The KVANT 1993 Archive: experiment plan and interim results

What this project is

The KVANT 1993 Archive is a research prototype on the border of film and play. The setting: an abandoned research laboratory from an imagined 1993, explored by the participant in first person. Technically it is a custom raycaster in the spirit of Wolfenstein 3D (320×200, software rendering, C++/SDL2) with a narrative system: text scenes, choices with consequences, film inserts, and procedural audio.

The point of the prototype is not to be a game but to be an instrument: it turns directorial parameters of interactive audiovisual storytelling (image density, fog, the pause of participation, montage, sound) into controllable variables that can be consciously switched on and off within the same space. Direction of participation thus becomes comparable and measurable — the practical foundation of a doctoral study in audiovisual art (field 021).

For the reader, in short: this page is a status report (September 2026) — what has been built, which interim results the synthetic phase has produced, and why every hypothesis remains honestly unverified until the human phase. How the method itself works — the practice, the variables, the hypothesis statuses, the honest limits — is unpacked step by step in a separate [methodological guide](/research/practice-based-research-audiovisual-art): this page answers “what has been done”, the guide answers “how it is done”.

Fig. 1. Technical foundation: C++ software raycaster, SDL2, internal 320×200 frame.

How the study is organised

The study follows practice-based design: every question gets its own contrast build of the prototype, frame dumps, automated tests, and an observation protocol. The method as a whole — the practice, the hypothesis statuses, the honest limits — is unpacked in a dedicated [methodological guide](/research/practice-based-research-audiovisual-art). Three directions cover the key directorial instruments of interactive storytelling:

Observation runs on two layers: synthetic observers (blind, counterbalanced sessions — for fast coverage) and human playtests with a ready-made package (for verification — currently deferred). All hypotheses honestly remain 🟡 until the human phase: synthetic data is never presented as verification.

Fig. 2. Exploration — choice or film insert — return: the rhythm of participation.

Interim results (September 2026)

1. A working laboratory. Artifact v0.4+: a 16×16 zoned space, boot/title screens, three terminal endings (TECH / CRITICAL / META), FMV-lite with voice, procedural audio with three BGM loops. All 7 automated tests pass.

2. Ascesis is comparable — and works in motion. HA-2/HA-4 dumps capture the objective difference between modes in a single frame; walkthrough sessions revealed the "a step teaches" mechanism: fog works as a depth cue that reveals itself through movement.

3. Without participation, the narrative reads as broken. The "flat" agency contrast produces the same verdict from every observer: a world without consequential pauses looks broken, not merely "calmer".

4. A measuring instrument now exists. The MV-1…MV-4 methodology (scales, a behavioural channel, pre-registered rules) passed synthetic saturation with 24 personas: the "base vs flat" distributions do not overlap; framing supplies the register of presentation; memory beats the frame — the authentic 1993 reads as a "modern filter" (thesis T7).

5. Theory and art studies. Theses T1–T7 were formulated (registers of presentation, the consequential pause, motor episteme, retroactive montage, distributed interpretation, presentation validity, prior inversion) together with an art-studies layer: formal analysis, the time of the work, archive poetics (Foster, Derrida, Boym) — all integrated into the working drafts of three papers.

What comes next: an academic publication

By the author's decision of 2026-09-14 the artifact stage is paused: the collected data is sufficient to move on to analysis and writing the research paper. The plan: consolidate the data from SYN sessions, MV saturation, think-alouds and walk logs; complete the paper on the basis of the three working texts (Papers A/B/C) and the art-studies layer; submit to a journal. After publication the project returns to the human phase: the HB-1 playtest package is already prepared, with moving FMV and the HC-4 "seamless mode" next in the queue.

The full working material — prototype, session journals, methodologies and paper drafts — is open in the repository: https://github.com/poberailo-alex/Research-01. The interactive web portal of the artifact (browser simulator, 16×16 world inspector, paper reader, media archive): https://archive-kvant-1993-web.vercel.app