Introduction

What questions did you have at the start of the project?
We started with the idea to talk about school. How does school work, and could it be better? Can we enable school children to describe this institution that dominates their life and to formulate, what’s good or bad about it?

In a broader sense: Should society form our personality / individuality / body to its needs or should it be as open as possible to allow us limitless freedom?

On the dramaturgical level, we wanted to give audiences choices that mattered. We had a very open gaming experience with Expectation: First Snow and kind of “fake choice” in Mission: K.L.I.M.A.. With Pinocchio A.I., we wanted to let the audience have real choice, while still having a scripted story progression.

What technology was the focus of the work, and why?
We knew ZOOM as software pretty well by then, knew how to stream from a small stage in an office room and had edited videos for Expectation: First Snow. Our stage designer Karin Waltenberger had worked on Alienation, and experienced shooting videos and designing a streamable stage. From Alienation, we knew basically how to use OBS and inspired by A.L.I.C.E lost in Cyberland by theatre company meinhardt&Kraus, we gathered that OBS can be used in ZOOM as well and wanted to experiment with that ourselves.

We knew the pitfalls of ZOOM (especially sound quality). But in the plus side, audiences were accustomed to it. If they were not, we already knew how to explain the onboarding. And it included a polling feature we wanted to employ again, after we had an idea of what we wanted to do subject-wise.

What did you hope to explore in terms of technology, form, and content?
Technologically, we wanted to play with OBS and Zoom – all the feeds (live acting, sound, videos) should be fed in through one Zoom window. As for the form, we wanted to explore the possibilities of a branching narrative in a theatre context further, add an RPG element and game balancing. Therefore, we did not only have to think about choices and effects, but on how to measure these effects. And content-wise, we tried to find out, how a collective mind of an audience would form a personality through choices and how to make that entertaining.

Creative Process

What forms of trial and error occurred?
The most important trial-and-error was possibly the first time we presented the early idea of the game to our dramaturg Nele Neitzke. The original plan was to let the audience create a RPG-like personality, based on the “Big Five” personality traits, let it run and ultimately fail a daily routine and then to be able to intervene as “consciousness” at certain points. This proved already too complicated to explain properly. After that, we turned it around: The personality was created by the choices the audience made during the daily routine.

We then developed the game further, played it through by letting choice (dice) decide the narrative path and finally did a pen-and-paper-like presentation to a new audience, where we narrated what they would ultimately see. This gave us clues, where our story had loopholes, our point system was unfair and, most importantly, that the most enjoyable part of the game was the groups discussion which path to take.

We filmed the first two sections as improvised videos to see if the media change affected the enjoyment and tested it with the youngsters that ultimately became our video cast.

We then had test audiences in the last days, after which we shortened the beginning monologue that explained the game and changed some wordings in the choices that proved misleading.

Tech-wise the biggest trial-and-error stage was sound. To get sound from OBS Videos into Zoom we implemented NDI, but it still proved shaky. Zoom tries to format sound to a ideal speaking output, which we tried to disable as much as possible. But NDI can also amplify sound or dynamically change it, as well as OBS has its own sound settings. This balancing was the most difficult one and kept being an issue. A workable solution might have been to change from Zoom to VOD.ninja, but at that stage we could not bring audiences to another platform anymore.

Did you use existing software?
We used OBS where we fed in a live video feed from a webcam and audio from a wireless microphone, as well as 98 videos.

We used NDI as link between OBS and Zoom (Video worked directly from OBS to Zoom, but not sound).

For video cutting we used Davinci Resolve.

Images and the easter egg animation via GIMP.

Did any of your questions or goals change, over the course of this project?
We first focussed more on the RPG-character (PINs personality) than the status (PINs happiness, energy, ambition). We gave the audience more agency and the personality aspect became a nice to have conclusion. But these changes actually helped to further our goal of giving audiences choices that mattered.

What were the key milestones in the development of the production?
The change from the character-driven agency to choice and status driven agency. The first playtests with pen-and-paper and with the whole session programmed and run via Zoom for the first time.

Reflections

What was the final outcome of the project, and how does it compare to any expectations at the start?
Audiences were immersed in the story (and very angry when they reached the unhappy ending where the mother is sending PIN to their room, because they came back too late and had eaten too much cake). The characters that were formed were never complete opposites, in five categories they usually fell in two of three options, which might be due to the communal nature of the game and the goal of balancing “ambition” (there was no selfish, carefree or introverted version of PIN).

In what ways was the production a success?
We had very good feedback and there was sufficient interest in an online game back in the first season where audiences could visit the theatre without many pandemic restrictions. The people involved gained skills and experience with OBS and got more comfortable with all technology involved.

What elements of the final production would you change?
We would probably look for another streaming platform than Zoom and work early on with the sound department.

If other people were to take a similar approach, what advice would you give?
Developing a game takes always more time than expected. Make early versions (pen-and-paper-narrations), test out the technology early on. Do not use Zoom, or test out if your sound ambitions match their latest updates.

The code for this project is open-source: https://github.com/obsproject/obs-studio

https://github.com/obs-ndi/obs-ndi