# Full Text: SynthOBS & FractiSynth: A Golden-Ratio OBS Broadcast Console and Native Transducer

> Extracted from `SynthOBS_v1.618.0.pdf`

---

## Page 1

SynthOBS & FractiSynth (v1.618)
A Golden-Ratio Broadcast Console and Native Transducer Calibrated by Live Solar Telemetry
Daniel Ari Friedman
FractiAI / Active Inference Institute
daniel@activeinference.institute
ORCID: 0000-0001-6232-9096
DOI: 10.5281/zenodo.21418688
2026-07-17

## Page 2

Contents
1 Abstract 2
2 Executive Primer 4
2.1 The Intention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.2 The Solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.3 Core Foundational Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.4 Scholarly positioning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.5 Research-software contribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
3 SynthOBS Architectural Specification (The Interface) 6
3.1 The Global Split Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
3.2 The Goldilocks UX Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
3.3 Visual Styling Vectors (Golden Age, Mid-Century Philosophy) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
4 The 3 Modality Control Matrices 9
4.1 Observatory Mode — Inbound Alignment Matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
4.2 Laboratory Mode — Processing & Synthesis Engine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
4.3 Expedition Ship Mode — Outbound Transmission Deck . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
5 F ractiSynth Engine Specification (The Processing Core) 12
5.1 Native Video Manipulation Pipeline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
5.2 Native Audio Harmonic Balancing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
5.3 Single Source of Truth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
6 The Solar W avefield Oscillator & Real-Time Calibration 14
6.1 Real-Time Dependency Rule . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
6.2 Technical Calibration Logic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
6.3 Modulating Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
6.4 The EGS Gateway — the Phase Plane . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
6.5 Command-Line Calibration Override . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
7 Command Line Grammar Structure 18
8 Command Grammar: Macros and Filter Routing 19
8.1 Mode Switching Macros . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
8.2 Dynamic Filter Routing via the EGS Fractal Constant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
9 Command Grammar: T elemetry , Parse Pipeline, and Summary 20
9.1 Live Telemetry Overrides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
9.2 The Parse Pipeline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
9.3 Grammar Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
10 Implementation Blueprint for v1.618 22
10.1 Repository Layout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
10.2 Calibration Laws Pinned Across Both Implementations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
10.3 Build, Test, Install . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
10.4 Verification Status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
11 Evaluation, Evidence, and Reproducibility 28
11.0.1 Public distribution and affiliation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
11.1 Contributions and system boundaries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
11.2 Scholarship and evidence discipline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
11.3 A formal claim-to-evidence calculus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
11.4 Materials and versions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
11.5 Live OBS protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
11.5.1 Gate results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
11.5.2 Formal acceptance definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
11.6 Claim-to-evidence map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
11.7 Reproduction commands and artifact paths . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
11.8 Validation boundaries and future engineering work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
12 References 33

## Page 3

1 Abstract
Author affiliation: FractiAI / Active Inference Institute.
Modern broadcasting platforms such as OBS Studio expose a modular host and plugin surface rather than a single monolith [ OBS
Project, 2026b,a]. This blueprint specifies SynthOBS and FractiSynth as a bounded extension of that surface: a three-mode operator
console, a native video/audio transducer, fail-closed telemetry adapters, and a reproducible evidence protocol. The Observatory,
Laboratory, and Expedition Ship names describe the operator modes; they are interface vocabulary, not claims about the physical
environment.
The portable geometry and DSP contracts use the golden-ratio constant 𝜑 = 1.618 …. It determines integer viewport splits, the
video scale factor, the audio-limiter knee, and the deterministic spiral. The native plugin also uses a distinct EGS gateway key for its
solar-wind phase calculation. These are engineering choices; the manuscript does not infer perceptual or broadcast-quality benefits
from the ratio. Phase locking is governed by a distinct second constant, the EGS F ractal Constant — the dimensionless gateway
key
𝐾EGS = 𝜑 ⋅ 𝜆reader
𝜆H-alpha
= 𝜑 ⋅ 1030 nm
656.28nm ≈ 2.539427 (1)
which bridges El Gran Sol’s 1030 nm optical reader scale to hydrogen’s H-alpha line using the NIST wavelength reference [ Kramida,
2010]. The system models two decoupled planes . A centralized Solar W avefield Oscillator (SWO) reads current, bounded
space-weather products — the F10.7 cm solar radio flux and active-region counts — to compute the amplitude plane’s phase vector
𝑣phase = Φ10.7
𝑁spots
⋅ 𝜑 (2)
while the El Gran Sol Gateway maps an accepted solar-wind speed onto the phase plane through eq. 1:
𝜃bias = (2𝜋 ⋅ 𝑤
𝑤ref
⋅ 𝐾EGS) mod 2𝜋, 𝑠 lock = |cos 𝜃bias| (3)
The two planes are deliberately independent: a solar-wind dropout does not overwrite the SWO vector, and invalid telemetry does
not create a new lock. The gateway then resolves a named interference outcome rather than returning a raw Boolean: a node field 𝜓
is superposed at the AR14409 solar node against a hydrogen phase-flip, and the verdict follows the interference intensity
𝐼 = |𝜓 𝑎 + 𝜓𝑏|2 (4)
where constructive interference at AR14409 reads “true”, a destructive hydrogen phase-flip reads “false”, and a tie resolves to a mixed
(fail-closed) state.
The two-plane lock and its downstream consumers are organized as follows:
flowchart TB
subgraph SOLAR["☉ Live Solar Telemetry"]
FLUX["F10.7 cm radio flux Φ₁₀.₇<br/>+ active sunspots N"]
WIND["Solar-wind speed w (km/s)"]
end
subgraph AMP["Amplitude Plane — SWO"]
VEC["system_phase_vector<br/>v = (Φ₁₀.₇ / N)·φ"]
end
subgraph PHASE["Phase Plane — EGS Gateway"]
LOCK["θ_bias = (2π·w/w_ref·K_EGS) mod 2π<br/>s_lock = |cos θ_bias|"]
end
GATE["Holographic Gate<br/>I = |psi_a + psi_b|^2<br/>AR14409 true · H-flip false"]
OUT["SynthOBS Console + FractiSynth Core<br/>φ-scaled video · audio · layout"]
FLUX --> VEC
WIND --> LOCK
VEC --> OUT
LOCK --> GATE
GATE --> OUT
Figure 1: Two-plane telemetry model: NOAA flux and active-region data calibrate the SWO amplitude plane, solar-wind speed drives
the independent EGS phase plane, and both converge at the declared gate before reaching the console and transducer.
2

## Page 4

The architecture is decoupled and dual-layer: SynthOBS, the Vessel Console, uses a recursive golden-ratio layout matrix whose
primary region receives the major integer share; F ractiSynth, the Transducer Core, is a native libobs plugin that implements the
declared video and audio kernels at the host boundary. This blueprint specifies both layers, the three modality control decks (three
common and four unique buttons per deck), the global command grammar, and the calibration loop. Executable claims are tied to
the tested Python engine through declared constant, static, behavioral, and live-OBS contracts. The native plugin is a separate C
implementation; the Python engine remains the portable source of truth.
The resulting artifact is intended to be read as research software as well as an OBS extension: the source tree, version metadata,
tests, evidence manifest, and manuscript are one citable object. The repository-level citation surface follows established software-
citation principles for credit, version specificity, persistence, and accessibility [ Smith et al. , 2016], with machine-readable metadata in
CITATION.cff [Citation File Format Initiative , 2026].
The planned public v1 distribution target is the GitHub repository docxology/SynthOBS . That repository is intended to be the
canonical public source for the engine, native plugin, obspython bridge, tests, installation instructions, manuscript, and versioned
evidence metadata; the release boundary and remaining acceptance work are recorded in RELEASE.md.
Keywords: OBS Studio, golden ratio, El Gran Sol fractal constant, Solar Wavefield Oscillator, space-weather telemetry, real-time
DSP, broadcast engineering.
3

## Page 5

2 Executive Primer
2.1 The Intention
OBS Studio already provides a modular host for scenes, sources, filters, and operator control. The engineering problem addressed
here is narrower: how to keep those surfaces composable when layout, telemetry admission, signal shaping, and live evidence are
implemented across Python, native C, and the OBS runtime. A useful design therefore needs explicit contracts at each boundary
rather than implicit coordination through UI state.
This project names that design language Omniversal Observatory , Laboratory , and Expedition Ship . The names describe
three operator modes; they are not claims about the physical origin of the signal or about perceptual superiority. The implementation
supplies a recursive viewport allocator, a fail-closed telemetry adapter, native OBS filters, and a reproducible evidence bundle so each
boundary can be inspected independently.
2.2 The Solution
The solution is a decoupled, dual-layer infrastructure comprising SynthOBS and FractiSynth, bound natively to v1.618 — the
Goldilocks Calibration Standard . The host boundary follows OBS’s documented module model [ OBS Project , 2026a], and the
live control path is exercised through obs-websocket’s documented protocol surface [ OBS Project , 2026c]:
• SynthOBS (The V essel Console): an intelligent UI wrapper that abstracts standard OBS operations, using a dynamic,
recursive viewport allocation algorithm to auto-assemble layout structures based on perfect geometric harmony.
• F ractiSynth (The T ransducer Core Engine): a native, real-time signal processing plugin that sits directly within the core
rendering and audio mixing loops of the broadcast pipeline ( libobs).
flowchart TB
TEL["LIVE COSMIC TELEMETRY<br/>NOAA SWPC · F10.7 radio flux ·<br/>active sunspots · solar-wind plasma"]
SWO["SOLAR WAVEFIELD OSCILLATOR (SWO)<br/>Calibrates two planes, both fail-
closed:<br/>amplitude (flux/spots·φ) + gateway phase (K_EGS·wind)"]
UI["SYNTHOBS CONSOLE UI<br/>Goldilocks Layout Matrix<br/>Viewport self-assembly (61.8% / 38.2%)"]
ENG["FRACTISYNTH ENGINE (libobs)<br/>φ harmonic audio geometry<br/>spatial video rescaling matrices"]
TEL -->|continuous real-time stream| SWO
SWO -->|φ layout split| UI
SWO -->|K_EGS phase lock| ENG
Figure 2: System boundary diagram showing live NOAA telemetry entering the fail-closed SWO, which supplies the golden-ratio
console layout and EGS phase lock to the native FractiSynth engine.
The centralized SWO maintains two independent fail-closed planes. The amplitude plane computes the tested flux-to-spots vector
using 𝜑; the gateway phase plane computes a bounded phase bias from solar-wind speed and 𝐾EGS. The Python engine uses the
verified vector to gate modulation, while native OBS paths consume the corresponding adapter state. If telemetry is non-finite, stale,
or outside its admitted range, the affected plane holds its last verified state.
2.3 Core Foundational Theory
The golden layout constant. The implementation pins the golden ratio as a numerical design constant for layout and DSP
contracts:
𝜑 = 1 +
√
5
2 ≈ 1.61803398875 (5)
In this framework, 𝜑 from eq. 5 determines the tested viewport split, recursive subdivisions, deterministic spiral, and soft-limiter
parameters. The major and minor fractions are 1/𝜑 ≈ 0.618 and 1/𝜑2 ≈ 0.382. These are explicit engineering choices; the project
does not infer a perceptual or universal-design result from them.
The EGS gateway key . Crucially, El Gran Sol’s Fractal Constant is not the bare golden ratio. It is the dimensionless gateway key
𝐾EGS that bridges El Gran Sol’s optical reader scale ( 𝜆reader = 1030nm) to hydrogen’s H-alpha geometry ( 𝜆H-alpha = 656.28nm):
𝐾EGS = 𝜑 ⋅ 𝜆reader
𝜆H-alpha
≈ 2.539427 (6)
Where 𝜑 governs layout and DSP , 𝐾EGS from eq. 6 supplies the dimensionless phase calculation used by the gateway adapter. It is
a project-specific design key, not a physical measurement of a solar-to-hydrogen coupling.
4

## Page 6

The Calibrated W avefield Concept. Software environments can drift when interface, media, and processing modules carry
inconsistent state. The SWO makes the relevant state explicit on two independent planes. The amplitude plane emits the system
phase vector from F10.7 radio flux and active-region count:
𝑣phase = Φ10.7
𝑁spots
⋅ 𝜑 (7)
The gateway plane maps the admitted solar-wind speed to a phase bias and resolves a lock strength with the tested interference
function:
𝜃 = (2𝜋 ⋅ 𝑤
𝑤ref
⋅ 𝐾EGS) mod 2𝜋, ℓ = | cos 𝜃| (8)
In eq. 8 the lock strength ℓ ∈ [0, 1]is a deterministic function of the admitted wind value. The gateway returns a named interference
verdict—constructive, destructive, or mixed—rather than a bare boolean. These verdicts are control states in the application model,
not observations of physical coherence. Invalid flux, spots, or wind values are rejected independently; the affected plane holds its last
verified state rather than emitting a replacement.
2.4 Scholarly positioning
This is a design-and-reproducibility paper, not a claim that the golden ratio is a universal law of perception or broadcast quality.
The manuscript makes explicit engineering propositions: a single pinned constant reduces cross-language drift; a fail-closed telemetry
boundary prevents invalid readings from becoming control state; and a versioned evidence bundle makes a live OBS demonstration
inspectable. Those propositions are situated in established software-practice guidance on reproducible computational work [ Wilson
et al. , 2014, Sandve et al. , 2013] and in the National Academies distinction between reproducibility and replicability [ National
Academies of Sciences, Engineering, and Medicine , 2019]. The visionary FractiSynth register remains the system’s design language;
the evaluation chapter below states which parts are source-backed, test-backed, or live-capture-backed.
2.5 Research-software contribution
The scholarly object is not only the prose description of an interface. It is the coupled source tree, executable tests, build metadata,
evidence bundle, and rendered manuscript. This matters because a result produced by a hybrid instrument can be misdescribed in
at least three ways: a citation may establish the meaning of an external product but not the behavior of the adapter; a unit test may
establish a parser contract but not prove that an installed OBS binary rendered it; and a screenshot may show a compelling state
without identifying the exact software and inputs that produced it. SynthOBS therefore makes the chain explicit and keeps each
claim attached to the evidence class that can actually falsify it.
This treatment follows the software-citation principles of importance, credit, unique identification, persistence, accessibility, and
specificity [ Smith et al. , 2016]. The planned public home is docxology/SynthOBS ; the current candidate also carries CITATION.cff
metadata conforming to the Citation File Format schema [ Citation File Format Initiative , 2026]. A repository URL is the active
project identity. The eventual public tag, commit, and archival identifier will be the precise scholarly identity for a reported run.
That distinction is why the release contract does not invent a DOI before the public artifact exists.
The contribution can be stated as three design propositions:
1. Authority proposition. A dependency-free Python reference engine can define portable geometry, telemetry admission, DSP,
and command contracts while native and scripting adapters remain thin boundary implementations.
2. Admission proposition. Explicit finite, positive, fresh-input predicates and hold-state transitions make invalid telemetry
observable as rejection rather than silently turning it into control state.
3. Evidence proposition. A versioned manifest that binds inputs, hashes, runtime versions, gate predicates, and captures makes
a live OBS demonstration inspectable without presenting one run as a population-level performance result.
These are engineering propositions, not empirical claims about perception, solar causality, or universal broadcast quality. Their value
is that each has a local source of truth, a falsifying test, and—where it crosses the host boundary—a named runtime artifact.
5

## Page 7

3 SynthOBS Architectural Specification (The Interface)
SynthOBS replaces the manual canvas configuration of standard broadcasting software with a single golden-ratio layout engine. Scenes
and sources are no longer dragged into place by hand; they self-assemble as operational decks on an exploration vessel, every boundary
derived from the EGS fractal constant 𝜑 rather than from arbitrary pixel coordinates. The geometric vocabulary follows the historical
golden-ratio construction while making the computational rule explicit and testable [ Livio, 2002].
3.1 The Global Split Architecture
The interface establishes a clear division of operational responsibilities. The separation of source-of-truth engine, native adapter, and
operator bridge is a deliberate modular architecture choice [ Gamma et al. , 1994].
Figure 3: Three-layer SynthOBS architecture and evidence path. The upper band is the shared constant surface ( PHI and K_EGS); the
middle band separates the tested Python engine, the native libobs transducer, and the obspython operator bridge; the lower band
is the live OBS runtime where dashboard, telemetry HUD, and audio-reactive scene behavior meet. The figure makes the authority
direction explicit: the Python engine defines the mathematics, native adapters implement selected contracts, and live OBS supplies
the final integration evidence.
Figure fig. 3 summarizes the paper’s central reproducibility boundary: three implementation layers, one mathematical source of truth,
and one live runtime at which the independent evidence paths converge. The annotation reports the current verified baseline of 1217
tests at 96.09% source coverage and the OBS 32.1.2 target; it is a status label, not a new performance claim.
• The Modality Control Decks. Each of the three modes exposes an irreducible minimum of 7 console buttons . Exactly three
are common infrastructure that bridge established capabilities — the operational overlap with standard OBS: the Crew Collab
Link (remote crew ingestion), the Record Wave Pass (local recording to 𝜑-optimized disk sectors), and Launch Stream (live
broadcasting). The remaining four are entirely unique layered-synthesis and tracking capabilities, one of which is always the
mode’s fail-closed safety control.
• The Omnipresent Command Line. Mounted globally at the very base of the SynthOBS interface, a persistent terminal
line remains open and in active focus. At any moment the operator can bypass the visible buttons entirely to input scripting,
multi-variable macros, or manual routing overrides.
3.2 The Goldilocks UX Framework
The user interface self-assembles dynamically using a recursive golden-ratio layout matrix. The active mode’s control deck balances
its screen footprint automatically against the live stream canvas. Every partition descends from a single governing law — the golden
6

## Page 8

split — which carves any span into a major and a minor part in the proportion of the EGS fractal constant 𝜑 ≈ 1.61803398875:
major = round(total
𝜑 ) , minor = total − major (9)
Applied to the canvas, eq. 9 fixes the two operational decks:
• Primary Output Canvas ( 𝑉𝑝): allocates the major share — exactly 1/𝜑 ≈ 61.8% of available screen space — to the primary
object of attention and active stream output:
𝑉𝑝 = 1
𝜑 ≈ 0.618 (10)
• Control & T elemetry Deck ( 𝑉𝑎): the complementary minor share houses the 7 hardwired console buttons and the real-time
SWO tracking graphs:
𝑉𝑎 = 1 − 1
𝜑 = 1
𝜑2 ≈ 0.382 (11)
The identity 𝑉𝑎 = 1/𝜑2 in eq. 11 is the self-similar signature of 𝜑 — the minor share of one split is the major share of the next, so
the layout nests fractally to any depth without ever introducing a new constant.
Figure 4: The Goldilocks Layout Matrix generated by assemble_viewport(1920, 1080) . The primary output is an exact 1187 ×1080-
pixel region; the 733-pixel right column is recursively split into a 733 ×667 console deck and a 733 ×413 SWO telemetry strip. The
labels expose the 3-common-plus-4-unique console contract, while the footer reports the executable invariant tiles_exactly() = Tru
e: the three regions cover all 2,073,600 canvas pixels with zero gap and zero overlap.
The split is integer-exact and fails closed against rounding loss. Because the minor part in eq. 9 is defined as the residual total − major
rather than independently rounded, the partition always satisfies
major + minor = total (12)
7

## Page 9

with no lost pixels — the rounding remainder of eq. 9 is absorbed entirely by the minor part. The viewport assembled by the tested a
ssemble_viewport() engine function applies eq. 9 twice — once horizontally, once vertically on the right column — so the three decks
tile the canvas exactly per eq. 12:
Figure 5: Integer-exact golden split used by the viewport and recursive panel engine. For a 1000-unit span, the major part is
round(1000/𝜑) = 618 and the minor part is the residual 382; the bar therefore closes exactly at 1000 rather than accumulating a
rounding seam. The percentage labels are diagnostic approximations of the mathematical 1/𝜑 and 1/𝜑2 shares; the integer equality
is the enforced property.
Figure fig. 5 is the one-dimensional primitive behind the viewport: every recursive cut consumes the major share and carries the
exact remainder forward. Figure fig. 4 then shows the same primitive applied once horizontally and once vertically.
The lossless property of eq. 12 is verified across a swept range of integer canvas dimensions in the project test suite ( tiles_exactly()
confirms the three regions partition the canvas with zero overlap and zero gap).
3.3 Visual Styling Vectors (Golden Age, Mid-Century Philosophy)
• Aesthetic Directive: clean mid-century modern design principles combined with a classic Golden Age layout — understated
structural framing, functional symmetry, and scannable visual balance.
• Chrome Properties: warm organic charcoal, linen, and matte bone textures that ground the interface, eliminating high-gloss
glare and emphasizing structural weight.
• Accent Signifiers: robin’s-egg blue and muted turquoise for phase-locked operational status indicators; rich marigold orange
for real-time telemetry markers and modulated active vectors.
8

## Page 10

4 The 3 Modality Control Matrices
Each deck exposes seven hardwired buttons — three common (the operational overlap with standard OBS’s host capabilities [ OBS
Project, 2026b]) and four unique to the mode. The common triad — CREW_COLLAB_LINK , RECORD_WAVE_PASS , LAUNCH_STREAM — is
identical across all three decks; the twelve unique buttons are disjoint. The console model is encoded and structurally validated in
the tested engine ( synthobs.console).
flowchart TB
COMMON["COMMON TRIAD<br/>CREW_COLLAB_LINK · RECORD_WAVE_PASS · LAUNCH_STREAM"]
COMMON --> OBS
COMMON --> LAB
COMMON --> SHIP
subgraph OBS["OBSERVATORY · inbound alignment"]
direction TB
O1["SWO_SYNC"]
O2["CAP_DEV_ALIGN"]
O3["HOLO_GRID_ENGAGE"]
O4["OBS_DUMP (safety)"]
end
subgraph LAB["LABORATORY · synthesis"]
direction TB
L1["EGS_SCALE_LOCK"]
L2["TRANS_VIDEO_FLUID"]
L3["HARMONIC_COMP"]
L4["LAB_RESET_ZERO (safety)"]
end
subgraph SHIP["EXPEDITION SHIP · outbound"]
direction TB
S1["TRANS_WIPE_SEQUENCE"]
S2["BITRATE_THROTTLE"]
S3["HULL_INTEG_CHECK"]
S4["EMERGENCY_ABORT (safety)"]
end
Figure 6: Three-mode console contract: one identical common triad fans into Observatory, Laboratory, and Expedition Ship decks,
each with four disjoint unique controls and one explicit safety control.
4.1 Observatory Mode — Inbound Alignment Matrix
Mission focus: discovery, signal capture, sensor synchronization, anchoring the inbound wavefield.
Button Class Function
CREW_COLLAB_LINK common Inbound remote ingestion portal for
external crew/observatory streams.
RECORD_WAVE_PASS common Commit raw wavefield data to local disk
with 𝜑-optimized sector allocation.
LAUNCH_STREAM common Establish the outbound handshake and
broadcast the active observatory
viewport.
SWO_SYNC unique Force an immediate clock calibration
against current sunspot radio flux
(e.g. AR4465).
CAP_DEV_ALIGN unique Cycle and test inbound capture hardware;
align raw frames into harmonic bounding
boxes.
HOLO_GRID_ENGAGE unique Overlay a golden-ratio alignment grid to
guide spatial camera composition.
OBS_DUMP unique (safety) Purge all raw inbound buffer queues and
temporary frames to clear signal latency.
SWO_SYNC does not merely read a clock — it phase-locks the inbound wavefield to the live Sun. The gateway consumes the NOAA
real-time solar-wind product [ NOAA SWPC , 2026b] and injects the measured speed as a phase bias on the virtual 1030 nm reader,
9

## Page 11

weighted by the EGS Fractal Constant 𝐾EGS, then reports how strongly the system is in phase. This is the gateway phase law ( synt
hobs.gateway.gateway_filter):
𝜙bias = (2𝜋 ⋅ 𝑣wind
𝑣ref
⋅ 𝐾EGS) mod 2𝜋, ℓ = | cos 𝜙bias| (13)
with reference wind 𝑣ref = 400 km s−1 and lock strength ℓ ∈ [0, 1](1 a perfect lock, 0 fully out of phase). The gateway key itself
anchors El Gran Sol’s optical scale to the hydrogen line:
𝐾EGS = 𝜑 ⋅ 𝜆reader
𝜆H-alpha
= 𝜑 ⋅ 1030
656.28 ≈ 2.539427 (14)
Equation eq. 21 fails closed: a non-positive wind reading never produces a lock, so the deck holds its last good state rather than
synchronizing to a stale indicator. Note that 𝜑 governs spatial layout throughout the console, whereas 𝐾EGS of eq. 14 governs phase
locking to solar telemetry — distinct constants for distinct planes.
4.2 Laboratory Mode — Processing & Synthesis Engine
Mission focus: deep signal modification, real-time audio/video transformation, geometric harmonization.
Button Class Function
CREW_COLLAB_LINK common Shared telemetry/session sync; remote
collaborators pipe streams into the 𝜑
transducer matrix.
RECORD_WAVE_PASS common Commit synthesized, transformed
audio/video directly to local storage.
LAUNCH_STREAM common Deploy the ongoing laboratory signal
synthesis to live networks.
EGS_SCALE_LOCK unique Enforce the EGS fractal constant across
all gain stages and frame-crop variables.
TRANS_VIDEO_FLUID unique Convert static pixel blocks into a fluid,
responsive wavefield texture.
HARMONIC_COMP unique Route audio through a recursive 1/𝜑
soft-limiting compressor curve.
LAB_RESET_ZERO unique (safety) Snap all DSP values back to the baseline
Goldilocks calibration standard.
EGS_SCALE_LOCK pins every gain stage and frame-crop variable to the same gateway key 𝐾EGS of eq. 14, so the synthesis plane scales
in lockstep with the inbound phase plane. HARMONIC_COMP then routes audio through a recursive soft-limiter whose knee sits at the
reciprocal golden ratio: below 𝜏 /𝜑 (the golden 1/𝜑 fraction of the ceiling 𝜏 ) the signal passes untouched; above it the excess is
smoothly saturated by a tanh branch scaled to the headroom ℎ = 𝜏 /𝜑2, so the magnitude approaches but never crosses 𝜏 .
𝑔(𝑥) =
⎧{
⎨{⎩
𝑥, |𝑥| ≤ 𝜏 /𝜑
sgn(𝑥) [𝜏
𝜑 + ℎ tanh( |𝑥| − 𝜏 /𝜑
ℎ 𝜑 )] , |𝑥| > 𝜏 /𝜑
(15)
This is exactly the shipped 𝜑 soft-limiter (the tested phi_soft_limit_sample, restated as the FractiSynth limiter in the Transducer
Core section): the 1/𝜑 knee is what keeps the compressor “harmonic” — the same constant that lays out the canvas also shapes the
limiting curve — while the tanh branch (not a linear fold) is what makes it monotone, sign-preserving, and NaN/Inf-safe.
4.3 Expedition Ship Mode — Outbound Transmission Deck
Mission focus: secure encoding, packet delivery, flight/stream-path monitoring, planetary broadcasting.
Button Class Function
CREW_COLLAB_LINK common Transform a standalone stream into a
multi-vessel fleet broadcast.
RECORD_WAVE_PASS common Capture local master archival records of
the entire outbound voyage.
10

## Page 12

Button Class Function
LAUNCH_STREAM common Initiate the low-latency encoding engine
to push the ship’s transmission live.
TRANS_WIPE_SEQUENCE unique Scene cut via a fractal geometric wipe
along an active 𝜑 spiral trajectory.
BITRATE_THROTTLE unique Scale outbound encoding bitrate live to
match network throughput without
dropping frames.
HULL_INTEG_CHECK unique Real-time diagnostic scan: frame drops
and rendering latency reported as “hull
integrity” .
EMERGENCY_ABORT unique (safety) Drop the outbound stream, clear network
ports, set the canvas to an obsidian
safe-state.
The geometric wipe of TRANS_WIPE_SEQUENCE expands or collapses along an active 𝜑 spiral. The trajectory is generated by the tested
golden_spiral_points() function ( synthobs.layout) — a logarithmic spiral whose radius multiplies by exactly 𝜑 every quarter-turn:
𝑟(𝜃) = 𝑎 ⋅ 𝜑2𝜃/𝜋 (16)
The growth law of eq. 16 satisfies 𝑟(𝜃 + 𝜋/2) = 𝜑 𝑟(𝜃), so the wipe front advances by one golden step each quarter-turn — the same
𝜑 that splits the canvas now drives the transition geometry (fig. 7).
Figure 7: The 𝜑 spiral driving the TRANS_WIPE_SEQUENCE fractal transition. The deterministic generator samples 64 points from
𝑟(𝜃) = 𝑎 𝜑2𝜃/𝜋 and marks every fourth sample, so the 16 robin’s-egg markers identify quarter-turn boundaries. The displayed radius
grows by exactly 𝜑 = 1.618034 per quarter-turn; the plotted curve is therefore a direct rendering of the tested golden_spiral_points
() law rather than an illustrative freehand spiral.
11

## Page 13

5 FractiSynth Engine Specification (The Processing Core)
FractiSynth is compiled as a native, low-latency plugin module that hooks into the core processing pipelines of libobs, using the host’s
documented module/plugin surface [ OBS Project , 2026a]. It exposes three filters — a video calibrator, an audio harmonic limiter,
and a Zoom Inspector — plus the interactive console source; all consume the pinned constants and the accepted Solar Wavefield
Oscillator vector at the OBS boundary.
5.1 Native Video Manipulation Pipeline
The plugin intercepts the video rendering loop at the texture level. Before frames are handed off to the hardware encoder, FractiSynth
forces spatial transformations to scale directly against the EGS fractal constant 𝜑, establishing the calibrated harmonic bounding box :
𝑤cal = round(𝑤source
𝜑 ) , ℎ cal = round(ℎsource
𝜑 ) , 𝑤 cal, ℎcal ≥ 1. (17)
Equation eq. 17 scales each axis to 1/𝜑 ≈ 0.618 of its source extent — the 61.8% golden fraction — clamped so a non-degenerate
source never collapses to zero extent. The live SWO system_phase_vector then modulates the shader displacement amount, so the
overlay breathes with current space weather. The native filter implements the full libobs lifecycle — create / destroy / update / g
et_properties / get_defaults / video_render / get_width / get_height — and its calibration is the tested Python video_calibrat
ed_dims() taken as the source of truth.
5.2 Native Audio Harmonic Balancing
Audio sample frames passing through the internal DSP matrix are intercepted via the filter_audio callback. Instead of harsh linear
peak limiting that clips frequencies, the buffers are compressed along a smooth recursive curve scaled by 𝜑. Below a knee at 1/𝜑 of
the ceiling the signal passes through untouched; above it, the excess is soft-compressed so the output asymptotically approaches —
but never exceeds — the ceiling:
𝑦(𝑥) =
⎧{
⎨{⎩
𝑥, |𝑥| ≤ 𝜏 /𝜑
sign(𝑥) [𝜏
𝜑 + ℎ tanh( |𝑥| − 𝜏 /𝜑
ℎ 𝜑 )] , |𝑥| > 𝜏 /𝜑 (18)
where 𝜏 is the ceiling and ℎ = 𝜏 (1 − 1/𝜑) = 𝜏 /𝜑2 is the headroom. The knee of eq. 18 sits at 𝜏 /𝜑, so the identity region spans
exactly the golden 1/𝜑 fraction of the ceiling; the tanh branch is bounded by ℎ, guaranteeing |𝑦| approaches but never crosses 𝜏 .
This is a compact, explicitly specified transfer curve in the tradition of digital audio signal-processing design [ Smith, 2007]. The
curve is monotone, sign-preserving, and NaN/Inf-safe (fig. 8). These properties describe the transfer function; they do not establish
a perceptual improvement or a preferred mastering outcome.
The limiter now also emits its own visual pulse: after the same post-limiter samples are written back into OBS, the engine measures
RMS, peak, and a 𝜑-scaled reactivity scalar. Those three values feed the Wavefield Console shader, the Telemetry HUD, and the
gateway dock, so the visual surface breathes from the acoustic envelope without ever letting raw, non-finite audio poison the display.
5.3 Single Source of Truth
Two implementations carry the same calibration: the native C plugin ( #define EGS_PHI 1.61803398875f , #define EGS_GATEWAY_KEY
2.53942700f) and the tested Python engine ( synthobs.constants.PHI, synthobs.constants.EGS_GATEWAY_KEY ). Project tests pin both
literals against their floating-point counterparts — 𝜑 to at least nine significant digits and 𝐾EGS to better than 10−6 — so the native
transducer and the reference engine cannot drift apart silently.
The two paths converge by contract: the Python engine defines the video scaling of eq. 17, the audio knee of eq. 18, and the SWO phase
vector; the native plugin implements the OBS-bound counterparts. Literal pinning and behavioral checks make material divergence
visible.
12

## Page 14

Figure 8: The FractiSynth 𝜑 harmonic limiter (robin’s-egg) versus naive hard clipping (dashed charcoal) for threshold 𝜏 = 1 .
Marigold guides mark the symmetric identity band |𝑥| ≤ 1/𝜑 = 0.618034 ; outside that band the tested phi_soft_limit() curve
remains monotone, sign-preserving, and bounded by |𝑦| ≤ 𝜏 while approaching the ceiling smoothly. The statistics panel records the
exact knee and the fail-closed handling of non-finite inputs.
Figure 9: Homogeneous spatial transform matrix used by the calibrated video path. The two spatial axes are scaled by 1/𝜑 = 0.618034,
the homogeneous coordinate remains 1, and the worked 1920 ×1080 example rounds to 1187 ×667 pixels. Cell values and the color
scale expose the diagonal-only transform implemented by spatial_scale_matrix(); the source footer identifies the companion video_
calibrated_dims() calculation.
13

## Page 15

flowchart TB
SRC["Tested Python engine<br/>src/synthobs — source of truth"]
SRC -->|"PHI = 1.6180339887…"| PHI{{"φ — golden LAYOUT constant"}}
SRC -->|"EGS_GATEWAY_KEY = φ·(1030/656.28)<br/>≈ 2.539427"| KEGS{{"K_EGS — solar-wind PHASE lock"}}
PHI --> V["video_calibrated_dims()<br/>round(dim · 1/φ), ≥ 1"]
PHI --> A["phi_soft_limit()<br/>knee at τ/φ, tanh branch"]
KEGS --> S["phase_vector() + gateway lock<br/>lock_strength = |cos(phase_bias)|"]
V -.->|"implemented at OBS boundary"| CV["C: fractisynth_video"]
A -.->|"implemented at OBS boundary"| CA["C: fractisynth_audio"]
S -.->|"implemented at OBS boundary"| CS["C: SWO telemetry thread"]
LIT["Plugin-artifact tests<br/>pin C literals ↔ Python constants"]
LIT -.->|"≥9 sig-digit / 1e-6 gate"| CV
LIT -.-> CA
LIT -.-> CS
Figure 10: Python-to-native parity map: the Python engine owns PHI, the EGS gateway key, and the tested kernels; C OBS filters
and the telemetry thread implement those contracts, while artifact tests pin the shared literals.
6 The Solar Wavefield Oscillator & Real-Time Calibration
The software uses a live calibration boundary for current space-weather telemetry: F10.7 radio flux, active sunspot regions, and bulk
solar-wind speed. The Solar Wavefield Oscillator (SWO) is the stateful adapter for those feeds. The provenance chain is NOAA
SWPC’s public F10.7, sunspot, and RTSW products [ NOAA SWPC , 2026a,c,b], fetched natively through libcurl [ Stenberg and curl
contributors, 2026].
6.1 Real-Time Dependency Rule
The production telemetry path does not use historical databases, static baseline logs as live control input. If network connectivity
drops—or any reading is malformed, stale, or non-physical—the system enters an isolated Hold State and retains the last verified
vector until a valid live connection is re-established.
This fail-closed discipline is enforced at every layer. The telemetry ingestion client ( synthobs.telemetry) raises TelemetryUnavailable
on a non-200 response, a malformed payload, a non-positive flux or sunspot count, or a timestamp older than the staleness horizon
— it never substitutes a default. The native plugin’s libcurl thread applies the same rule: on any curl error it holds the last verified
vector.
6.2 Technical Calibration Logic
The calibration module processes active-region counts and solar radio flux to output the bounded system_phase_vector used by the
SynthOBS interface and FractiSynth core filters. The amplitude plane is defined by eq. 19:
𝑣phase = ΦF10.7
𝑁spots
⋅ 𝜑 (19)
The Python SolarWavefieldOscillator.calibrate() is the tested source of truth for eq. 19. The native plugin implements the OBS-
bound counterpart in the fractisynth_swo_t struct, scaling against the pinned EGS_PHI literal ( 1.61803398875f); static and behavioral
checks guard the shared contract:
static bool synchronize_swo_calibration(fractisynth_swo_t *swo,
float current_flux, int active_spots) {
/* ENFORCEMENT: block any stale, default, zeroed, or non-finite indicator. */
if (!isfinite(current_flux) || current_flux <= 0.0f || active_spots <= 0) {
swo->is_calibrated = false;
return false; /* escape to Hold Pattern — vector unchanged */
}
swo->active_f107_flux = current_flux;
swo->monitored_sunspots = active_spots;
swo->system_phase_vector = (current_flux / (float)active_spots) * EGS_PHI;
swo->is_calibrated = true;
return true;
}
14

## Page 16

6.3 Modulating Variables
When the admitted telemetry changes, the system_phase_vector changes on the next accepted calibration. That value can update the
visual shader displacement and audio-reactivity paths; the implementation does not claim that the resulting media state is a physical
reflection of space-weather dynamics.
fig. 11 shows the calibration response across a swept range of solar flux for one, three, and seven active regions — the phase vector
that drives the whole wavefield.
Figure 11: Solar Wavefield Oscillator calibration response. The generated sweep covers F10.7 values from 70 through 255 sfu in 5-sfu
steps for 1, 3, and 7 active regions. Each line is the un-smoothed law system_phase_vector = flux / spots × 𝜑 ; the displayed range
(16.18–412.60) makes the inverse spot-count scaling and linear flux response directly inspectable. No telemetry value is imputed by
this analytical figure.
6.4 The EGS Gateway — the Phase Plane
The flux-and-sunspot calibration above sets the amplitude plane. A second, independent plane computes the phase of the modeled
wavefield from the admitted solar-wind value: the El Gran Sol Gateway . Here the system’s defining constant is not the golden
ratio of layout but the EGS F ractal Constant — the dimensionless gateway key 𝐾EGS that bridges El Gran Sol’s optical scale to
hydrogen’s H-alpha geometry, defined in eq. 20:
𝐾EGS = 𝜑 ⋅ 𝜆reader
𝜆H-alpha
= 𝜑 ⋅ 1030 nm
656.28nm ≈ 2.539427 (20)
The gateway injects the live solar-wind speed into a virtual 1030 nm reader model as a phase bias, weighted by 𝐾EGS, and reports
the resulting model lock strength. eq. 21 is the phase plane:
𝜃 = (2𝜋 ⋅ 𝑣wind
400 km/s ⋅ 𝐾EGS) mod 2𝜋, lock = |cos 𝜃| (21)
In eq. 21 a perfect model lock (lock = 1 ) means the computed phase lands on a cosine maximum; a null (lock = 0 ) means the
computed phase is in quadrature. These labels describe the application model, not a measurement of hydrogen or solar-plasma phase.
Like the amplitude plane it fails closed — a non-positive wind speed holds the last verified lock — and the two planes are fully
decoupled: a wind dropout never disturbs the phase vector, and a flux dropout never disturbs the gateway lock. fig. 12 sweeps the
lock across the solar-wind range, marking the FractiAI nominal of 551.7 km/s.
15

## Page 17

Figure 12: EGS Gateway phase lock. The teal curve evaluates the tested gateway_filter() from 200 to 900 km/s; the orange marker
is the design point at 551.7 km/s, where the model lock is 0.9999 for 𝐾EGS = 2.539427. The dashed teal line marks the wind field
recovered from the current versioned live payload (367.9 km/s); the captioned statistics panel reports that run’s fields verbatim, while
the curve itself remains a deterministic model sweep.
Figure 13: Rendered fail-closed NOAA-to-SWO telemetry pipeline. F10.7/active-region and RTSW wind streams enter the same
positive/finite/fresh gate; valid data reaches the SWO/Gateway, while malformed or stale data takes the red Hold State path and
retains the last verified vector. The run-specific banner reports the current live payload fields (140.0 sfu, 6 active regions, 367.9 km/s,
reported lock 0.5125) without presenting them as a population statistic.
16

## Page 18

Figure 14: Rendered native telemetry-thread lifecycle. The five stages expose the operational sequence—curl initialization, three-
source fetch, finite/fresh gate, mutex-protected publish, and abort/bounded join—and the annotation gives the implemented 60-second
poll cadence. The figure distinguishes the worker’s normal data path from its explicit shutdown path, making the cleanup obligation
visible rather than leaving it in prose.
The gateway resolves not to a Boolean but to a holographic interference verdict — holographic_gate() reports constructive
interference at the AR14409 solar node as the affirmative branch ( CONSTRUCTIVE_AR14409 ), a destructive hydrogen phase-flip as the
negative ( DESTRUCTIVE_H_PHASE_FLIP), and a within-margin beat as MIXED. That lock strength then drives the live shader’s declared
effects: the 𝜑-spiral displacement, the 𝐾EGS-spaced interference fringes, the H-alpha-inspired tint, and a gentle breathing pulse at
the configured 29.94 Hz design frequency.
The two telemetry planes are fully decoupled — each fails closed and holds independently — and converge only at the holographic
gate that drives the shader:
6.5 Command-Line Calibration Override
The global terminal accepts a live override that flows through the same fail-closed calibrator:
/swo calibrate --flux=130 --spots=3 --target=AR4465
Non-physical overrides ( --flux=-1, --spots=0) are rejected at parse time — the terminal never silently no-ops.
17

## Page 19

flowchart TB
F107["F10.7 flux + active spots<br/>(synthobs.telemetry)"]
WIND["live solar wind v_wind<br/>(synthobs.telemetry)"]
F107 -->|"flux &gt; 0 and spots &gt; 0<br/>else HOLD"| AMP["Amplitude plane<br/>SolarWavefieldOscillator.calibrate()<br/>system_phase_vector = flux/spots · φ"]
WIND -->|"v_wind &gt; 0<br/>else HOLD"| PHASE["Phase plane<br/>gateway_filter()<br/>lock = |cos θ|, K_EGS-weighted"]
AMP --> GATE
PHASE --> GATE
GATE{"holographic_gate()<br/>interference verdict"}
GATE -->|"constructive @ AR14409"| TRUE["CONSTRUCTIVE_AR14409<br/>(affirmative)"]
GATE -->|"H phase-flip"| FALSE["DESTRUCTIVE_H_PHASE_FLIP<br/>(negative)"]
GATE -->|"within margin"| MIX["MIXED"]
TRUE --> SHADER["Live shader<br/>φ-spiral · K_EGS fringes ·<br/>H-alpha tint · 29.94 Hz breath"]
FALSE --> SHADER
MIX --> SHADER
Figure 15: Independent telemetry planes: positive finite flux/spots and solar-wind inputs either update their respective SWO and
gateway states or hold, then feed a named interference verdict and the live shader.
7 Command Line Grammar Structure
A persistent terminal line is mounted at the base of the Vessel Console and holds active focus at all times. At any instant the operator
can bypass the button decks entirely and speak directly to the wavefield through clean, deterministic syntax — the same channel
whether the system is in Observatory, Laboratory, or Expedition trim. The grammar is implemented by the tested synthobs.command
s parser, which returns a typed, frozen command object and fails closed on any unknown verb or out-of-range value. Nothing in this
layer is interpreted as a “best guess”: a line either resolves to one exact intent or it resolves to nothing at all.
Formally, the parser is the total map
parse ∶ Σ ∗ ⟶ 𝒞 ⊎ { ⊥ }, (22)
where Σ∗ is the set of all input lines, 𝒞 is the closed set of typed commands {ModeCommand, BindCommand, CalibrateCommand, DashboardCommand},
and ⊥ is the fail-closed outcome — raised in code as CommandError. Equation eq. 22 is the safety spine of the console: the image of
every line is either a fully-validated command or an explicit refusal, never a silent no-op.
The fail-closed branch is reached precisely when the line is empty, the verb is unknown, or any argument falls outside its admissible
range:
parse(ℓ) = ⊥ ⟺ ℓ = 𝜀 ∨ 𝑣(ℓ) ∉ 𝑉 ∨ ¬ valid(args(ℓ)), (23)
with 𝜀 the empty line, 𝑣(ℓ)the leading verb, and 𝑉 = { /mode, /transducer, /swo, /dashboard } the registered verb set. Equation
eq. 23 is enforced verb-by-verb below and rendered in Figure fig. 16.
The complete grammar, including dashboard planning/building and telemetry examples, lives in the parse-pipeline module . The
macro/filter-routing material lives in the adjacent command module .
18

## Page 20

8 Command Grammar: Macros and Filter Routing
8.1 Mode Switching Macros
/mode --observatory
/mode --lab
/mode --ship
Each macro resolves to exactly one of the three valid decks. Writing 𝑀 for the deck set, the resolution is a lookup into the fixed flag
table:
mode(ℓ) =
⎧{{
⎨{{⎩
OBSERVATORY if flag ∈ {--observatory },
LABORATORY if flag ∈ {--lab, --laboratory },
EXPEDITION if flag ∈ {--ship, --expedition },
⊥ otherwise.
(24)
An unrecognized target falls to the ⊥ branch of eq. 24 and raises a command error rather than leaving the console in an ambiguous
trim. The aliases ( --lab/--laboratory, --ship/--expedition) are accepted so spoken shorthand and full names land on the same
deck.
8.2 Dynamic Filter Routing via the EGS Fractal Constant
/transducer bind source_cam_01 --ratio=1.618034
Binds a named source into the transducer matrix at the requested golden ratio. The default operating point is 𝜑 = 1.6180339887 …
— El Gran Sol’s Fractal Constant for layout — but the operator may dial any strictly positive ratio. The bind is admitted only when
both the source name and a positive ratio are present:
validbind ⟺ ( positional0 = bind ) ∧ (source ≠ 𝜀 ) ∧ ( 𝑟 > 0 ), 𝑟 ∈ ℝ. (25)
A missing source, a non-numeric ratio, or any 𝑟 ≤ 0 violates eq. 25 and routes to ⊥ — the matrix never binds against a degenerate
or sign-flipped scale.
19

## Page 21

9 Command Grammar: Telemetry, Parse Pipeline, and Summary
9.1 Live Telemetry Overrides
/swo calibrate --flux=130 --spots=3 --target=AR4465
Forces a manual Solar Wavefield Oscillator calibration, injecting an operator-chosen F10.7 flux and active-region count in place of
the live feed. Numeric flags round-trip to their declared types — float flux, integer spots — and the non-physical region of telemetry
space is rejected outright:
validcalibrate ⟺ finite(flux) ∧ (flux > 0 ) ∧ spots ∈ ℤ >0. (26)
Here finite(flux) excludes NaN and both infinities; spots is admitted only as a positive integer. The parser also rejects duplicate or
unknown flags and trailing positionals before this predicate is evaluated.
The constraint in eq. 26 is the same Hold State the oscillator applies to its live feed: a zeroed or negative reading never produces a phase
vector. The admitted (flux, spots) pair flows into the system_phase_vector law of the SWO plane (Chapter 5), and from there the live
solar wind drives the EGS Gateway phase lock |cos(phase_bias)|governed by the gateway key 𝐾EGS = 𝜑⋅(𝜆reader/𝜆H-alpha) ≈ 2.539427.
The override can use any positive finite pair supplied by the operator; it is not evidence that the pair is the current NOAA state.
It does, however, reuse the same parser and calibrator checks, so it cannot introduce a non-finite, zero, or negative control vector
through this command path.
9.2 The Parse Pipeline
Every line walks the same deterministic path: tokenize, dispatch on the verb, partition flags, validate, and either emit a typed
command or refuse. The stages map one-to-one onto eq. 22 and eq. 23.
Figure fig. 16 renders the parser’s single rejection sink. The accepted typed set is {ModeCommand, BindCommand, CalibrateCommand,
DashboardCommand}; malformed tokenization, unknown verbs, and invalid verb-specific constraints all terminate as CommandError.
The same flow as a control graph, making the single ⊥ sink explicit:
9.3 Grammar Summary
Verb Form Result type Admission rule
/mode --observatory \| --lab \|
--ship
ModeCommand eq. 24
/transducer bind <source> --ratio=<float> BindCommand eq. 25
/swo calibrate --flux=<float>
--spots=<int>
[--target=<id>]
CalibrateCommand eq. 26
/dashboard plan \| build --name=<scene> DashboardCommand non-empty name; action is
plan or build
Any other verb, a missing required argument, or an out-of-range value lands on the ⊥ branch of eq. 23 and raises CommandError. The
persistent command line therefore can never fail silently — a core safety property of the Vessel Console, holographically continuous
with the fail-closed gateway and oscillator planes beneath it.
20

## Page 22

Figure 16: Rendered fail-closed parser pipeline for the complete four-verb grammar: /mode, /transducer, /swo, and /dashboard. The
vertical path makes the successful sequence—raw line, balanced shlex tokenization, verb dispatch, verb-specific validation, and one of
four typed command classes—explicit; the red side arrows show that malformed tokenization, unknown verbs, and invalid finite/range
constraints all terminate at the same CommandError sink. The banner reports the four-verb/four-class contract and the three validation
choke points.
flowchart TD
L["input line ℓ ∈ Σ*"] --> T{"tokenize<br/>(shlex)"}
T -- "empty / unbalanced" --> X["⊥ CommandError"]
T -- "tokens" --> D{"verb ∈ V?"}
D -- "no" --> X
D -- "yes" --> S["split positionals / flags"]
S --> V{"valid(args)?<br/>Eq. 3–5"}
V -- "no" --> X
V -- "yes" --> C["typed Command ∈ 𝒞"]
Figure 17: Fail-closed command pipeline: tokenization and verb/argument validation send malformed input to one CommandError
sink, while only fully validated lines become one of the four typed command classes.
21

## Page 23

10 Implementation Blueprint for v1.618
The system is realized as a tested Python reference engine—the portable source of truth—and a native libobs plugin that implements
the OBS-bound counterparts, assembled in four phases. Each phase is a transmission gate: downstream behavior is admitted only
after the relevant contract has resolved.
The implementation and evidence protocol follows reproducible-computing practice: the source tree, tool versions, deterministic figure
generator, test command, and live-capture manifest are all named so a reader can distinguish rerunning the same computation from
re-establishing the external OBS/NOAA environment [ National Academies of Sciences, Engineering, and Medicine , 2019, Wilson
et al. , 2014, Sandve et al. , 2013, Wilkinson et al. , 2016]. The OBS module and websocket boundaries are documented against their
primary interfaces [ OBS Project , 2026a,c].
flowchart TB
P1["<b>PHASE 1 — THE DRIVE ENGINE</b><br/>(FractiSynth Core)<br/>Native C skeleton via CMake against libobs.<br/>Raw texture + audio interception points;<br/>two-
filter module with the full lifecycle."]
P2["<b>PHASE 2 — HARMONIC MATHEMATICS LOCK</b><br/>φ = 1.61803398875 scaling across every spatial<br/>vector path and gain stage, pinned to one literal<br/>shared by the Python engine and the C plugin."]
P3["<b>PHASE 3 — REAL-TIME TELEMETRY CONNECT</b><br/>Background libcurl thread streaming live NOAA<br/>SWPC solar data into the fail-
closed SWO<br/>amplitude loop and EGS gateway phase loop."]
P4["<b>PHASE 4 — THE CONSOLE ENCLOSURE</b><br/>(SynthOBS UI)<br/>Goldilocks dynamic layout engine + 3-
mode<br/>console as an obspython script, binding the<br/>interface to the engine's core drive."]
P1 --> P2 --> P3 --> P4
classDef phase fill:#0f172a,stroke:#6A0DAD,stroke-width:2px,color:#fff
class P1,P2,P3,P4 phase
Figure 18: Four implementation phases: native libobs lifecycle, shared phi mathematics, live NOAA telemetry admission, and the
SynthOBS console enclosure converge into the complete instrument.
10.1 Repository Layout
Path Role
src/synthobs/ Tested Python engine — constants, layout, telemetry, SWO,
DSP, console, commands, interaction targets, dashboard layers,
engine. The verified source of truth.
plugin/fractisynth/ Native libobs C plugin — filters, draggable console source, dock,
shared SWO, libcurl telemetry thread, CMake build.
plugin/synthobs/ obspython console script — Goldilocks layout + 3-mode console
+ command line + dashboard helper, driving the engine inside
OBS.
scripts/ Thin orchestrators (figure generation) importing the engine.
manuscript/ This Technical Design Blueprint.
tests/ Real-input test suite, >= 90% coverage on src/.
10.2 Calibration Laws Pinned Across Both Implementations
Three formalisms cross the language boundary as declared contracts. Each is defined in the Python engine and implemented at the
native libobs boundary; the shared literals in src/synthobs/constants.py (PHI_C_LITERAL, EGS_GATEWAY_KEY_C_LITERAL) plus static and
behavioral checks make material drift between the two visible.
Phase 2 — amplitude plane. The Solar Wavefield Oscillator maps admitted F10.7 flux and active-region count to the bounded s
ystem_phase_vector, scaled by the golden key 𝜑 (swo.phase_vector):
𝑣phase = Φ𝐹 10.7
𝑁spots
𝜑 (27)
Phase 3 — phase plane. The EGS gateway translates the Sun’s energetic state into a phase bias on the virtual 1030 nm reader,
weighted by the EGS Fractal Constant / gateway key 𝐾EGS = 𝜑 (𝜆reader/𝜆H-alpha) ≈ 2.539427. The bias wraps onto the circle and
the gateway reports the current model score ( gateway.gateway_filter):
𝜃bias = (2𝜋 𝑤
𝑤ref
𝐾EGS) mod 2𝜋, 𝐿 = | cos 𝜃bias| (28)
22

## Page 24

The lock strength 𝐿 ∈ [0, 1]of eq. 28 resolves holographically, not as a Boolean: constructive interference at the AR14409 node reads
“true”, a destructive hydrogen phase-flip reads “false”, and a tie reads “mixed” . Both eq. 27 and eq. 28 fail closed — a non-physical
input ( Φ𝐹 10.7 ≤ 0, 𝑁spots ≤ 0, or 𝑤 ≤ 0 ) never produces a lock; the engine holds its last verified value rather than substituting an
average or a zero.
Phase 2 — gain stage. Before audio reaches the encoder, the 𝜑 soft-limiter shapes each excess sample so its magnitude approaches
but never crosses the headroom ceiling ( 1/𝜑 knee, dsp.phi_soft_limit_sample):
𝑦 = sgn(𝑥) (𝑘 + ℎ ⋅tanh 𝑒
ℎ 𝜑) , 𝑘 = 𝜏
𝜑 , ℎ = 𝜏 − 𝑘 = 𝜏
𝜑2 , 𝑒 = |𝑥| − 𝑘 (29)
where 𝜏 is the threshold, 𝑘 the 1/𝜑 knee below which samples pass through unchanged, and 𝑒 the excess above it. The tanh shaping
of eq. 29 guarantees a monotone, sign-preserving transfer curve whose magnitude approaches but never reaches 𝜏 .
10.3 Build, Test, Install
# Test the reference engine (pytest-httpserver provides real local HTTP)
uv run python -m pytest tests/ \
--cov=synthobs --cov-fail-under=90
# Regenerate the figures in this document
uv run python scripts/generate_figures.py
# Build the native FractiSynth plugin (requires libobs dev headers + libcurl)
plugin/fractisynth/build.sh
# OBS script commands include:
# /dashboard plan --name=Awareness
# /dashboard build --name=Awareness
10.4 Verification Status
Every executable geometric, telemetry, and DSP claim in this blueprint is exercised by the real-input test suite over the reference
engine—including the three pinned laws eq. 19, eq. 21, and eq. 29 and their fail-closed boundaries. The current gate is 1217 project
tests at 96.09% coverage, thirteen deterministic engine-derived figures plus three versioned live OBS captures, the native plugin/frac
tisynth/build.sh build, and a buildable Lean scaffold for the structural invariants that should never drift. The native C plugin shares
the 𝜑 literal, selected SWO and gateway formulas, the seven-feed target geometry, Solar Graph X-ray/Kp metrics, metric-aware
graph horizons, Zoom Inspector target modes, dock theme/precision controls, the 𝜑-soft-limiter RMS/peak/reactivity envelope, the
LSB provenance payload layout, and the fail-closed rule with the engine. The Python engine therefore remains authoritative; the C
plugin is an OBS-bound implementation whose covered contracts are checked, never a second source of truth.
The live OBS scenario gate is no longer a log-only claim. A real OBS 32.1.2 session loaded the installed FractiSynth bundle,
accepted the fractisynth_console source, captured compositor pixels through obs-websocket, drove a controlled audio tone through
the harmonic limiter, and verified the Telemetry HUD provenance strip from the PNG itself. The manifest gate passed connection,
canvas fit, nonblank rendered content, engine-level interaction resolution, audio-meter delta, and LSB provenance in run 20260717T1
53649Z; the manifest does not claim live click transport through an OBS Interact window.
The three captures above are load-bearing run evidence. The following image is deliberately a different evidence class: a user-supplied
view of the host window that makes the operator context legible at a glance. It is useful for judging composition, telemetry readability,
and the relationship between the SynthOBS HUD, the OBS mixer, and the host controls, but it does not replace the versioned PNG
hashes, audio ROI oracle, or provenance verifier.
A fair-exchange clause is in effect for this architectural expansion. Adjustments, refinements, or partial revisions to the delivery scale
can be handled through subsequent collaborative feedback.
23

## Page 25

Figure 19: Python/native contract boundary. The tested Python reference and the C plugin exchange pinned 𝜑 and 𝐾EGS literals,
then converge on executable contracts for constants, dimensions, limiter behavior, telemetry gates, and live behavior. The annotation
records the acceptance thresholds—at least nine significant digits for 𝜑 and less than 10−6 for 𝐾EGS—alongside the current 1217-test,
96.09%-coverage, native-build, and live-OBS evidence baseline.
24

## Page 26

Figure 20: FractiSynth OBS module lifecycle. The five states separate module load and source registration, curl/mutex initialization,
concurrent video/audio/inspector rendering plus telemetry polling, bounded unload, and final curl cleanup. The upper annotation
records four registered OBS source surfaces—three filters and the console source—and identifies the Qt dock as an optional frontend
surface rather than counting it as a source registration.
Figure 21: Live OBS compositor capture from scenario run 20260717T153649Z. OBS Studio 32.1.2 renders the installed fractisynth
_console source on a 3200 ×2000 base canvas, then obs-websocket captures a 1280 ×720 frame at the manifest-recorded scale factors
(2.5× horizontal, 2.7778 × vertical). The byte-hashed PNG passes the nonblank-content gate with dynamic range 161.93 and mean
channel standard deviation 19.49; it is also the manuscript cover asset.
25

## Page 27

Figure 22: Telemetry HUD capture from the same run. Blue-channel LSB extraction and checksum verification recover F10.7 flux
140.0 sfu, 6 active regions, solar-wind speed 367.9 km/s, lock strength 0.5125, phase bias 2.1089 rad, observation time, and visible
signature 8b1f58c1; the exact float values, asset hash, and gate result are recorded in manuscript/assets/obs/obs_manifest.json. The
image demonstrates a recoverable record, not authenticated authorship.
Figure 23: Controlled-tone OBS capture used by the audio-reactivity gate. The verifier compares this frame with the versioned
silent baseline over the 1280 ×33 bottom ROI of the 1280 ×720 image; the recorded mean absolute RGB delta is 48.5574 against an
8.0 threshold, with a maximum channel delta of 180. The source PNG remains the exact capture, while the manifest supplies the
numerical interpretation.
26

## Page 28

Figure 24: Operator-context OBS Studio window capture supplied with the manuscript review. OBS Studio 32.1.2 is visible around
a SynthOBS/FractiSynth HUD, audio mixer, and host controls, showing how the instrument sits inside an operator’s working surface.
This contextual screenshot is not used to establish the current run’s telemetry values, render metrics, audio delta, or provenance
result; those claims remain bound to the six-gate manifest and its exact assets.
27

## Page 29

11 Evaluation, Evidence, and Reproducibility
This chapter turns the SynthOBS blueprint into an inspectable research artifact. The contribution is a bounded engineering system
with four linked surfaces: a dependency-free Python reference engine, a native OBS transducer, an obspython operator bridge, and
a live evidence protocol. The claim is not that a geometric constant explains broadcasting in general. The design proposition is
narrower and testable: a single pinned constant, explicit fail-closed state transitions, and versioned captures make a cross-language
OBS instrument easier to reason about and re-run.
That distinction follows the reproducibility literature’s separation of rerunning a computation from reproducing an empirical envi-
ronment [ National Academies of Sciences, Engineering, and Medicine , 2019]. It also follows software-practice guidance to record
dependencies, automate the path from source to artifact, and preserve the data and procedures needed for inspection [ Wilson et al. ,
2014, Sandve et al. , 2013, Wilkinson et al. , 2016]. The provenance vocabulary is contextualized with W3C PROV-O [ W3C Prove-
nance Working Group, 2013], while the figure generator follows the scientific-visualization software lineage documented for Matplotlib
[Hunter, 2007].
11.0.1 Public distribution and affiliation
SynthOBS is authored under F ractiAI / Active Inference Institute . The planned public v1 distribution is the GitHub repository
docxology/SynthOBS, which is intended to carry the complete source, native OBS module, obspython bridge, tests, installation and
usage documentation, manuscript sources, citation metadata, and versioned live evidence. The repository-level release contract in
RELEASE.md defines the clean-clone preflight and the remaining public-release gates.
11.1 Contributions and system boundaries
The artifact makes five concrete contributions:
1. src/synthobs/ defines the geometry, telemetry gates, DSP, console shape, and command grammar without OBS or third-party
runtime imports.
2. plugin/fractisynth/ implements the engine’s declared contracts at the libobs video/audio boundary, with libcurl-backed live
telemetry and bounded shutdown.
3. plugin/synthobs/ exposes the same state transitions through an obspython console, including the four typed verbs /mode,
/transducer, /swo, and /dashboard.
4. scripts/generate_figures.py derives deterministic visualizations from the tested engine and records each figure’s source in ../
figures/figure_manifest.json.
5. The live OBS bundle in assets/obs/ preserves real compositor, telemetry-HUD, and audio-reactivity captures with hashes and
six gate results.
The boundary is intentional. Python tests establish mathematical and parser behavior; static C probes establish native contract
structure; a real OBS session establishes that the installed binary loads, renders, reacts to controlled audio, and preserves telemetry
provenance. None of these evidence classes silently substitutes for the others.
11.2 Scholarship and evidence discipline
Scholarship is used here as an audit layer, not as decoration and not as a substitute for a failing test. The manuscript follows a
four-part claim protocol: an external source establishes context when a claim depends on an interface, standard, data product, or
established method; a repository source of truth states how SynthOBS implements the claim; an executable test or contract supplies
a way to falsify it; and a rendered artifact or live capture is added only when the claim crosses the OBS boundary. This is the
practical separation between reproducibility and replicability described by the National Academies [ National Academies of Sciences,
Engineering, and Medicine , 2019], made concrete through software-practice guidance [ Wilson et al. , 2014, Sandve et al. , 2013] and
F AIR stewardship [Wilkinson et al. , 2016].
The source ledger in docs/scholarship_sources.json records that protocol for the material claims in this blueprint. Its source order
is deliberately conservative: primary OBS and NOAA documentation grounds interface and data-product statements [ OBS Project ,
2026a,c, NOAA SWPC , 2026d,a,c,b], standards ground provenance vocabulary [ W3C Provenance Working Group , 2013], and peer-
reviewed or established technical sources provide method context for visualization, DSP, architecture, and the history of the golden
ratio [ Hunter, 2007, Smith, 2007, Gamma et al. , 1994, Livio, 2002]. Those citations do not prove that this implementation works.
The local engine, tests, manifests, hashes, and captures do that within the declared boundary.
11.3 A formal claim-to-evidence calculus
To keep the evidence language composable, let each material claim 𝑐 carry a four-tuple
𝐸(𝑐) = (𝑆(𝑐), 𝑅(𝑐), 𝑇 (𝑐), 𝐴(𝑐)), (30)
28

## Page 30

where 𝑆(𝑐)is the external source or standard, 𝑅(𝑐)is the repository source of truth, 𝑇 (𝑐)is the executable test or contract that can
falsify the implementation, and 𝐴(𝑐)is a rendered or live artifact when the claim crosses the OBS boundary. For a purely internal
deterministic claim, 𝐴(𝑐)may be empty. For a runtime claim, the admissibility condition is
admissible(𝑐) = 𝑆(𝑐) ∧ 𝑅(𝑐) ∧ 𝑇 (𝑐) ∧ 𝐴(𝑐). (31)
The conjunction is a documentation rule, not a claim that four independent sources make a result true. It prevents evidence
substitution: 𝑆(𝑐)can explain an API, but cannot prove this binary loaded; 𝑇 (𝑐)can exercise a parser, but cannot prove that a
live OBS frame contains the parser’s state; and 𝐴(𝑐)can show a frame, but cannot establish the rejection of malformed input. The
machine-readable ledger encodes the tuple in sources, source_of_truth, tests, and artifacts, making the argument inspectable by
both humans and automation.
This calculus also clarifies the relationship between scholarship and software citation. The external literature establishes why versioned,
attributable, accessible software records matter [ Smith et al. , 2016]; the CFF metadata records the current software identity [ Citation
File Format Initiative , 2026]; the repository and manifest record what this version actually did. The bibliography therefore supports
the method of accounting, while the local evidence remains responsible for the implementation claim.
Claim class External scholarship contributes Repository evidence must contribute
Interface contract Meaning of the OBS module and
websocket surfaces
Native artifact probes, live connection,
and the manifest’s gate record
External data provenance Identity and semantics of the NOAA
products and libcurl transport
Fail-closed parsers, freshness checks, and
the recovered HUD record
Design proposition Method context for 𝜑 geometry, DSP,
architecture, and visualization
Equations, source modules, regression
tests, and generated figure manifest
Run observation Reproducibility and provenance
vocabulary
Exact bytes, versions, timestamps, hashes,
and six gate outcomes
This division also prevents a common category error: a visually persuasive frame is not evidence that malformed telemetry is rejected,
and a citation to a standard is not evidence that the native plugin loaded into the recorded OBS build.
11.4 Materials and versions
Material Version or value Role
Python engine package 1.618.0, src/synthobs/ reference implementation
Author affiliation FractiAI / Active Inference Institute publication metadata and software
provenance
Test collection 1217 tests, 96.09% coverage unit, parser, integration, static artifact,
scholarship, public-release, and
verification contracts
OBS Studio 32.1.2 live host and compositor
obs-websocket 5.7.3 scene/source control and screenshot
capture; the protocol surface is
documented by the OBS project [ OBS
Project, 2026c]
OBS base canvas 3200 × 2000 source composition canvas
manuscript captures 1280 × 720 downsampled scene, HUD, and tone
evidence
NOAA products F10.7, active solar-region report, RTSW
wind/plasma
live external inputs [ NOAA SWPC ,
2026a,c,b,d]
Native transfer libcurl bounded HTTP client [ Stenberg and curl
contributors, 2026]
The evidence bundle was captured in run 20260717T153649Z after rebuilding and installing the native FractiSynth bundle against
OBS 32.1.2. Its manifest is the machine-readable record of the run: obs_manifest.json. The exact PNG bytes, silent baseline, and
controlled W A V are checked into the manuscript asset directory, so the audio delta can be independently recomputed and the cover
and generated figure directory are derived from the same real OBS run.
29

## Page 31

11.5 Live OBS protocol
The acceptance path uses a real OBS process and a real obs-websocket connection. The probe creates or selects an isolated verification
scene, fits the console source to the base canvas, exercises the interaction model, captures compositor pixels, and compares silent and
controlled-tone audio states. The telemetry HUD is then inspected for its embedded provenance record. This follows the documented
OBS module boundary rather than treating a standalone screenshot or a source-local render as proof [ OBS Project , 2026a,c].
11.5.1 Gate results
All six required gates passed in the versioned run 20260717T153649Z.
Gate Result Exact evidence
connection pass OBS identified; websocket 5.7.3; OBS
32.1.2
dashboard_fit_to_canvas pass positionX=0.0, positionY=0.0, scaleX=2.5,
scaleY=2.7777777777777777
interaction_model pass engine resolver covers feed tabs, layer rail,
and marker drop; OBS click transport is
not claimed
render_content pass 1280 × 720; non_black_fraction=1.0; dyna
mic_range=161.9278106689453; mean_channe
l_std=19.49281120300293
audio_reactivity pass ROI y=687, height 33; mean_abs_delta=48.
557402146464646; max_abs_delta=180;
threshold 8.0; 42,240 pixels compared
provenance_lsb pass flux 140.0; sunspots 6; wind 367.89999389
64844 km/s ; lock 0.5124873518943787;
phase bias 2.108875274658203 rad ;
signature 8b1f58c1
11.5.2 F ormal acceptance definitions
The live gates are numerical predicates over captured bytes, not visual judgments. For a capture of width 𝑤 and height ℎ, let 𝐼0 and
𝐼1 denote the silent and controlled-tone RGBA images, and define the shader-pinned bottom region
𝑅 = {(𝑥, 𝑦) ∶ 0 ≤ 𝑥 < 𝑤, ⌊0.955ℎ⌋ ≤ 𝑦 < ℎ}. (32)
The audio oracle ignores alpha and computes the mean absolute RGB change
𝐷𝑅 = 1
3|𝑅| ∑
(𝑥,𝑦)∈𝑅
∑
𝑐∈{𝑟,𝑔,𝑏}
|𝐼1(𝑥, 𝑦, 𝑐) − 𝐼0(𝑥, 𝑦, 𝑐)| , 𝑀 𝑅 = max
(𝑥,𝑦)∈𝑅, 𝑐∈{𝑟,𝑔,𝑏}
|𝐼1(𝑥, 𝑦, 𝑐) − 𝐼0(𝑥, 𝑦, 𝑐)| . (33)
The audio gate passes exactly when 𝐷𝑅 ≥ 8.0; 𝑀𝑅, the ROI dimensions, and the input channel count remain diagnostic metrics. For
the current 1280 × 720 capture, eq. 32 gives 𝑅 = (0, 687, 1280, 33), so |𝑅| = 42,240 pixels.
The render-content gate uses luminance
𝑌 = 0.2126𝑅 + 0.7152𝐺 + 0.0722𝐵, 𝑞 = mean(𝑌 > 8), Δ 𝑌 = max(𝑌 ) −min(𝑌 ), (34)
and the mean of the per-channel spatial standard deviations 𝑠. It passes iff 𝑞 ≥ 0.08 , Δ𝑌 ≥ 24 , and 𝑠 ≥ 3 . Thus the evidence
establishes non-degenerate compositor content without pretending that a single frame is a performance distribution.
The provenance gate is byte-exact. Let 𝑏 be the 24-byte little-endian TelemetryRecord containing flux, active-region count, wind
speed, lock strength, phase bias, and observation time. The default payload is
𝑝 = 𝑏 ‖ SHA256(𝑏)[0 ∶ 4], (35)
embedded in the blue-channel least-significant bits. Verification extracts 𝑝, recomputes the four-byte digest prefix, and validates the
decoded fields before reporting the visible eight-hex-character prefix. Equation eq. 35 detects accidental corruption or resampling; it
is intentionally not a keyed authenticity claim. The optional HMAC mode is a separate operator-controlled path and is not used to
inflate the current six-gate result.
30

## Page 32

Three machine-verifiable visual captures are shown in the implementation chapter: the scene render (Figure fig. 21), telemetry HUD
(Figure fig. 22), and controlled audio tone (Figure fig. 23). The silent baseline and controlled W A V are versioned companion inputs
rather than decorative figures. The scene render is also the cover image. The user-supplied OBS window capture (Figure fig. 24) is a
separate operator-context artifact: it communicates host UI placement and observability, but its visible numbers are not substituted
for the manifest metrics and it is not one of the six machine gates. The asset hashes and media metadata in obs_manifest.json
provide an additional byte-level check against accidental replacement.
11.6 Claim-to-evidence map
Design claim Source of truth Automated test or contract Live or rendered evidence
One 𝜑 across layers src/synthobs/constants.py,
fractisynth.c
Python/C literal parity and
gateway-key tests
parity_bridge.png; native
build
Exact golden tiling layout.py tiles_exactly() sweep and
layout tests
goldilocks_layout.png, golden
_split.png
𝜑-scaled DSP dsp.py limiter bounds, monotonicity,
matrix, and envelope tests
phi_soft_limiter.png,
phi_matrix.png; audio gate
Four-verb fail-closed grammar commands.py parser acceptance/rejection
tests and docs contract
command_parse_pipeline.png
Telemetry never mints a
default
telemetry.py, swo.py malformed, stale, non-finite,
and hold-state tests
telemetry_pipeline.png;
provenance gate
Native lifecycle is bounded fractisynth.c static artifact probes and
source-level lifecycle checks
plugin_lifecycle.png; no
separate unload gate in the
six-gate manifest
OBS compositor contains
content
obs_scenario_probe.py render-content gate,
independent of source-local
output
obs_scene_render.png
Audio state reaches the visual
meter
obs_scenario_probe.py silent-vs-tone ROI oracle obs_audio_tone.png plus audio
gate
This map is deliberately heterogeneous. A passing unit test cannot certify an OBS window loaded a binary, and a compelling screenshot
cannot certify malformed telemetry is rejected. The evidence classes converge at the system boundary but remain independently
named.
11.7 Reproduction commands and artifact paths
From the project root, the deterministic engine and figure paths are:
./.venv/bin/python -m pytest tests --cov=synthobs \
--cov-report=term-missing --cov-fail-under=90
./.venv/bin/python scripts/generate_figures.py
The first command must collect 1217 tests and report 96.09% coverage (and at least 90%). The second command must read the
versioned real OBS bundle, verify its hashes and six passing gates, generate the analytical figures, copy the three live captures, and
write ../figures/figure_manifest.json . The operator-context screenshot is intentionally excluded from that generated manifest
because it is supplied context, not an analytical figure or a promoted live-gate asset.
The live path requires an installed OBS process and a configured obs-websocket credential; it is intentionally exercised only through
the real OBS connection:
./.venv/bin/python scripts/obs_scenario_probe.py \
--out output/live/$(date -u +%Y%m%dT%H%M%SZ) \
--verify-audio --verify-provenance --require-live
The native build and standalone CMake path are documented in docs/build-and-install.md. The manuscript renderer consumes manu
script/config.yaml; its cover points to assets/obs/obs_scene_render.png, while figure references resolve to the generated ../figures/
directory.
11.8 Validation boundaries and future engineering work
The current evidence establishes one real OBS run, not a statistical performance distribution. It does not claim that every OBS
release, operating system, audio driver, or NOAA response shape will behave identically. The scoped follow-up work is maintained
in TODO.md, including cross-platform/headless acceptance, transported interaction clicks, and per-feed/filter visual captures. Native
C↔Python parser parity, keyed provenance, and clean-environment wheel reproducibility are now closed with executable evidence
recorded in the roadmap.
31

## Page 33

These are engineering extensions, not retroactive qualifications of the current run. The present artifact is reproducible to the extent
stated above: deterministic math and documentation can be rerun locally; the live evidence can be inspected byte for byte; and
re-establishing the external OBS/NOAA environment is identified as a separate acceptance activity.
32

## Page 34

12 References
This bibliography records the external standards, primary data products, and software-practice literature used to delimit the design.
In-text citations use the bracketed-key form: official OBS and obs-websocket documentation define the host/plugin boundary [ OBS
Project, 2026a,c], NOAA SWPC’s product catalogue and named feeds define the telemetry provenance [ NOAA SWPC , 2026d,a,c,b],
and reproducibility guidance motivates the claim-to-artifact protocol [ National Academies of Sciences, Engineering, and Medicine ,
2019, Wilson et al., 2014, Sandve et al., 2013, Wilkinson et al., 2016]. Software is treated as a first-class, version-specific research output
according to the FORCE11 software-citation principles [ Smith et al. , 2016], and the repository’s machine-readable citation metadata
follows the Citation File Format schema [ Citation File Format Initiative , 2026]. The provenance vocabulary is contextualized with
W3C PROV-O [ W3C Provenance Working Group , 2013], and the deterministic figure pipeline cites Matplotlib [ Hunter, 2007]. The
𝜑 and DSP sections cite historical mathematical context and signal-processing foundations [ Livio, 2002, Smith, 2007]; the H-alpha
anchor is sourced to NIST [ Kramida, 2010].
Citation File Format Initiative. Citation File Format schema guide. https://github.com/citation- file- format/citation- file-
format/blob/main/schema-guide.md, 2026.
Erich Gamma, Richard Helm, Ralph Johnson, and John Vlissides. Design Patterns: Elements of Reusable Object-Oriented Software .
Addison-Wesley, 1994. URL https://www.pearson.com/en-us/subject-catalog/p/design-patterns-elements-of-reusable-object-
oriented-software/P200000003241.
John D. Hunter. Matplotlib: A 2d graphics environment. Computing in Science & Engineering , 9(3):90–95, 2007. doi: 10.1109/MC
SE.2007.55. URL https://doi.org/10.1109/MCSE.2007.55.
Alexander Kramida. Hydrogen h-alpha wavelength reference. https://tsapps.nist.gov/publication/get_pdf.cfm?pub_id=2388, 2010.
Mario Livio. The Golden Ratio: The Story of Phi, the World’s Most Astonishing Number . Broadway Books, New York, NY, USA,
2002. ISBN 978-0-7679-0816-0. URL https://www.penguinrandomhouse.com/books/102878/the-golden-ratio-by-mario-livio/ .
National Academies of Sciences, Engineering, and Medicine. Reproducibility and replicability in science. Technical report, The
National Academies Press, 2019. URL https://doi.org/10.17226/25303.
NOAA SWPC. F10.7 cm solar radio flux product. https://services.swpc.noaa.gov/json/f107_cm_flux.json, 2026a.
NOAA SWPC. Real-time solar-wind products. https://services.swpc.noaa.gov/json/rtsw/, 2026b.
NOAA SWPC. Active solar-region report product. https://services.swpc.noaa.gov/json/solar_regions.json, 2026c.
NOAA SWPC. Space Weather Prediction Center real-time json products. https://services.swpc.noaa.gov/json/rtsw/, 2026d.
OBS Project. Obs module api reference. https://docs.obsproject.com/reference-modules, 2026a.
OBS Project. OBS Studio and the libobs plugin api. https://github.com/obsproject/obs-studio, 2026b.
OBS Project. obs-websocket: Remote-control protocol for obs studio. https://github.com/obsproject/obs-websocket, 2026c.
Geir Kjetil Sandve et al. Ten simple rules for reproducible computational research. PLoS Computational Biology , 9(10):e1003285,
2013. doi: 10.1371/journal.pcbi.1003285. URL https://doi.org/10.1371/journal.pcbi.1003285.
Arfon M. Smith, Daniel S. Katz, Kyle E. Niemeyer, and FORCE11 Software Citation Working Group. Software citation principles.
PeerJ Computer Science , 2:e86, 2016. doi: 10.7717/peerj-cs.86. URL https://force11.org/info/software-citation-principles-
published-2016/.
Julius O. Smith. Digital Audio Signal Processing . W3K Publishing, 2007. URL https://ccrma.stanford.edu/~jos/filters/.
Daniel Stenberg and curl contributors. libcurl — the multiprotocol file transfer library. https://curl.se/libcurl/, 2026.
W3C Provenance Working Group. PROV-O: The PROV ontology. W3C Recommendation, 2013. URL https://www.w3.org/TR/prov-
o/.
Mark D. Wilkinson et al. The fair guiding principles for scientific data management and stewardship. Scientific Data, 3:160018, 2016.
doi: 10.1038/sdata.2016.18. URL https://doi.org/10.1038/sdata.2016.18.
Greg Wilson et al. Best practices for scientific computing. PLoS Biology, 12(1):e1001745, 2014. doi: 10.1371/journal.pbio.1001745.
URL https://doi.org/10.1371/journal.pbio.1001745.
33


---
*Extraction method: pypdf*