot see the world?” This has made me consider different visual aesthetics. At current I think I’m going to do the initial artwork in black and white with charcoal, and then distort it with filters that make it seem digital. I like the idea of having something digital with organic undertones. I think it will match the content of the piece well.
Overall Experience & Goal
My recent tests have involved testers reading the Twine web app draft while I record their brain activity on a separate computer with my template application. This has been really helpful because the experience for the user is much closer to what it will be like with the final version. The testers are always very interested to see their results after they’ve experienced the work. From this, I’ve really been sold that feedback, and lots of it, is very important to the resonance of my work. It will be important for me to give feedback during the story so people are aware of the level of the metric from which I am controlling the story. At this point, I don’t know how subtle or abstract I want that feedback to be.
How would a robot see the world?
I haven’t written the alternative endings yet. And even if I had I probably wouldn’t give them to you.
Play Testers of Thesis Draft (while recording brain activity)
So I had three people test the above web-based draft of the work while I recorded their attention levels using a NeuroSky Mindband. I had each tester read the story through once while their brain activity was being recorded, and then I had them read the story a second time while recording the same parameter. In each of the three tests, the attention level was more variable during the first session. Additionally in each participant’s first session, their attention level was higher in the first half of the story as opposed to the second half.
The feedback that I got from each reader was that they seemed to lose focus with the story during the pages with longer passages. They wanted the feeling of progression without having to do as much work. The tests are below:
This paper examines the foundations and challenges of AI: http://www.rr.cs.cmu.edu/aaai.pdf
I’m using it as a reference to edit the language of Rob3115 in the narrative. I want him to have robotic mannerisms, especially at the start of the story. His speech may evolve as his mental state changes over the course of the story, depending on the reader’s ability to drive his level consciousness upward (with higher EEG attention lvl).