TTS-AGI/raw-voice-annotation-data

Raw Voice Annotation Data

A curated dataset of 9,491 audio samples across 58 voice dimensions, each rated on a 0-6 scale (or categorical for Language/Accent). Every sample has been validated through a multi-stage pipeline: FAISS candidate mining from SAE features, LLM relevance filtering, and audio-based confirmation with Gemini and Qwen3-Omni.

Dataset Structure

voice_annotation_data/
  {DIM}/
    {VALUE}/
      sample_00.mp3
      sample_01.mp3
      ...

• Format: MP3, mono, 64 kbps, 44.1 kHz
• Max samples per bucket: 25
• Total dimensions: 58
• Total samples: 9,491

Quick Statistics

Metric	Value
Total dimensions	58
Numeric dimensions (0-6 scale)	57
Categorical dimensions	1
Total buckets	400
Total audio samples	9,491
Buckets with 25 samples	364
Buckets with < 25 samples	36

Temporal Dimensions

Some dimensions describe static properties of the voice (e.g., pitch range, gender, warmth), while others explicitly measure temporal changes within the audio clip. Understanding this distinction is critical for feature extraction and model design.

Trajectory Dimensions (measure directional change over the clip)

These dimensions track a shift or trajectory from the beginning to the end of the clip. Their value describes the direction and magnitude of change, not a static property.

Code	Name	What it tracks
VFLX	Velocity Flux	Measures acceleration/deceleration trend across the clip
DARC	Dynamic Arc	Measures loudness trajectory from start to finish
ARSH	Arousal Shift	Measures change in physiological activation over the clip
VALS	Valence Shift	Measures change in emotional polarity across the segment

Temporal Pattern Dimensions (require temporal extent to evaluate)

These dimensions measure properties that unfold over time -- rhythmic patterns, pause structures, onset characteristics, or fluctuation rates. While they don't measure a directional shift, they cannot be assessed from a single instantaneous snapshot.

Code	Name	What it measures
TEMP	Tempo	Measures speed of word/syllable production over time
CHNK	Chunking	Measures breath-group frequency and pause patterns
SMTH	Smoothness	Measures timing regularity and transition fluidity
EMPH	Emphasis	Measures dynamic stress patterns across words
DFLU	Disfluency	Measures density of fillers/hesitations over the utterance
STRU	Structure	Measures macro-level organization of ideas across the clip
VOLT	Volatility	Measures how much the emotional state shifts within the clip
COGL	Cognitive Load	Measures real-time cognitive processing friction during speech
RESP	Respiration	Measures breathing frequency and laboriousness over the utterance
ATCK	Attack	Measures onset characteristics as sound initiates from silence

Static Dimensions

All remaining dimensions describe properties that can be assessed from the overall quality of the voice without tracking a temporal trajectory: spectral characteristics (brightness, warmth, roughness), speaker identity (age, gender, register), emotional state (valence, arousal), resonance placement, speaking style, and recording quality.

Dimension Categories

Rhythm & Timing

Code	Name	Samples
TEMP	Tempo	166
CHNK	Chunking	159
SMTH	Smoothness	175
CLRT	Articulation Clarity	175
RANG	Pitch Range	150
EMPH	Emphasis	175
DFLU	Disfluency	175
STRU	Structure	152

Social & Interpersonal

Code	Name	Samples
STNC	Stance	175
FOCS	Focus	175
VULN	Vulnerability	175

Speaker Identity

Code	Name	Samples
GEND	Perceived Gender	175
AGEV	Voice Age	175
REGS	Register	175

Emotion & Affect

Code	Name	Samples
VALN	Valence	175
AROU	Arousal	175
VOLT	Volatility	175

Physical Production

Code	Name	Samples
RESP	Respiration	175
TENS	Tension	157
COGL	Cognitive Load	175
ATCK	Attack	175

Spectral & Timbral

Code	Name	Samples
BRGT	Brightness	145
ROUG	Roughness	144
HARM	Harmonicity	155
FULL	Fullness	131
WARM	Warmth	140
METL	Metallic Character	167
ESTH	Esthetics	168

Temporal Dynamics

Code	Name	Samples
VFLX	Velocity Flux	175
DARC	Dynamic Arc	175
ARSH	Arousal Shift	175
VALS	Valence Shift	175

Language & Recording

Code	Name	Samples
LANG	Language	100
ACNT	Accent	--
RCQL	Recording Quality	175
BKGN	Background Noise	175
EXPL	Explicitness	175

Resonance Placement

Code	Name	Samples
R_CHST	Chest Resonance	165
R_THRT	Throat Resonance	175
R_ORAL	Oral Resonance	152
R_MASK	Mask Resonance	157
R_NASL	Nasal Resonance	104
R_HEAD	Head Resonance	165
R_MIXD	Mixed Resonance	150

Speaking Style

Code	Name	Samples
S_CASU	Casual Style	141
S_CONV	Conversational Style	159
S_FORM	Formal Style	167
S_DRAM	Dramatic Style	175
S_NARR	Narrator Style	154
S_NEWS	Newsreader Style	165
S_TECH	Teacher/Didactic Style	147
S_AUTH	Authoritative Style	163
S_PLAY	Playful Style	175
S_CART	Cartoonish Style	175
S_ASMR	ASMR Style	158
S_WHIS	Whisper-Talk Style	175
S_RANT	Ranting/Angry Style	175
S_STRY	Storytelling Style	160
S_MONO	Monologue Style	175

---

Full Dimension Reference

Below is a comprehensive description of every dimension and every bucket value.

TEMP -- Tempo [TEMPORAL PATTERN]

Tempo measures the mechanical speed of word and syllable production, representing the linguistic density per unit of time. It ignores emotional valence or pauses, focusing strictly on how fast the vocal apparatus is moving through phonemes.

Samples: 166 total (0=24, 1=25, 2=25, 3=25, 4=25, 5=25, 6=17)

Value	Description
0	The speech is glacially slow, with syllables stretched to their absolute breaking point to create a near-static acoustic environment. Each isolated word floats in heavy silence, lacking standard linguistic momentum. The fundamental velocity of the vocal apparatus is labored and drawn out, making the linguistic content nearly unrecognizable.
1	The word production is unusually deliberate and labored, characterized by excessive time taken to resolve every single phoneme. The pace feels incredibly heavy and sedate, creating a dragging, low-density acoustic stream with greatly elongated vowel durations. It sounds like someone reading a solemn decree or struggling through deep lethargy.
2	The vocal delivery is relaxed and unhurried, moving at a pace slightly below the conversational average but maintaining functional momentum. Syllables are articulated comfortably without any sense of compression, rush, or artificial urgency. The overall acoustic speed is comparable to a gentle storyteller or a patient teacher explaining a concept.
3	The audio features a standard, unremarkable human speech tempo typical of an everyday, efficient conversation. The linguistic density and syllable rate are perfectly balanced, creating a comfortable and natural acoustic flow. There is no perceived stretching or squeezing of the phonemes.
4	The speech has a brisk, elevated momentum that suggests high engagement and a healthy forward conversational push. Syllables are delivered with slight forward energy without sounding panicked or artificially rushed. The linguistic stream moves swiftly and efficiently, keeping the listener highly engaged.
5	Words and syllables are noticeably compressed, running into each other to create a fast, high-density stream of acoustic information. The speaker is vibrating with rapid physical energy, forcing the vocal apparatus to move at a highly accelerated velocity. This rapidly trims all excess duration from the vowels, making the listener work hard to track the phonemes.
6	The tempo reaches the absolute human limit of linguistic speed, resulting in a blistering, hyper-accelerated wall of sound. Phonation is extremely dense, causing individual word boundaries to completely blur into a continuous, high-velocity acoustic vibration. The delivery sounds like a professional auctioneer or an individual in a state of high-speed panic.

CHNK -- Chunking [TEMPORAL PATTERN]

Chunking describes how the speaker groups words into breath units and the frequency of pauses between those groups. It reveals the speaker's respiratory management and the cognitive architecture of their thought process.

Samples: 159 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=9)

Value	Description
0	The speech is severely fragmented into single syllables or isolated words, broken apart by massive, jarring gaps of silence. It sounds like a malfunctioning vocal mechanism or someone physically struggling to piece together a single thought. There is absolutely no continuous respiratory flow or cohesive acoustic phrasing present.
1	The audio consists of very short, choppy two-word bursts that create a percussive, stop-and-go rhythmic structure. Every minor thought is separated by distinct, unnecessary pauses that completely interrupt the natural breathing cycle. The overall acoustic effect is highly disjointed and lacks any continuous flow.
2	Breath units are consistently shorter than natural phrasing, characterized by cautious, highly separated groupings of words. The speaker frequently pauses after small clauses, breaking the sentence into small, easily digestible acoustic blocks. This results in a pedantic, heavily segmented delivery style that feels artificially paced.
3	The speech is grouped into natural, medium-length sentence-sized clauses that align perfectly with standard human breath cycles. Pauses occur naturally at grammatical boundaries, creating an easy-to-follow, balanced conversational rhythm. The respiratory management and acoustic spacing feel completely unremarkable and organically structured.
4	The phrasing is noticeably extended, often linking multiple sentences together into a single, continuous, sweeping breath unit. The speaker utilizes high lung capacity to bridge normal stopping points, creating a smooth, long-range acoustic flow. Pauses are infrequent but deliberate, supporting a large-scale structural delivery.
5	The audio features very long, dense streams of words poured out with only the briefest of snatched inhalations. The linguistic density between pauses is incredibly high, driven by an urgent need to finish large thoughts without resting. This creates a relentless, unbroken wave of speech that stretches the limits of normal breath support.
6	The speaker produces a massive, continuous wall of words that lasts the entire duration of the audio clip without a single pause or breath. It is a relentless, unrelenting flood of dense linguistic output that offers absolutely no acoustic gaps for rest or processing. The listener is overwhelmed by the sheer, uninterrupted continuity of the vocalization.

SMTH -- Smoothness [TEMPORAL PATTERN]

Smoothness gauges the regularity of the timing grid and the fluidity of transitions between linguistic units. It differentiates between a legato flow, where sounds melt into one another, and a staccato flow, where sounds are sharp and jerky.

Samples: 175 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=25)

Value	Description
0	The rhythmic timing of the speech is completely chaotic and spasmodic, filled with random, jarring glitches and acoustic micro-stops. The vocal delivery moves in disconnected, highly irregular pulses that suggest severe vocal instability or mechanical breakdown. There is an absolute lack of continuous, predictable acoustic flow.
1	Every syllable is sharply detached from its neighbor, creating a percussive, staccato, machine-gun-like acoustic effect. The timing is extremely rigid, with a total lack of fluid connection or legato transition between sounds. The delivery feels intensely harsh and mathematically segmented into isolated acoustic hits.
2	The rhythmic grid is uneven and bumpy, characterized by frequent micro-hesitations and clumsy acoustic transitions. The speech lacks a steady, predictable pulse, sounding like an unprepared or organically flawed read. The acoustic envelope stumbles awkwardly from word to word without natural grace.
3	The audio features the standard, slightly flexible rhythm of natural human speech, complete with minor, organic timing imperfections. It flows smoothly but retains the natural breathing and slight elasticity characteristic of spontaneous thought processes. The timing is comfortable, balanced, and conventionally fluid.
4	The speech timing is noticeably consistent, practiced, and regular, locking into a predictable and pleasing rhythmic pulse. It sounds like the highly controlled delivery of a professional narrator who smoothly connects phrases with practiced ease. The transitions between words are clean, measured, and highly stable.
5	The vocal delivery flows like silk, featuring a beautiful, continuous, wave-like legato quality in the acoustic signal. Every word melts seamlessly into the next with exceptional grace and perfectly modulated timing transitions. The rhythmic structure is highly pleasing, fluid, and entirely uninterrupted by micro-stops.
6	The rhythm is mathematically perfect and entirely metronomic, completely devoid of natural human timing deviations. The delivery sounds highly mechanical, like a high-end digital synthesizer or an AI voice strictly locked to a digital quantization grid. It possesses an inhuman, flawless acoustic regularity that feels completely sterile.

CLRT -- Articulation Clarity

Articulation focuses on the precision with which the vocal apparatus produces consonants and vowels, measuring the sharpness or blurriness of the acoustic boundaries between sounds.

Samples: 175 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=25)

Value	Description
0	The speech is an indecipherable, blurry hum where distinct consonants and vowel boundaries are completely invisible in the acoustic signal. It sounds like a heavily muffled voice heard through a thick wall, lacking any high-frequency transient bites. The listener cannot extract any clear phonemes from this continuous, mushy sound wave.
1	Consonants are severely swallowed and lack all precision, resulting in a heavily mumbled, fuzzy acoustic texture. The word boundaries are entirely soft, sounding like the speaker is heavily sedated or speaking with a highly restricted oral cavity. The severe lack of articulatory effort makes acoustic transcription incredibly difficult.
2	The articulation is consistently soft and overly relaxed, lacking the crisp transient edges typical of active communication. While most words are intelligible, the overall delivery is mushy, similar to tired, domestic mumble-core speech. The vocal apparatus is applying minimal physical effort to shape the sound waves.
3	The audio exhibits a neutral, standard level of articulation perfectly suited for everyday, clear acoustic communication. Words are easily separable with well-defined vowel shapes and appropriately struck consonants. The speaker puts in just enough mechanical effort to be naturally understood without sounding overly precise.
4	Every syllable is pronounced with crisp, distinct clarity, ensuring highly intelligible and clean phoneme boundaries. The consonants have a clear, precise bite, typical of a high-quality, professional narrator focused on maximum audio legibility. The delivery is sharp and clean without crossing into artificial exaggeration.
5	The articulation is incredibly precise and deliberate, with consonants carved out using high physical energy and perfectly shaped vowels. It has the sharp, commanding acoustic clarity of a classically trained stage actor or an elite broadcaster. Every acoustic boundary between sounds is meticulously managed and highly defined.
6	The phonemes are hyper-articulated to an extreme, unnatural degree, completely over-enunciating every single consonant and vowel. The delivery sounds like a rigid caricature of a formal speaker, with exaggerated plosive pops and sharply cut word endings. This level of mechanical precision is highly distracting and completely abandons organic human blending.

RANG -- Pitch Range

Range describes the vertical movement of the fundamental frequency (pitch) across the segment. It captures the melody of the speech, ranging from a monotone flat line to wide, sweeping sing-song curves.

Samples: 150 total (1=25, 2=25, 3=25, 4=25, 5=25, 6=25)

Value	Description
0	The fundamental frequency remains on a perfectly flat, horizontal line with absolutely zero micro-tonal deviation. The voice is completely devoid of melodic movement, sounding like a singular synthetic beep or a completely lifeless drone. It represents the absolute mathematical minimum of pitch variance in an acoustic signal.
1	The pitch exhibits only microscopic, severely suppressed movements, hovering heavily around a single monotonous baseline note. The vocal melody feels clinical, drained of energy, and aggressively flattened out. It is a highly restricted acoustic profile that completely lacks natural frequency inflection.
2	The intonation moves within a very narrow, highly controlled, and restrained fundamental frequency window. Pitch rises and falls are present but intentionally minimized, typical of a speaker remaining intensely calm or delivering dry technical instructions. The acoustic melody is functional but kept strictly clamped down.
3	The audio features the moderate, standard intonation range of a typical human conversation, complete with organic pitch rises and falls. The fundamental frequency moves smoothly across a standard melodic window, emphasizing thoughts naturally. It is a balanced, conventionally expressive acoustic contour.
4	The pitch moves expressively across a wide range of notes, actively utilizing significant fundamental frequency highs and lows to project personality. The voice sounds lively, colorful, and deeply engaged, creating a dynamic and engaging melodic map. The sweeping pitch jumps clearly highlight the emotional intent of the speaker.
5	The delivery is highly melodic, characterized by sweeping, dramatic pitch jumps and beautifully pronounced sing-song contours. The voice utilizes nearly its full musical range, bouncing rapidly between high acoustic peaks and deep valleys. It possesses the vibrant, animated theatricality of a professional voice actor or storyteller.
6	The pitch swings wildly and hyperbolically between the absolute extreme highs and lows of the speaker's vocal register. It is an aggressively stylized, operatic delivery that sounds larger than life and entirely theatrical. This massive variance in fundamental frequency creates a highly exaggerated, nearly musical vocal performance.

EMPH -- Emphasis [TEMPORAL PATTERN]

Emphasis refers to how the speaker creates an informational hierarchy by stressing specific words or syllables using micro-adjustments in loudness, pitch, and duration.

Samples: 175 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=25)

Value	Description
0	Every single word is delivered with identical acoustic weight, perfectly flat volume, and exactly uniform duration. There is absolutely no informational hierarchy, making the speech sound like an unanchored, monotonous list of syllables. The complete absence of dynamic stress renders the sentence entirely flat.
1	The emphasis is incredibly weak, with only the faintest, nearly undetectable shifts in volume or pitch to mark key words. The hierarchy of information is deeply buried, making it difficult to discern the core focus of the sentence through the audio alone. The dynamic delivery feels aimless and severely under-stressed.
2	Important words are only lightly and subtly marked by minor increases in vocal energy or slight pitch bumps. This gentle emphasis provides a basic, low-stakes structural map for the listener to follow the logic. It is a soft, polite acoustic hierarchy without any aggressive highlighting.
3	The speech features a clear, natural emphasis pattern where key informational anchors are distinctly and comfortably stressed. The listener can immediately identify the core message through organic, well-placed spikes in acoustic volume and pitch. The dynamic weighting is perfectly balanced for everyday comprehension.
4	The speaker heavily and effectively marks important words, creating a strong, highly structured acoustic hierarchy. Key concepts are underlined with distinct vocal force, elongated syllable duration, and clear pitch spikes. The delivery sounds like a highly capable teacher giving crucial, high-stakes instructions.
5	Content words are punched with aggressive, intense energy that aggressively forces the listener's attention to specific points. The emphasis pattern is highly percussive, utilizing sharp, forceful bursts of volume to highlight critical concepts. It is a commanding, highly active dynamic structure that bites through the audio mix.
6	The emphasis is applied with violent, explosive intensity, creating massive acoustic disparities between stressed and unstressed words. Highlighted syllables erupt with jarring volume and prolonged duration, resembling a high-pressure sales pitch or a furious argument. The dynamic contrast within the sentence is stretched to its absolute extreme limit.

DFLU -- Disfluency [TEMPORAL PATTERN]

This dimension tracks the presence of linguistic noise or interference in the delivery, including non-lexical fillers like 'um', false starts, and self-corrections.

Samples: 175 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=25)

Value	Description
0	The audio clip is a pristine, flawless stream of speech with absolutely zero fillers, hesitations, or vocal restarts. It sounds like a perfectly edited studio recording or a masterfully rehearsed script reading. The linguistic signal is completely devoid of any cognitive or mechanical interference.
1	The delivery is highly polished and professional, containing perhaps only a single, negligible micro-hesitation in a long stream of speech. The linguistic output is exceptionally clean, indicating a highly prepared and focused speaker. There is almost no acoustic junk interrupting the primary message.
2	The speech is highly fluent but contains an occasional, organic filler sound that naturally bridges a thought without causing distraction. It represents the smooth, intelligent flow of a highly articulate spontaneous conversation. The tiny acoustic imperfections feel deeply human but do not degrade the legibility.
3	The audio contains a standard, entirely natural amount of non-lexical fillers like 'um' or 'uh', along with minor self-corrections. The delivery feels highly authentic and spontaneous while remaining perfectly easy to comprehend. It represents the exact baseline of unscripted, everyday human processing noise.
4	The speech is noticeably hesitant, with vocal fillers and noticeable processing pauses occurring in nearly every acoustic phrase. The speaker frequently interrupts their own flow to search for words, causing a distinctly staggered and bumpy acoustic output. The high density of linguistic noise significantly slows down the transfer of information.
5	The delivery is overwhelmingly messy, dominated by frequent, distracting restarts, stammers, and heavy usage of vocal fillers. The core message is buried under a thick layer of cognitive interference and broken audio sentences. The listener must actively filter out the high volume of acoustic junk to understand the speaker.
6	The speech is entirely shattered into a chaotic, virtually incoherent web of stammers, abandoned words, and continuous filler sounds. The speaker completely fails to maintain a stable linguistic thread, making it physically difficult to extract any meaning from the audio wave. It is a pure acoustic manifestation of profound cognitive or linguistic breakdown.

STRU -- Structure [TEMPORAL PATTERN]

Structure looks at the macro-level organization of ideas within a clip, differentiating between a linear, logical progression and a scattered, jumpy, or circular delivery.

Samples: 152 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=2)

Value	Description
0	The speech lacks any logical connection between words, phrases, or ideas, presenting as an entirely scattered acoustic word salad. It consists of completely unrelated, jarring outbursts that possess zero macro-level organization. The listener cannot map any linear progression or coherent thought process in the audio.
1	The delivery is incredibly jumpy, with the speaker constantly abandoning topics mid-sentence and pivoting without any acoustic transitions. It is an erratic, highly disorganized stream of audio that totally fails to connect point A to point B. The cognitive map of the speech is fractured and nearly impossible to track.
2	The audio features a very loose, rambling structure where a main idea is heavily buried under layers of irrelevant tangents and side-thoughts. The delivery sounds like an unorganized, wandering stream-of-consciousness that struggles to reach an acoustic destination. The flow is meandering, lacking clear boundaries or logical guideposts.
3	The speech maintains a basic, coherent logical flow that successfully moves from one idea to the next without losing the listener. The speaker utilizes standard conversational markers to link thoughts together in a naturally progressing chain. It is a functional, easily summarizable delivery of audio information.
4	The delivery is distinctly linear and well-organized, marching along a highly predictable and clear path from introduction to conclusion. The speaker expertly avoids tangents, making the logical explanation effortlessly easy to follow. It has the distinct acoustic markers of a well-planned, sequential thought process.
5	The speech is exceptionally organized, utilizing very clear verbal transitions, strong signposting, and disciplined returns to the main topic. It sounds like a highly prepared, professional address that effortlessly manages complex acoustic information. The macro-level structure is robust, elegant, and highly effective.
6	The delivery exhibits perfect, high-level architectural clarity, functioning like a flawlessly structured masterclass or formal oral essay. The speaker sets up clear acoustic introductions, deliberate developmental steps, and highly defined summary closures. The logical map of the audio is an absolute masterpiece of communicative engineering.

STNC -- Stance

Stance describes the speaker's social posture and their behavior in the conversation space, ranging from a Submissive style to a Dominant style that claims the acoustic floor.

Samples: 175 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=25)

Value	Description
0	The voice sounds incredibly tiny, submissive, and heavily apologetic, actively attempting to take up as little acoustic space as possible. It lacks all projection and confidence, signaling a complete surrender of social power to the listener. The vocal delivery feels like the speaker is trying to make themselves entirely invisible.
1	The audio features a very reserved, cautious, and quiet delivery that respectfully waits for long silences before initiating sound. The phrase endings are deeply softened and drop in volume, indicating a distinct hesitation to claim the conversational floor. It is a highly guarded, socially deferential acoustic posture.
2	The speaker adopts a highly cooperative, friendly, and open tone that actively invites back-and-forth exchange. The acoustic markers suggest a balanced, polite awareness of a conversational partner, utilizing warm intonation to share the space. It is an inviting, egalitarian vocal stance.
3	The delivery utilizes a neutral, standard interactional style focused entirely on relaying information rather than establishing social hierarchy. The speaker confidently occupies their space without either imposing on or deferring to the listener. It is a balanced, highly functional interpersonal acoustic baseline.
4	The voice is highly assertive, confident, and direct, utilizing strong projection that signals an absolute expectation to be heard. The speaker firmly claims the acoustic floor and drives the interaction forward without hesitation. It is a proactive, secure, and commanding conversational posture.
5	The speech carries a heavy, authoritative weight, sounding exactly like a person in absolute command of the acoustic environment. The phrasing is intensely final, utilizing loud, downward-inflected endings that allow virtually no room for doubt or interruption. It is a deeply dominant, uncompromising display of vocal power.
6	The speaker completely dominates and monopolizes the acoustic space, delivering a relentless, aggressive lecture that talks entirely at the listener. There is absolutely zero room left for a response, effectively crushing a dialogue into a forcefully imposed wall of sound. It is an overwhelming, highly aggressive display of pure interpersonal dominance.

GEND -- Perceived Gender

This dimension measures the perceived gender of the voice based on the psychoacoustic interaction of pitch and resonance (the size of the perceived vocal tract). High values represent traditional masculine signatures, while low values represent traditional feminine signatures.

Samples: 175 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=25)

Value	Description
0	The acoustic profile features an extremely high fundamental frequency combined with the bright, small-cavity resonance characteristic of a hyper-feminized vocal tract. The voice carries a very light physical weight and delicate harmonic texture, entirely devoid of low-end chest frequencies. It sounds like a high soprano, projecting a bright, purely head-dominant acoustic signature.
1	The audio presents a clearly feminine vocal signature with standard female fundamental pitch ranges and proportionally scaled formant frequencies. The tonal balance is strongly centered in the head and oral cavities, lacking heavy lower-harmonic rumble. The acoustic weight and spectral footprint are unmistakably characteristic of a typical adult woman.
2	The voice is predominantly feminine but exhibits subtle neutral or slightly darkened resonant qualities in its acoustic profile. It operates at a slightly lower fundamental frequency, resembling a female alto or a speaker with a marginally larger perceived vocal tract. The tonal weight is slightly thicker than standard feminine speech, adding a gentle acoustic warmth.
3	The acoustic signature sits perfectly centered on the androgynous spectrum, making it completely ambiguous and difficult to categorize by traditional gender markers. The audio lacks both the sharp, high-frequency brightness of typical feminine speech and the heavy, thick chest rumble of masculine speech. It is a highly balanced, perfectly neutral vocal tract manifestation.
4	The vocal profile is predominantly masculine but features a slightly elevated fundamental pitch and a somewhat smaller resonant cavity than a typical male. The harmonic weight is relatively light, producing a bright, tenor-like acoustic quality that lacks heavy sub-harmonic rumble. It sounds like a youthful male or a man speaking with a noticeably lifted, head-forward resonance.
5	The audio delivers a standard masculine vocal signature characterized by a naturally low fundamental frequency and distinct, robust chest resonance. The spectral balance features strong low-mid frequency support, creating a thick, authoritative acoustic weight. It is the unmistakable, highly resonant sound profile of a typical adult man.
6	The vocal tract produces a hyper-masculinized acoustic profile, heavily dominated by an extremely low fundamental frequency and massive, booming chest resonance. The vocal folds sound incredibly thick, generating an overpowering, low-frequency rumble that fills the entire acoustic space. It mimics the larger-than-life, deeply resonant depth of a cinematic hero or a classic sub-bass radio announcer.

AGEV -- Voice Age

Agev tracks the perceived biological age of the speaker based on vocal maturity, elasticity, and wear. It ranges from the pure, small-tract resonance of infancy to the ossified, unstable acoustic profile of the elderly.

Samples: 175 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=25)

Value	Description
0	The acoustic signal is dominated by the extremely small vocal tract and unstable, high-pitched fundamental frequency of an infant or toddler. The audio entirely lacks complex linguistic structure, consisting primarily of pure, high-frequency biological resonance and primitive vocalizations. The physical elasticity of the vocal folds produces bright, highly piercing neonatal sound waves.
1	The voice carries the clear, bright, and highly energetic acoustic signature of a pre-pubescent child. The resonant cavity is distinctly small, yet the articulation shows developing structure, producing a lightweight, highly elastic harmonic profile. The audio entirely lacks the grounded, heavy chest frequencies characteristic of adult physical maturity.
2	The vocal profile exhibits the transitional acoustics of an adolescent, blending remnants of childhood brightness with newly developing lower-frequency depth. The fundamental pitch may occasionally destabilize or 'crack,' revealing the rapid biological scaling of the vocal tract. The overall resonance is thickening, but has not yet settled into full adult acoustic weight.
3	The audio showcases a voice at peak biological vigor, producing perfectly elastic, incredibly clean harmonic definition across all frequencies. There is absolutely no acoustic evidence of vocal fatigue, muscular wear, or aging-related texture in the signal. The fundamental pitch is rock-solid, and the resonance is highly resilient, fresh, and perfectly supported.
4	The acoustic signature reflects a fully developed, mature adult, characterized by a highly settled, grounded, and authoritative resonant weight. The vocal folds exhibit a very slight, natural thickening, which adds a warm, robust low-mid frequency presence to the audio. The overall physical support is incredibly stable, projecting a commanding and complete spectral profile.
5	The voice begins to display the distinct acoustic markers of middle-to-late aging, including a slight loss of respiratory support and a narrowing of the functional pitch range. A noticeable 'grainy' or weathered texture permeates the audio, resulting from the natural, slow desiccation of the vocal folds over decades. The physical momentum of the speech is perceptibly heavier and slightly less elastic.
6	The acoustic profile is heavily defined by extreme biological senescence, characterized by a pronounced, uncontrollable physiological tremor or 'wobble' in the fundamental frequency. The resonance sounds highly ossified, thin, and reedy, with significant breathiness indicating a massive loss of muscular vocal support. The audio captures the labored, fragile phonation typical of very advanced age.

REGS -- Register

Register classifies the dominant musical tessitura or pitch-class of the voice using classical domain terms. It maps the vertical acoustic space the speaker naturally occupies.

Samples: 175 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=25)

Value	Description
0	The fundamental frequency resides at the extreme low end of human phonation, producing a heavy, cavernous Basso Profondo acoustic profile. The audio is overwhelmingly dominated by thick, low-frequency chest vibrations that create a dark, rumbling physical weight. The vocal baseline sits so low it almost borders on continuous vocal fry.
1	The vocal gravity sits comfortably in the standard Baritone range, blending a rich, grounded low-end depth with highly functional mid-range clarity. The acoustic profile is deeply warm and stabilizing, producing a robust harmonic footprint that easily cuts through a mix. It is a highly common, pleasing male musical register.
2	The pitch center is elevated into the Tenor register, characterized by a light, agile, and noticeably 'bright' fundamental frequency. The audio entirely lacks heavy sub-harmonic chest rumble, instead relying on a highly projected, ringing presence in the upper-mid frequencies. The voice possesses a buoyant, physically lifted acoustic quality.
3	The acoustic signature occupies the ambiguous 'bridge' register, manifesting either as a uniquely deep, husky female Contralto or a high, head-voiced male Countertenor. The fundamental pitch hovers in a highly distinct transitional zone, blending upper-chest warmth with lower-head resonance. The resulting audio footprint is rich, complex, and deeply textured.
4	The vocal center of gravity naturally rests in the standard female Mezzo-Soprano range, offering a perfectly balanced, versatile mid-frequency tone. The audio contains more physical 'body' and lower-harmonic warmth than a high soprano, yet remains cleanly situated in the upper-mid acoustic spectrum. The resonance is highly adaptable and functionally clear.
5	The fundamental frequency is brightly elevated into the Soprano register, generating a soaring, highly brilliant acoustic profile. The audio is overwhelmingly centered in the upper head resonance, entirely stripping away lower-frequency chest weight to produce a light, airy sound. The voice shines with high-frequency harmonic presence.
6	The pitch sits at the absolute upper limit of biological phonation, producing an impossibly thin, piercing Coloratura or 'whistle' register. The acoustic signal is intensely bright and completely devoid of lower harmonics, mimicking the high-frequency chirp of a bird. The vocal baseline operates entirely outside the boundaries of normal conversational norms.

VALN -- Valence

Valence measures the global emotional polarity of the performance, tracking the underlying acoustic signals of pleasure (positivity) or pain (negativity).

Samples: 175 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=25)

Value	Description
0	The acoustic signal is completely saturated with pure, unadulterated emotional suffering, broadcasting extreme hostility, deep agony, or pathological bitterness. Every micro-inflection is heavily weighed down by absolute aversion, completely stripping the audio of any warmth or comfort. It is the raw, highly disturbing sound of a speaker trapped in a state of total negativity.
1	The vocal delivery is clearly anchored in an unhappy, frustrated, or heavily saddened emotional state. A distinct negative 'weight' pulls down the pitch contours, coloring every single phoneme with an acoustic sense of gloom and deep dissatisfaction. The audio profile is dark, heavily burdened, and distinctly unwelcoming.
2	The speech exhibits a subtle, negative-leaning emotional bias, marked by faint acoustic traces of boredom, mild disappointment, or simmering unease. While functioning near the baseline, the vocal tone lacks any upward energy, maintaining a slightly heavy, 'scowling' micro-expression. The psychological state feels slightly closed off and subtly uncooperative.
3	The audio presents a perfectly balanced, emotionally neutral baseline completely devoid of any detectable positive or negative affective bias. The speaker delivers the linguistic content with a highly objective, purely informative acoustic profile that lacks any personal emotional involvement. It is a sterile, functionally flat emotional landscape.
4	The voice demonstrates a positive-leaning, highly polite, and mildly interested emotional state. The overall acoustic delivery sounds perceptibly 'lighter' and more open than the standard baseline, often featuring subtle upward pitch inflections that signal approachability. It is a warm, gently inviting, and entirely unthreatening vocal posture.
5	The acoustic profile is infused with a clear, undeniable 'smile,' radiating a cheerful, happy, and highly warm emotional state. The fundamental frequency dances with positive, upward-leaning energy, creating a highly engaging and deeply inviting auditory atmosphere. The positivity is physically audible in the brightened shape of the vowels.
6	The vocal delivery explodes with pure, euphoric, and deeply ecstatic emotional joy. The voice sounds as if it is completely overflowing with manic pleasure, utilizing incredibly bright resonance and soaring, uninhibited pitch peaks. It is a wildly unconstrained, highly contagious acoustic manifestation of absolute happiness.

AROU -- Arousal

Arousal measures the physiological 'engine speed' of the speaker, tracking autonomic nervous system activation from near-sleep to total fight-or-flight.

Samples: 175 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=25)

Value	Description
0	The physiological engine of the speaker is operating at an absolute minimum, producing a severely lethargic, nearly comatose acoustic signal. The voice physically drags, lacking the fundamental breath energy required to sustain clear phonation, resulting in heavily faded, trailing audio. It is the sound of total physiological exhaustion or deep drowsiness.
1	The vocal energy is extremely low, characterized by a highly sedate, physically relaxed, and carefully muted acoustic delivery. Everything in the biological apparatus is slowed down, mirroring a deep meditative state or an incredibly quiet, hushed 'library' environment. The overall sound wave is deeply subdued and gentle.
2	The speaker is entirely awake but is consciously holding back their physiological energy, maintaining a highly reserved and strictly controlled vocal output. The intensity of the acoustic signal is deliberately suppressed, typical of a cautious, highly formal, or hyper-professional setting. The physical drive of the voice is kept on a very tight leash.
3	The audio captures the standard, baseline physiological energy level of a fully alert human engaged in a normal conversation. The speaker sounds entirely present, utilizing a healthy, balanced physical force that perfectly matches the conversational context. The autonomic nervous system is active, stable, and completely unexceptional.
4	The physiological activation is clearly elevated, driving the vocal apparatus with an energized, slightly urgent physical momentum. A perceptible 'hum' of high alertness buzzes through the audio, making the speech sound highly insistent, excited, or mildly anxious. The acoustic transients hit with noticeable, fast-paced physical force.
5	The autonomic nervous system is in a state of high activation, pushing the vocal folds with intense physical pressure and heavily increased acoustic volume. The energy borders on shouting, demonstrating an aggressive, high-stakes physical drive that begins to actively strain the biological limits of the throat. The audio is saturated with highly mobilized, forceful vocal power.
6	The vocal apparatus is pushed into maximum physiological mobilization, resulting in a hysterical, violently extreme acoustic output. The energy is so overwhelmingly high—resembling a full panic attack or a blood-curdling scream—that the biological system sounds on the verge of total failure. The sound wave is a highly alarming, explosive manifestation of absolute fight-or-flight survival.

RESP -- Respiration [TEMPORAL PATTERN]

Respiration measures the audibility and laboriousness of the speaker's breathing, indicating physical exertion or respiratory distress.

Samples: 175 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=25)

Value	Description
0	The speaker's respiratory cycle is completely invisible and inaudible within the acoustic signal, allowing for impossibly long, seamless linguistic phrases. They possess seemingly infinite breath control, characteristic of a highly trained singer or an elite public speaker operating with absolute physiological efficiency. The 'wind' of the lungs never once interrupts the 'music' of the voice.
1	Breathing is strictly controlled and occurs only as highly efficient, barely perceptible micro-inhalations taken at long, calculated intervals. The respiratory function operates entirely in the background of the audio mix, never pulling focus from the acoustic clarity of the words. It is a highly disciplined, remarkably quiet respiratory profile.
2	The audio features standard, completely natural conversational breathing where inhalations occur at logical grammatical pauses. The sound of the breath is mildly audible but entirely unobtrusive, perfectly blending into the organic rhythm of the speech. The respiratory engine sounds healthy, unforced, and entirely typical.
3	The speaker's breathing is noticeably elevated, with inhalations occurring more frequently and audibly than strictly necessary for basic phonation. The slightly heavier air intake suggests a subtle increase in physical engagement, mild psychological stress, or active physical movement. The sound of the respiratory cycle begins to actively assert itself into the audio stream.
4	The respiratory cycle is highly labored, characterized by deep, clearly audible gasps and heavy heaving between sentences. The speaker is rapidly running out of air, forcing them to physically truncate their phrase lengths to accommodate their desperate need for oxygen. The acoustic signal is frequently and aggressively interrupted by the struggle to breathe.
5	The audio is dominated by rapid, highly shallow gasping that violently and unpredictably chops the linguistic stream into fragmented pieces. The speaker sounds heavily winded, as if they have just finished a full-speed sprint or are experiencing severe, acute physical distress. The violent rush of air is almost louder than the actual phonation.
6	The physical ability to speak is almost entirely destroyed by catastrophic respiratory failure, resulting in an agonal, suffocating acoustic profile. The audio captures the terrifying sounds of choking, hyperventilation, or a desperate fight for survival where air cannot be processed into words. It is an extreme, highly distressing biological breakdown of the breathing mechanism.

TENS -- Tension

Tension measures the perceived muscular 'tightness' in the throat, jaw, and neck, ranging from completely relaxed and floppy to rigidly locked and strangled.

Samples: 157 total (0=7, 1=25, 2=25, 3=25, 4=25, 5=25, 6=25)

Value	Description
0	The vocal apparatus exhibits absolutely zero muscular resistance, producing an impossibly 'floppy,' wide-open, and entirely tension-free acoustic signal. The voice sounds profoundly relaxed, mirroring the deep physiological release of a post-massage state or heavy sedation. The resulting sound wave is incredibly soft, sprawling, and physically uninhibited.
1	The throat and jaw muscles are highly loose and comfortable, producing a deeply warm, physically effortless vocal tone. There is absolutely no noticeable constriction in the biological pipeline, allowing the audio to resonate with a highly pleasing, 'open' acoustic geometry. The speech is physically incredibly easy to produce.
2	The voice utilizes standard, neutral conversational muscle tone, applying just enough physical firmness to clearly articulate without squeezing the vocal tract. There is no distracting or uncomfortable physical tension audible in the acoustic signal. The throat remains appropriately relaxed but functionally active.
3	A faint acoustic 'edge' or mild muscular squeeze becomes audible, indicating a subtle tightening of the neck or jaw muscles. This mild tension profile often suggests a speaker who is deeply concentrating, slightly annoyed, or operating under minor psychological stress. The voice loses a fraction of its natural warmth due to this slight physical constriction.
4	The vocal tone sounds noticeably pressed and restricted, as if the audio is being physically pushed through a narrowed, tightly clenched biological pipe. The speaker is clearly carrying heavy stress in their jaw and neck, resulting in a firm, mildly metallic, and highly constrained acoustic signal. The lack of open resonance is physically palpable.
5	The muscles surrounding the vocal folds are heavily strained, producing a hard, grinding, and intensely uncomfortable acoustic texture. The sheer physical effort required to force sound through the tightly constricted throat is blatantly audible, robbing the voice of all organic softness. The audio feels highly stressful and physically painful to listen to.
6	The somatic tension is so extreme that the vocal musculature is entirely rigid and locked, actively crushing the vocal folds and causing the voice to crack or break. The delivery sounds brutally strangled, heavily distorted by the immense external pressure applied by the neck and jaw. It is the acoustic manifestation of a biological system paralyzed by high-stakes stress.

COGL -- Cognitive Load [TEMPORAL PATTERN]

Cognitive Load tracks the mental friction involved in formulating speech, measuring the real-time struggle to turn complex thoughts into spoken words.

Samples: 175 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=25)

Value	Description
0	The speech output is perfectly fluid and completely effortless, possessing absolutely zero cognitive hesitation or processing friction. It sounds as though the speaker has completely memorized the text and is reciting it from deep neural pathways without a single micro-second of real-time thought. The thought-to-speech pipeline is flawlessly executed.
1	The cognitive delivery is highly articulate and incredibly fast, demonstrating a near-perfect, highly efficient mental processing speed. The speaker easily translates complex ideas into a perfectly sequenced acoustic stream with almost zero visible mental labor. The intellectual machinery operates with highly impressive, uninterrupted smoothness.
2	The audio features standard, healthy conversational flow where the speaker experiences only occasional, completely natural micro-pauses to search for the correct word. The mental processing is entirely manageable and feels organically easy, creating a comfortable, naturally paced acoustic rhythm. The brain-to-mouth connection is stable and typical.
3	The speaker's real-time mental processing becomes audibly noticeable, creating an uneven acoustic rhythm as they actively build their sentences block by block. There is a distinct 'searching' quality in the voice, accompanied by minor hesitations that reveal the brain working to organize the logic. The cognitive load is elevated but successfully managed.
4	The cognitive machinery is heavily burdened, resulting in a delivery weighed down by frequent vocalized pauses, long gaps, and a distinct acoustic 'heaviness.' The speaker is visibly struggling to formulate their thoughts, heavily slowing down the linguistic output as their brain attempts to solve the communication problem. The mental effort significantly degrades the pacing of the audio.
5	The speaker's mental capacity is severely overloaded, causing them to frequently lose their train of thought, self-correct, and acoustically backtrack. The heavy cognitive friction makes the internal struggle blatantly obvious, heavily fracturing the flow of information and causing the speaker to stumble over their own logic. The brain is clearly failing to keep up with the mouth.
6	The cognitive burden is so catastrophically high—due to extreme shock, complex fabrication, or intellectual overload—that the speaker's linguistic output completely breaks down. The brain entirely fails to string together a coherent sentence, resulting in a shattered, deeply confused, and logically incoherent acoustic wave. The mental operating system has essentially crashed.

FOCS -- Focus

Focus measures the direction of the speaker's interpersonal attention, ranging from totally dissociated and inward to obsessively laser-focused on the listener.

Samples: 175 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=25)

Value	Description
0	The acoustic energy is entirely internalized and heavily dissociated, sounding exactly like a person sleep-talking or mumbling strictly to themselves. The speaker demonstrates zero awareness of an external audience, 'dropping' their words aimlessly rather than projecting them outward. It is a profoundly isolated, completely inward-facing vocal geometry.
1	The delivery is completely emotionally detached and highly impersonal, resembling someone coldly reading a script into an empty void. While the voice is technically audible, the 'soul' of the speaker is entirely disconnected from the listener, producing a chillingly distant acoustic profile. There is absolutely zero interpersonal 'reach' in the audio wave.
2	The speaker's attention is clearly split, resulting in a distracted, slightly 'hollow' acoustic delivery where their mind is visibly wandering. They are physically speaking to someone, but the vocal projection lacks true interpersonal grounding, creating a slightly unfocused and floaty audio presence. The communicative intent is weak and continually drifting.
3	The audio captures a standard, polite interactional focus where the speaker is casually aware of the listener without applying intense psychological pressure. The vocal energy is appropriately directed outward, maintaining a comfortable, socially standard interpersonal connection. The acoustic projection is functional, balanced, and appropriately present.
4	The speaker employs clear, highly engaged 'Direct Address,' actively ensuring their acoustic energy firmly lands on the listener. The voice possesses a strong forward projection that metaphorically 'looks' right at the audience, demanding their attention and cognitive participation. The interpersonal connection is tight, highly active, and deeply communicative.
5	The vocal delivery utilizes an intense, highly precise laser-focus, locking onto the listener with overwhelming interpersonal intent. Every single syllable is aggressively 'thrown' at the audience to persuade, threaten, or seduce, creating a deeply penetrating acoustic presence. The psychological weight of the audio is completely inescapable.
6	The speaker's focus becomes deeply obsessive and almost uncomfortably hypnotic, treating the listener as the sole existing object in their entire reality. The acoustic delivery is overwhelmingly intimate and trance-like, completely invading the listener's psychological space with an unblinking, hyper-focused vocal intensity. It is an extreme, profoundly overwhelming manifestation of directed presence.

VOLT -- Volatility [TEMPORAL PATTERN]

Volatility measures how much the emotional state shifts or vibrates within the clip, tracking the psychological stability of the performance.

Samples: 175 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=25)

Value	Description
0	The emotional state is absolutely static and heavily frozen, presenting as an unyielding block of psychological ice. The voice exhibits zero micro-shimmer or affective reaction to the actual words being spoken, resulting in a disturbingly flat and highly rigid acoustic profile. The emotional trajectory is a completely dead, unmoving flatline.
1	The vocal delivery is highly steady and exceptionally consistent, locking into a single, unvarying mood with intense professional discipline. The affective tone is deeply grounded and completely immune to external or internal disruption, producing a highly reliable and unshakeable acoustic signal. The psychological baseline is rock solid.
2	The audio displays a standard, highly stable level of human emotional consistency, featuring only minor, entirely logical affective shifts based directly on the text. The overall psychological state remains deeply firm, completely avoiding any erratic or unpredictable acoustic behavior. The emotional baseline is perfectly anchored.
3	The speech features the natural, organic 'breathing' of standard human emotion, allowing the affective state to dynamically flex and respond to the conversation in real-time. The emotional shifts feel highly spontaneous and healthy, creating a gently undulating, realistic acoustic landscape. The psychological volatility is perfectly balanced.
4	Minor, uncontrollable 'cracks' begin to appear in the speaker's emotional facade, introducing a distinct psychological 'flickering' into the acoustic signal. The speaker is actively attempting to remain steady, but underlying surges of emotion cause rapid micro-shifts in pitch and volume that leak into the audio. The emotional stability is noticeably straining under pressure.
5	The vocal delivery is highly unstable, characterized by jarring, unpredictable shifts in affective tone that drastically alter the acoustic profile. The speaker sounds emotionally untethered, sounding as if they could violently snap from laughing to crying in a fraction of a second. The psychological vector is highly erratic and physically difficult to predict.
6	The emotional state is completely chaotic and deeply shattered, featuring violent, rapid-fire affective switching that fundamentally destabilizes the entire acoustic signal. The speaker cycles through manic, extreme emotional extremes every few seconds, entirely destroying any sense of a consistent psychological baseline. The audio is a terrifying, highly volatile map of total emotional collapse.

VULN -- Vulnerability

Vulnerability measures the 'openness' or 'permeability' of the speaker, ranging from completely armored and defended to totally raw and emotionally naked.

Samples: 175 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=25)

Value	Description
0	The acoustic profile is completely impenetrable and heavily armored, projecting an absolute, unyielding emotional defense wall. The tone is hard, shiny, and highly guarded, mimicking the untouchable vocal posture of a mob boss or a drill sergeant where absolutely nothing emotional 'leaks out.' The psychological barrier is actively hostile to any connection.
1	The delivery is highly guarded, deeply cautious, and strictly professional, meticulously maintaining a hard interpersonal boundary. The speaker tightly controls their acoustic footprint, sharing only the absolute minimum amount of personal resonance required to communicate. The psychological 'mask' is firmly and deliberately strapped on.
2	The voice utilizes a highly composed, standard adult emotional mask that keeps deep personal vulnerability politely and safely hidden. The tone is perfectly functional and socially appropriate, providing a smooth, well-maintained acoustic surface that prevents the listener from peering too deeply inside. The emotional core remains safely guarded.
3	The acoustic delivery is open, friendly, and highly accessible, featuring a distinct 'soft' edge to the vocal resonance that invites connection. The speaker sounds entirely willing to be emotionally affected by the interaction, lowering their vocal defenses to create a warm, unguarded acoustic space. The psychological boundary is comfortably lowered.
4	The speaker's voice is highly empathetic and distinctly permeable, sounding as if they are actively absorbing and reacting to the emotional energy of the listener. The acoustic profile is deeply responsive, stripped of standard defensive armor, and highly sensitive to the interpersonal environment. The psychological state is deeply engaged and highly unprotected.
5	The speaker's emotional mask is completely removed, resulting in an exposed, deeply defenseless acoustic profile often heard during profound confessions or moments of genuine awe. The voice is entirely stripped of all social protection, allowing raw, unfiltered emotional resonance to completely dictate the pitch and volume. The listener is granted absolute, unhindered access to the speaker's core state.
6	The acoustic signal is profoundly raw, completely naked, and overwhelmingly permeable, completely lacking any psychological filter or protective barrier. The speaker's sensitivity is dialed so high that the mere act of speaking seems to cause physical emotional pain, resulting in a deeply trembling, completely shattered vocal output. It is the absolute extreme of terrifying, unshielded emotional exposure.

BRGT -- Brightness

Brightness refers to the prominence of high-frequency energy (treble) within the voice. It describes the sheen or brilliance of the signal, determined by the presence of upper harmonics and sibilance.

Samples: 145 total (0=8, 1=12, 2=25, 3=25, 4=25, 5=25, 6=25)

Value	Description
0	The acoustic signal is completely devoid of high-frequency energy, sounding heavily muffled as if submerged underwater or blocked by a thick physical barrier. Transient consonants and sibilance are entirely absorbed and invisible, leaving only a dark, formless blob of audio. The spectral balance is severely rolled off at the top, resulting in a muddy low-end rumble.
1	The treble frequencies are significantly reduced, creating a woolly, dark acoustic texture that entirely lacks a sharp articulatory edge. Sibilant sounds are soft and heavily suppressed, rendering the vocal signal distinctly dull and top-heavy. While marginally more intelligible than a total high-frequency cutoff, the voice remains heavily veiled and muddy.
2	High-frequency energy is present but gently attenuated, resulting in a warm, rounded, and earthy acoustic profile. This slight treble suppression creates a pleasingly smooth, vintage sound that lacks modern analytical crispness but is exceptionally easy on the ears. Sibilance is naturally controlled and softly absorbed into the frequency mix.
3	The spectral balance represents the perfect, neutral baseline of natural human speech with an entirely even distribution of high and low frequencies. There is a highly functional mix of low-end body and high-end clarity that ensures perfect legibility without harshness. Sibilants and transients hit with standard, unexaggerated acoustic presence.
4	The high frequencies are distinctly well-defined and highly articulate, giving the voice a modern, crisp, and extremely present production quality. Every consonant slices through the audio mix with sharp transient clarity and beautifully distinct sibilance. The acoustic profile feels bright and forward without ever causing auditory fatigue.
5	Treble frequencies are heavily emphasized, adding a brilliant, airy shimmer to the absolute top of the vocal spectrum. The voice sits extremely forward in the mix, characterized by highly pronounced sibilance and a sparkling, ultra-clear acoustic sheen. The continuous high-frequency excitation creates a dazzling, though potentially fatiguing, listening experience.
6	The high-frequency energy is overwhelmingly harsh, resulting in a sharply piercing, sizzling acoustic output that feels physically painful to hear. Sibilant consonants act like breaking glass, dominating the entire spectrum with aggressive, ear-drilling treble spikes. The signal is entirely unbalanced, relying exclusively on an abrasive, highly metallic top-end.

ROUG -- Roughness

Roughness measures the presence of rapid amplitude modulations or grit within the vocal fold vibration itself. It captures the biological texture of the voice, independent of recording distortion.

Samples: 144 total (0=2, 1=17, 2=25, 3=25, 4=25, 5=25, 6=25)

Value	Description
0	The vocal fold vibration is impossibly smooth and pure, closely resembling the clean acoustic geometry of a pure sine wave or a synthetic flute. The signal contains absolutely zero aperiodic noise, biological grit, or vocal fry, resulting in a perfectly clean, uninterrupted tonal stream. It is a flawless, glassy acoustic emission completely free of organic friction.
1	Phonation is exceptionally consistent and velvety, exhibiting only the absolute faintest microscopic traces of organic human texture. The vocal folds vibrate with incredibly healthy, polished regularity, generating a highly soothing and buttery acoustic output. The sound wave glides effortlessly with almost zero internal biological friction.
2	The voice features the standard, healthy level of organic texture typical of everyday human speech and natural biology. A slight, highly natural acoustic 'bite' prevents the tone from sounding synthetic, rooting the vocal production in grounded human reality. There is no distracting grit, fry, or harshness present in the highly stable signal.
3	A distinct, grainy texture noticeably permeates the vocal tone, introducing rapid amplitude modulations into the fundamental frequency. The voice sounds as though it contains a layer of acoustic sand, typical of mild vocal fry or slightly desiccated vocal folds. This gentle grit adds a rough, tactile characteristic to the otherwise stable phonation.
4	The acoustic signal is heavily raspy and scratchy, revealing a highly weathered biological apparatus operating with significant internal friction. The vocal folds vibrate with a continuous, grinding noise floor, sounding like a speaker with a mild cold or a long-term smoking habit. This heavy grit is an inescapable, defining feature of the overall sonic footprint.
5	The vocal texture is aggressively harsh and distinctly growling, driven by intense physical friction and heavy aperiodic interference at the vocal folds. The fundamental tone is deeply disrupted by a heavy, rattling rumble that heavily distorts the biological sound wave. It is a highly abrasive, physically forceful acoustic grind that dominates the spectrum.
6	The fundamental tone is almost completely obliterated by violent, aperiodic noise and extreme biological crunch. The vocal folds are in a state of chaotic, shredding friction, producing an acoustic output akin to a death-metal growl or severe physical tissue breakdown. It is a jarring, maximally rough audio signal utterly devoid of clean periodicity.

HARM -- Harmonicity

Harmonicity measures the ratio of periodic signal (clear pitch) to aperiodic noise (hiss or turbulence), describing how tunable the voice is.

Samples: 155 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=5, 6=25)

Value	Description
0	The acoustic signal is entirely composed of broadband aperiodic noise, lacking even the absolute faintest trace of a discernible fundamental pitch. It represents a totally unvoiced, absolute whisper where all acoustic energy is derived from air turbulence rather than vocal fold vibration. The sound wave is a ghost made purely of friction and white-noise static.
1	A microscopic, ghostly hint of fundamental pitch barely survives inside a dense, overwhelming cloud of high-frequency white noise. The acoustic output is primarily breath and air leakage, with only a tiny, highly fragile fraction of energy successfully converted into tonal resonance. It sounds like a heavy, desperate whisper struggling to find its voice.
2	The audio is distinctly breathy, featuring a clear fundamental pitch that is continuously wrapped in a thick, highly audible layer of air turbulence. Known psychoacoustically as 'bedroom voice,' the vocal folds fail to close completely, allowing a steady stream of hissing noise to mix with the tone. The resulting sound is highly soft, deeply diffused, and highly aerated.
3	The signal achieves the standard, healthy ratio of periodic tone to aperiodic noise characteristic of normal human speaking voices. The fundamental pitch is solid and highly clear, acting as the primary acoustic carrier while only a negligible, natural amount of breath noise escapes. The vocal fold closure is biologically balanced and perfectly functional.
4	The voice is highly resonant and exceptionally focused, featuring a massive fundamental pitch and a remarkably low acoustic noise floor. The tonal energy is highly efficient and distinctly ringing, characteristic of a highly trained public speaker maximizing their vocal fold closure. The audio wave is deeply concentrated and musically tunable.
5	The acoustic output is exceptionally tonal and brilliantly bell-like, successfully converting virtually all breath energy into a pure, ringing fundamental frequency. There is absolutely no audible air leakage or turbulent hiss, resulting in an incredibly clean, almost musical vocal tone. The harmonic stack is incredibly rich, highly stable, and mathematically beautiful.
6	The harmonicity is so absolute and mathematically perfect that the voice sounds entirely computer-generated or digitally synthesized. It is completely devoid of the organic 'fuzz', breath, or aperiodic noise inherently attached to human biology. The resulting tone is a sterile, digitally pure acoustic ideal that feels entirely artificial.

FULL -- Fullness

Fullness describes the size or volume of the sound's spectral footprint, capturing the perceived weight of the audio across the frequency spectrum.

Samples: 131 total (0=2, 1=25, 2=25, 3=25, 4=25, 5=25, 6=4)

Value	Description
0	The voice is overwhelmingly thin and heavily paper-like, completely lacking any spectral body, warmth, or lower-frequency foundation. The acoustic footprint is a flat, one-dimensional sliver of sound that resembles audio forced through a tiny, severely broken watch speaker. It entirely fails to occupy any significant portion of the sonic frequency spectrum.
1	The spectral bandwidth is heavily restricted, artificially rolling off both the low-end bodily thump and the high-end airy frequencies. The resulting audio is highly narrow, severely small, and distinctly boxed-in, perfectly mirroring the heavy frequency compression of a low-quality telephone call. The signal lacks both warmth and brilliance, feeling highly confined.
2	The vocal profile is highly intelligible and clear but remains distinctly slender and physically lightweight in its overall spectral footprint. It lacks the rich low-mid body required to actively command the acoustic space, feeling functionally present but physically very small. The voice does not push significant air or project heavy acoustic mass.
3	The audio features a solid, naturally healthy spectral bandwidth that feels organically complete and appropriately sized for a human speaker. It perfectly balances enough low-mid energy to sound grounded with enough high-end to sound distinctly realistic and present. The acoustic volume occupies a standard, highly comfortable amount of space.
4	The vocal signal is incredibly rich and highly wide, spreading out across a broad frequency spectrum to create a highly expensive acoustic texture. It possesses a strong, thick physical body that feels highly present and structurally robust in the audio mix. The resulting sound wave occupies a massive amount of highly pleasing sonic territory.
5	The audio footprint is massive, physically thick, and highly commanding, featuring a heavily reinforced spectral body that projects immense power. The voice sounds perfectly captured by a high-end studio microphone, fully utilizing a deep proximity effect to generate a heavy, weighty acoustic mass. It effortlessly and entirely dominates the low and mid-frequency ranges.
6	The vocal fullness reaches an overwhelming, IMAX-level cinematic scale, completely saturating the entire acoustic field with unbelievable depth and width. The sound is astronomically big, creating a highly exaggerated, rumbling voiceover effect that vibrates deep within the listener's physical chest. It is a hyper-saturated, omnipresent wall of spectral density.

WARM -- Warmth

Warmth refers to the emphasis on the low-mid frequencies (around 200Hz to 500Hz), providing a sense of coziness, organic depth, and intimacy.

Samples: 140 total (0=14, 1=25, 2=25, 3=25, 4=25, 5=25, 6=1)

Value	Description
0	The acoustic profile is entirely devoid of low-mid frequency resonance, resulting in a distinctly cold, metallic, and surgically sterile sound. The voice completely lacks any organic 'cuddle' or bodily depth, sounding like a harsh dental drill or a highly detached, emotionless machine. It is an actively uninviting, heavily rigid sonic texture.
1	The audio is distinctly tinny and highly top-heavy, severely lacking the low-mid bodily warmth required to sound natural or comforting. This heavily skewed frequency response creates a cheap, highly hollow, and distinctly irritating acoustic signature. The voice feels entirely disconnected from the physical warmth of the human chest.
2	The vocal resonance is highly functional but distinctly neutral-cool, operating with a very dry acoustic profile that prioritizes hard information over comfort. The low-mid frequencies are present but distinctly suppressed, preventing the voice from sounding naturally inviting or physically relaxed. It is a highly utilitarian, slightly stiff acoustic delivery.
3	The resonance perfectly strikes the standard human baseline, offering a highly balanced, universally comfortable acoustic profile. There is exactly enough low-mid frequency presence to sound biologically grounded and human, but not so much that the audio becomes muddy or muffled. The natural warmth of the chest and throat is perfectly proportioned.
4	The voice exhibits a highly pleasant, distinctly woody, and cozy resonance heavily supported by a smooth, gentle low-mid frequency emphasis. This targeted boost in the lower harmonics creates a deeply soothing, physically close acoustic sensation that feels instantly comfortable to the listener. The sonic texture is round, inviting, and inherently relaxing.
5	High levels of rich, low-mid resonance heavily saturate the audio, creating a deeply velvety, highly attractive 'late-night radio DJ' acoustic profile. The vocal folds produce a heavy, exceptionally warm hum that perfectly rounds off all harsh transients and fills the sonic space with comfortable depth. It is a highly engineered, deeply attractive resonant glow.
6	The warmth is so overwhelmingly massive that the voice essentially melts into a gigantic, enveloping physical hug of pure low-frequency resonance. It is a completely bass-heavy, hyper-cozy, 'chocolate-like' acoustic environment that completely swallows the listener's ears. The sound wave entirely sacrifices sharp intelligibility in favor of absolute, overwhelming sonic comfort.

METL -- Metallic Character

Metallic character describes the presence of clanging or ringing inharmonic frequencies, making the voice sound steel-like, hard, or robotic.

Samples: 167 total (0=17, 1=25, 2=25, 3=25, 4=25, 5=25, 6=25)

Value	Description
0	The voice is purely organic and impossibly soft, completely lacking any hard acoustic edges or inharmonic ringing frequencies. The resonance is distinctly wooden or wool-like, perfectly absorbing all high-frequency transients and preventing any harsh reflections within the vocal tract. The audio profile is the absolute, unyielding antithesis of a metallic machine.
1	The acoustic output is deeply grounded in a soft, earthy resonance with only the absolute minimal, microscopic trace of any metallic ping. The sound wave is highly natural, prioritizing a warm, fleshy biological dampening over hard, reflective acoustic amplification. The voice feels intimately human and completely devoid of artificial stiffness.
2	The audio features a standard, highly typical human resonance containing only a negligible, entirely organic amount of high-frequency acoustic 'ping.' The biological vocal tract functions normally, beautifully balancing soft tissue absorption with the hard reflection of the hard palate. The overall sonic footprint remains firmly and functionally natural.
3	A distinct firmness begins to visibly infiltrate the acoustic profile, giving the voice a stiff, slightly 'hard' physical quality. The speaker sounds as though they are operating with a tightly clenched jaw, causing higher frequencies to reflect harshly off the internal vocal tract. This introduces a subtle but completely un-ignorable metallic sheen.
4	The audio is deeply characterized by a clear, highly audible metallic resonance that sounds exactly like a voice bouncing directly off a steel plate. A pronounced, highly inharmonic 'ring' or 'ping' actively interferes with the fundamental tone, highly typical of extreme vocal tension or aggressive projection. The sound wave feels intensely rigid and unyielding.
5	The vocal output is fiercely clanging and distinctly piercing, heavily saturated with sharp, iron-like inharmonic frequencies that actively assault the ear. The resonance is incredibly hard and physically dangerous, completely stripping away any illusion of soft, organic biological tissue. The acoustic edge is highly cold, severely abrasive, and deeply metallic.
6	The voice is entirely consumed by a pure, robotic metallic resonance, sounding exactly like a synthetic machine or a heavily armored entity. It is a violently hard, continuously clanging acoustic signal that lacks even a single trace of organic, human softness. The sonic texture is a perfect, terrifying manifestation of cold, vibrating steel.

ESTH -- Esthetics

Esthetics is a highly consistent dimension measuring the overall attractiveness or pleasantness of the sound, combining recording quality, vocal health, and charismatic delivery.

Samples: 168 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=18)

Value	Description
0	The audio is physically intolerable and viscerally repulsive, instantly triggering an immediate, desperate desire to stop the playback. It consists of highly painful acoustic elements like violent distortion, extreme high-pitched screeching, or deeply disturbing biological noises. The sonic experience is an absolute, unmitigated assault on the listener's auditory system.
1	The vocal quality is deeply unpleasant and highly grating, actively forcing the listener to endure harsh, whining, or exceptionally 'ugly' acoustic textures. The signal is deeply irritating, feeling audibly 'dirty' or aggressively fatiguing to process over any length of time. Listening to this audio requires significant effort and active tolerance of severe sonic flaws.
2	The audio profile is entirely mundane and fundamentally boring, heavily lacking any discernible acoustic appeal or engaging characteristics. While technically free of painful distortion, the voice sounds highly cheap, exceptionally dry, and thoroughly uninteresting to the human ear. It is a highly forgettable, perfectly mediocre, and strictly utilitarian sonic experience.
3	The recording presents a highly acceptable, perfectly functional standard of audio quality typical of average human voices or standard consumer microphones. The signal is cleanly intelligible and entirely unobtrusive, though it completely lacks any professional polish or distinct vocal charisma. It is the absolute baseline for safe, everyday auditory consumption.
4	The acoustic experience is highly pleasant, featuring a genuinely good human voice flawlessly captured by a clean, professional recording chain. The audio is incredibly easy on the ears, highly well-balanced, and completely devoid of any distracting technical or biological flaws. It provides a highly satisfying, premium-quality listening environment.
5	The vocal delivery radiates massive 'star power,' flawlessly combining an exceptionally charismatic, rich voice with a beautifully polished studio recording. The acoustic texture is deeply soothing, highly engaging, and undeniably attractive to the listener's ear. It is a premium, highly engineered auditory product that actively and effortlessly commands positive attention.
6	The audio achieves a sublime, almost holy level of absolute acoustic beauty, representing the absolute pinnacle of perfect vocal and technical synergy. The voice is an ecstatic, physically thrilling listening experience that flawlessly balances deep warmth, striking clarity, and massive emotional resonance. It is a flawless, deeply transcendent sonic masterpiece.

VFLX -- Velocity Flux [TEMPORAL TRAJECTORY]

VelocityFlux measures the change in speed over the duration of the clip, capturing the acceleration or deceleration trend rather than the average speed.

Samples: 175 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=25)

Value	Description
0	The vocal trajectory begins at a noticeably rapid, high-energy pace before undergoing a massive, continuous deceleration, progressively grinding down to an extremely slow, labored crawl by the end. The speech starts rushed and compressed, then steadily loses all forward momentum as syllables stretch wider and wider apart. The speed differential from start to finish is extreme, covering the full range from 1.5x to 0.5x normal pace.
1	The speech begins at a clearly elevated pace and undergoes a heavy, sustained deceleration across the full duration. The initial brisk delivery gradually and obviously slows down, with the speaker audibly losing rhythmic momentum as the phrase progresses. The tempo drops from roughly 1.3x to 0.7x normal speed, creating a pronounced but not extreme winding-down effect.
2	The temporal dynamics exhibit a subtle, natural easing of the tempo, starting slightly faster than normal and gently braking to a slightly slower pace by the end. This mild deceleration functions as a deliberate, comfortable cadence, carefully reducing momentum from about 1.1x to 0.9x normal speed. The slowdown is controlled and rhythmically satisfying, like a speaker naturally winding down a thought.
3	The velocity of the speech remains absolutely locked and perfectly steady from the very first syllable to the last. There is exactly zero detectable acceleration or deceleration, resulting in a strictly linear, metronomic temporal delivery. The acoustic momentum is held at a completely constant speed throughout the entire clip.
4	The speech trajectory features a subtle, natural forward lean, beginning slightly slower than normal and gently pushing the tempo faster as the phrase concludes. This mild acceleration from about 0.9x to 1.1x normal speed suggests a rising tide of engagement or growing confidence. The acoustic momentum subtly builds, creating a gently forward-moving vector.
5	The audio begins at a noticeably slow, deliberate pace and captures a clear, sustained acceleration across the full duration, with the speaker audibly speeding up from roughly 0.7x to 1.3x normal speed. The acoustic rhythm compresses progressively as the phrase progresses, driven by an escalating surge of energy or urgency. The temporal vector is distinctly forward-thrusting with an obvious speed buildup.
6	The tempo vector begins extremely slowly, with syllables heavily stretched and drawn out, before undergoing a massive, continuous acceleration that ends at a rapid, high-energy pace. The speech transforms from a crawl at 0.5x to a rush at 1.5x normal speed, with the acoustic momentum building relentlessly throughout. The speed differential from start to finish is extreme, the mirror image of bucket 0.

DARC -- Dynamic Arc [TEMPORAL TRAJECTORY]

DynamicArc measures the trajectory of the loudness (volume) across the clip, tracking the volume envelope from start to finish.

Samples: 175 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=25)

Value	Description
0	The acoustic volume begins at a highly audible level before suffering a total, catastrophic collapse, fading away into absolute dead silence. The audio completely dissolves and dies out, perfectly tracking a highly linear trajectory toward complete acoustic nothingness. The volume envelope is a pure, absolutely unrecoverable fade to black.
1	The speech vector is characterized by a clear, incredibly steady, and highly pronounced diminuendo that drops the volume drastically across the clip. The speaker sounds exactly as if they are rapidly losing all conviction or physically backing away from the microphone mid-sentence. The dynamic envelope heavily deflates, bleeding massive acoustic energy continuously.
2	The volume trajectory features a gentle, highly natural softening of acoustic power at the very end of the linguistic phrase. This minor dip in decibels is a standard, deeply organic human cadence used to gently signal the conclusion of a thought. The dynamic envelope 'lands' smoothly and softly without entirely losing its structural integrity.
3	The decibel level remains absolutely flatlined, exhibiting exactly zero change in audio volume from the beginning to the very end of the clip. The dynamic range is perfectly compressed and highly static, flawlessly maintaining an unwavering, mathematically constant acoustic output. The volume envelope is entirely frozen in place.
4	The dynamic envelope perfectly traces a highly controlled, deeply satisfying bell curve of rising and falling acoustic volume. The speech begins quietly, smoothly swells into a louder, highly energetic peak in the middle, and then gracefully softens back down at the end. It is a deeply musical, highly planned trajectory of deliberate vocal power.
5	The volume vector tracks a clear, highly aggressive crescendo, building massive acoustic loudness and physical insistence across the duration of the clip. The speaker is actively 'powering up,' visibly driving the decibel level significantly higher with every passing word. The dynamic envelope is highly ascendant, continuously gathering forceful, undeniable momentum.
6	The dynamic arc covers an impossibly extreme range, starting as a tiny, barely audible whisper and exploding directly into a massive, floor-shaking scream. This violent trajectory acts as a sheer vertical ramp of acoustic power, completely shattering the initial quiet with an overwhelming surge of volume. The decibel differential is absolutely massive and highly shocking.

ARSH -- Arousal Shift [TEMPORAL TRAJECTORY]

Arousal Shift measures the change in physiological activation over the duration of the clip, tracking if the speaker is waking up or crashing.

Samples: 175 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=25)

Value	Description
0	The physiological trajectory begins in a state of high activation before entirely collapsing into a near-comatose, heavily sedated acoustic state. The speaker's biological engine suffers a total loss of vitality, perfectly mirroring a person passing out or entirely surrendering to absolute exhaustion. The energy vector is a catastrophic, completely unrecoverable downward plunge.
1	The audio tracks a highly significant de-escalation of autonomic nervous system energy, moving rapidly from high alert down to a state of deep, relative calm. This deep drop in physical intensity perfectly mimics a massive 'sigh of relief' or the immediate, breathless aftermath of a resolved conflict. The acoustic engine spins down aggressively and audibly.
2	The physiological vector exhibits a subtle, highly comforting settling of bodily tension as the audio smoothly progresses to its end. The vocal energy softens and relaxes slightly towards the final words, clearly suggesting the speaker is becoming physically more comfortable or gently tired. The acoustic momentum gently decelerates into a highly relaxed bodily posture.
3	The autonomic activation level remains entirely static and completely constant, demonstrating exactly zero change in baseline physiological arousal. The exact energy level at the start of the clip is perfectly mirrored at the end, firmly locking the speaker into an unyielding physical state. The biological engine neither revs up nor physically spins down.
4	The speech trajectory tracks a subtle, highly engaging 'perking up' of physiological alertness and rising bodily energy. The speaker sounds perceptibly more awake, highly interested, and physically engaged by the end of the phrase than they did at the exact beginning. The acoustic engine gently revs up, applying a light, highly active forward-moving pressure.
5	The energy vector traces a clear, highly driving, and intensely aggressive escalation of physical intensity and massive autonomic arousal. The vocal folds are pushed significantly harder, louder, and faster as the speaker actively works themselves up into a heightened state of severe excitement or anger. The biological engine rapidly accelerates toward an explosive boiling point.
6	The physiological trajectory is a completely violent, 0-to-100 explosion of extreme autonomic mobilization that fundamentally shatters the acoustic baseline. The clip begins in a state of calm and literally detonates into a full scream, shout, or full-blown panic attack within seconds. The energy vector is a highly terrifying, totally combustible biological acceleration.

VALS -- Valence Shift [TEMPORAL TRAJECTORY]

Valence Shift tracks the change in emotional polarity across the segment, measuring whether the psychological state is moving toward positive or negative.

Samples: 175 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=25)

Value	Description
0	The emotional trajectory suffers a catastrophic, heavily tragic turn, plunging instantly from a positive or neutral state into deep despair or absolute, sheer horror. The acoustic 'smile' is violently stripped away mid-sentence, instantly replaced by the heavy, trembling tones of absolute psychological agony or shock. It is the highly distressing sonic equivalent of receiving completely devastating, life-altering news.
1	The mood vector tracks a highly significant souring and intense darkening of the psychological state, rapidly deteriorating into distinct, heavy negativity. Initial optimism or polite interaction audibly curdles into deep doubt, aggressive hostility, or profound sadness by the end of the clip. The underlying 'light' completely and noticeably drains out of the acoustic signal.
2	The emotional arc features a subtle, highly perceptible clouding of the initial positive or perfectly neutral affective baseline. A distinct hint of worry, minor disappointment, or mild unease slowly creeps into the final acoustic phrases, gently pulling the mood slightly downward. The psychological vector leans gently but completely undeniably toward the negative.
3	The emotional polarity remains absolutely stable and heavily sustained throughout the entire duration of the targeted audio clip. There is exactly zero detectable shift in the overall affective state; a happy speaker ends distinctly happy, and an angry speaker ends distinctly angry. The psychological baseline is perfectly locked and highly consistent.
4	The affective trajectory traces a subtle, highly encouraging brightening of the emotional state, gently shifting towards a distinctly positive resolution. A highly perceptible hint of a smile, deep relief, or renewed optimism gracefully appears at the very end of the acoustic phrase. It sounds exactly like the sun gently breaking through heavy, oppressive emotional clouds.
5	The mood vector undergoes a clear, highly powerful resolution, actively lifting the speaker heavily out of a negative or neutral state into distinct, bright positivity. Heavy acoustic tension is visibly released, allowing the voice to become significantly warmer, much happier, and deeply relieved by the end of the audio. It perfectly captures the deeply satisfying acoustic sound of a complex problem being solved.
6	The emotional arc executes an extreme, completely euphoric redemption, rapidly launching the speaker from deep despair into absolute, manic joy. The psychological state violently swings from highly negative to totally ecstatic in a single, massive, and highly explosive acoustic trajectory. The audio captures a burst of highly overwhelming, completely unexpected emotional delight.

ATCK -- Attack [TEMPORAL PATTERN]

Attack measures the edge of the sound as it initiates from silence, describing how hard or soft the speaker hits the first syllable.

Samples: 175 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=25)

Value	Description
0	The acoustic onset is completely ghostly and practically invisible, oozing into existence so perfectly softly that the exact millisecond of initiation is impossible to pinpoint. The vocal folds close with absolute delicacy, executing a perfect, mathematically smooth fade-in from pure, dead silence. The audio entirely lacks even the slightest transient hit or percussive edge.
1	The speech initiates firmly with a highly audible rush of air and H-like friction before the actual fundamental tone gradually joins the acoustic signal. This heavily breath-first onset heavily dampens the initial vocal transient, making the exact start of the word sound deeply diffused, soft, and distinctly airy. The acoustic boundary is heavily smeared by leading turbulence.
2	The vocal onset is distinctly soft, highly gentle, and extremely polite, carefully avoiding any harsh percussive pops, clicks, or aggressive hits. The vocal folds perfectly meet with a smooth, highly controlled biological grace that initiates the sound wave cleanly but entirely without aggressive physical force. The initial acoustic edge is highly rounded and deeply comfortable.
3	The audio features a perfectly neutral, highly standard conversational onset that is appropriately confident without sounding physically aggressive. The vocal folds close with a healthy, completely unremarkable firmness that cleanly establishes the exact start of the acoustic phrase. It represents the highly functional, perfectly balanced baseline of healthy human speech initiation.
4	The onset is highly clear, distinctly square, and physically hard, executing a highly definite, perfectly exact start to the acoustic waveform. The vocal folds close with a highly perceptible, remarkably clean 'tap' that creates a strong, highly legible transient spike at the exact beginning of the word. The acoustic edge is sharp, highly precise, and immensely focused.
5	The sound violently erupts from perfect silence with a heavy, highly percussive 'bark' that physically punches the listener's ear. The vocal folds unexpectedly slam together with high energy, creating a deeply startling, deeply aggressive transient spike that heavily dominates the onset of the phrase. The acoustic attack is highly forceful, fast, and borderline hostile.
6	The acoustic onset is a violent, deeply explosive glottal slam that sounds exactly like a heavy, high-velocity biological vocal gunshot. The vocal folds are brutally and painfully crushed together before air aggressively forces them apart, creating an extremely hard, severely jarring transient spike. The initial edge of the sound wave is a highly aggressive, physically shocking acoustic impact.

LANG -- Language

Identifies the dominant language being spoken in the clip, focusing purely on the vocabulary, phonetics, and grammatical structure utilized.

Samples: 100 total (Americas=25, Asian_Pacific=25, European=25, Middle_Eastern_African=25)

Category	Description
Americas	The acoustic profile captures the specific vowel shifts, distinct rhoticity (R-sounds), and fluid consonant linking of languages native to or evolved within the Americas. The rhythm tends to be fluid and stress-timed, with highly active intonation curves and distinctly relaxed vowel reductions in unstressed syllables. The phonetic output reflects the unique historical blending of indigenous and colonial linguistic roots.
Asian_Pacific	The audio is heavily defined by the tonal pitch contours, distinct vowel lengths, and syllable-timed rhythms characteristic of Asian and Pacific languages. The fundamental frequency actively dictates lexical meaning, creating a highly melodic, mathematically structured acoustic map of pitch rises, falls, and flat tones. The phonetic inventory often features precise unaspirated stops and a unique distribution of nasal consonants.
European	The acoustic stream is characterized by the distinct consonant clusters, varied vowel inventories, and stress-timed or syllable-timed rhythms typical of European language families. The phonetic profile features a wide range of fricatives, plosives, and complex syllable structures common in English, German, French, or Slavic tongues. The overall prosody relies heavily on intonation contours to signal grammatical intent and emotional nuance.
Middle_Eastern_African	The speech signal is rich with deep pharyngeal and uvular consonants, creating a highly resonant, beautifully guttural acoustic texture typical of Middle Eastern and African languages. The rhythmic structure is highly complex, often featuring distinctly measured vowel durations and sharp, percussive consonant strikes. The phonetics rely heavily on active manipulation of the deep throat and back of the tongue.

ACNT -- Accent

The Accent dimension identifies the specific regional, social, or non-native pronunciation pattern. Accents are prosodic fingerprints involving vowel shifts, consonant realizations, and unique intonation contours.

Category	Description
Native_English_UK	The acoustic signal is defined by the crisp, highly articulated consonants and specifically unpronounced, 'dropped' post-vocalic 'R' sounds typical of many United Kingdom dialects. Vowels are often distinctly shaped, with high precision in the oral cavity and distinct pitch contours that frequently swoop dramatically at the ends of phrases. The physical delivery is highly active in the lips and front of the mouth.
Native_English_US	The audio features the distinct rhotic 'R' sounds, flattened 'T's (flapping), and specific vowel raising characteristic of United States English dialects. The rhythmic delivery is distinctly stress-timed, heavily reducing unstressed vowels into neutral 'schwa' sounds to create a rolling, highly connected acoustic momentum. The placement of the voice often favors a slightly wider, more relaxed oral posture.
Native_German	The speech is characterized by highly crisp, unvoiced plosives, distinctly precise glottal stops before initial vowels, and a firm, rhythmic acoustic timing. The pronunciation features a strong, guttural 'R' sound and beautifully rounded front vowels that require active, muscular lip involvement. The overall delivery is highly articulate, structured, and physically anchored in a solid, grounded resonance.
Non_Native_Asian	The vocal delivery is deeply influenced by the syllable-timed or tonal architecture of Asian languages, resulting in a highly measured, slightly staccato acoustic rhythm. The speaker may struggle to differentiate specific consonant pairs (like L and R) or may shorten vowel durations to match their native phonological grid. The overall prosody heavily relies on distinct pitch shifts rather than dynamic volume stress.
Non_Native_Romance	The audio reveals the acoustic interference of a native Romance language, heavily characterized by a distinct syllable-timed rhythm that gives every vowel equal duration. The speaker utilizes purely unaspirated stops (soft P, T, K) and distinctly trilled or tapped 'R' sounds that heavily flavor the target language. The overall intonation remains highly melodic, smoothly linking words together without harsh glottal stops.
Non_Native_Slavic	The acoustic fingerprint is heavily marked by a distinct lack of articles, flattened intonation contours, and highly active, deeply resonant consonant clusters. The speaker frequently unvoices final consonants and utilizes a heavily rolled or distinct 'R', bringing the robust, chest-heavy resonance of Slavic phonetics into the target language. The speech rhythm is highly grounded, slightly rigid, and physically dark.

RCQL -- Recording Quality

RCQL measures the technical integrity of the audio signal chain. It evaluates the microphone quality, sample rate fidelity, and absence of digital artifacts or distortion.

Samples: 175 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=25)

Value	Description
0	The audio file is severely corrupted and of completely abysmal technical fidelity, completely dominated by destructive digital artifacts, heavy clipping, and low-bitrate swirling. The acoustic signal sounds like a deeply degraded cellular connection passing through intense electronic interference, rendering the speech nearly unintelligible. The frequency spectrum is entirely decimated, lacking both high-end clarity and low-end stability.
1	The recording quality is highly amateur and significantly flawed, captured by a cheap, deeply inferior microphone with an incredibly narrow frequency response. The voice sounds severely muffled, highly tinny, and completely lacks any sense of acoustic depth or dynamic range. It is a completely flat, highly compromised signal highly typical of an old laptop's internal microphone.
2	The technical fidelity is completely below average but remains barely functional for basic human communication and intelligibility. The audio suffers from a distinct lack of professional clarity, often sounding slightly boxed-in, dynamically squashed, or poorly gain-staged. It is the highly unremarkable, slightly flawed sound of older consumer electronics.
3	The acoustic capture represents the highly standard, decent quality of modern consumer electronics, such as a standard smartphone voice memo or a clean web-conference call. The voice is completely legible and adequately clear, though it heavily lacks the rich, transparent polish and deep frequency response of a professional setup. It is a completely acceptable, strictly utilitarian recording.
4	The audio boasts highly good, prosumer-level recording quality, featuring a very clean signal path, appropriate leveling, and a pleasing overall acoustic clarity. The microphone captures a highly solid representation of the speaker's voice, avoiding any distracting distortion or harsh digital artifacts. While not absolutely pristine, it is a highly capable, very strong recording.
5	The technical fidelity is of distinct broadcast quality, delivering an incredibly crisp, highly clear, and deeply rich acoustic signal. The recording chain utilizes excellent pre-amplification and a high-quality microphone to perfectly capture the nuanced transients and full frequency spectrum of the voice. There is absolutely zero audible distortion, clipping, or technical imperfection.
6	The recording quality is completely pristine and of the absolute highest audiophile grade, offering a flawless, perfectly transparent window into the acoustic event. The signal path is entirely immaculate, captured by ultra-high-end studio microphones that deliver profound depth, perfect high-frequency 'air', and zero noise floor. It is an absolutely perfect, uncolored, and breathtakingly realistic technical capture.

BKGN -- Background Noise

BKGN measures the cleanliness of the silence behind the voice. It tracks the presence of room tone, hiss, environmental ambience, or interfering sounds.

Samples: 175 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=25)

Value	Description
0	The environment behind the speaker is an absolute, completely dead acoustic vacuum, totally devoid of any room tone, hiss, or background interference. The recording was clearly captured in a heavily sound-treated, professional isolation booth that perfectly absorbs all ambient reflections. The silence between the words is completely pure, black, and strictly dry.
1	The background noise is almost entirely imperceptible, consisting only of the absolute faintest, microscopic trace of a high-quality room tone. The environmental isolation is incredibly strong, creating a highly focused acoustic space that remains deeply quiet without feeling completely unnatural or dead. The speaker's voice completely and effortlessly dominates the audio file.
2	A highly slight, entirely ignorable layer of background hiss or extremely faint room ambience is technically audible in the spaces between the speech. This noise floor perfectly mimics a highly quiet, undisturbed home office or a perfectly still residential bedroom. The environmental sound is completely benign and does not actively compete with the vocal signal.
3	The audio contains a distinctly noticeable, clearly audible room ambience that actively reveals the physical size and acoustic reflections of the recording space. The listener can easily detect the steady, low-level hum of air conditioning, the faint drone of distant traffic, or the slight echo of an untreated room. The background is completely present but remains strictly secondary to the voice.
4	The recording environment is highly noisy and actively distracting, heavily polluting the audio with strong, distinct background clatter, loud street traffic, or heavy mechanical hums. The environmental sounds aggressively compete with the primary vocal signal, forcing the listener to actively separate the speech from the chaos. The signal-to-noise ratio is severely compromised.
5	The background environment is completely chaotic and deeply overwhelming, severely drowning the speaker in a highly loud, aggressively disruptive acoustic setting. The audio captures the intense noise of a wildly busy cafe, a heavy construction site, or a howling windstorm that actively masks the phonetic details of the voice. The speech is heavily buried within a highly hostile acoustic environment.
6	The environmental noise entirely overpowers the human voice, completely burying the acoustic signal under a catastrophic wall of static, chaos, or explosive sound. The background interference is significantly louder than the speaker, rendering the speech completely impossible to isolate or understand. The audio is essentially a recording of severe environmental noise with a faint human voice trapped inside.

EXPL -- Explicitness

EXPL measures the suitability of the content for general audiences, tracking the presence of profanity, sexual themes, violence, or disturbing subject matter.

Samples: 175 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=25)

Value	Description
0	The audio content is completely innocent and perfectly safe for all possible audiences, entirely devoid of any swearing, controversy, or suggestive themes. The vocabulary is highly clean, strictly informational, and totally appropriate for a completely unrestricted, G-rated public broadcast. The semantic and acoustic delivery entirely avoids any mature or abrasive elements.
1	The speech contains only the most highly mild, completely benign slang or extremely low-level interpersonal tension. The content remains entirely safe for standard workplaces and children, completely avoiding any harsh profanity or explicit mature concepts. The audio is highly gentle, slightly casual, but fundamentally very clean.
2	The audio captures standard, everyday adult conversation that may contain highly moderate, completely un-graphic mature themes or extremely mild, culturally acceptable profanity. The vocabulary reflects the slightly rougher edges of realistic human interaction without crossing into overtly shocking or explicit territory. It is exactly comparable to a standard PG-13 cinematic dialogue.
3	The speech is distinctly mature and highly restricted, heavily featuring strong, undeniable language or actively discussing complex, non-graphic adult themes. The content is absolutely Not Suitable for Work in a strict professional setting, requiring a clear, active warning for younger or highly sensitive listeners. The vocabulary is significantly rougher and explicitly mature.
4	The audio content is highly explicit and deeply aggressive, completely saturated with frequent, very strong profanity, highly graphic descriptions of violence, or overtly sexual dialogue. The semantic themes are entirely adult-oriented and designed to shock, perfectly reflecting an R-rated or highly restricted media environment. The linguistic delivery completely abandons all polite social filtering.
5	The recording contains extremely heavy, highly graphic, and distinctly disturbing adult content that is deeply inappropriate for almost all standard listening environments. The audio captures highly intense, explicit themes such as realistic pornographic sounds, deeply brutal auditory violence, or highly aggressive, completely unfiltered hate speech. It is a severely restricted, highly extreme acoustic file.
6	The content represents the absolute extreme of deeply disturbing, highly illegal, or completely banned auditory material. The audio captures catastrophic, highly traumatizing events such as actual physical torture, severe gore, or completely prohibited, highly restricted extreme adult interactions. The semantic and acoustic nature of the file is profoundly dark and strictly universally condemned.

R_CHST -- Chest Resonance

Chest Resonance describes sound that vibrates primarily in the sternum and ribcage, creating warmth, rumble, and acoustic weight.

Samples: 165 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=15)

Value	Description
0	The acoustic signal is entirely devoid of low-end weight, sounding completely disembodied, tiny, or entirely trapped in the upper nasal and head cavities. The fundamental frequency completely fails to engage the sternum or ribcage, producing a thin, helium-like voice that lacks any physical grounding. It is an extremely lightweight, physically unsupported acoustic footprint.
1	The vocal delivery contains only the absolute faintest trace of bodily warmth, resulting in a distinctly thin adult voice that lacks natural authority or physical grounding. The low-frequency harmonics are severely underdeveloped, preventing the audio from projecting any significant acoustic mass. The resonance is highly superficial and feels distinctly disconnected from the speaker's torso.
2	The voice utilizes a standard, lighter speaking resonance where the chest is only mildly participating in the overall acoustic output. The sound is predominantly centered in the mouth, offering just enough low-mid frequency support to sound functionally human without projecting heavy warmth. It is a highly common, slightly physically detached conversational tone.
3	The acoustic profile features a perfectly balanced, deeply healthy, and beautifully grounded speaking voice that fully utilizes the natural resonance of the body. The listener can clearly hear the sternum supporting the fundamental tone, providing a robust, highly stable low-mid frequency presence. It represents the ideal, highly functional standard for male and female broadcast warmth.
4	The vocal signal exhibits a distinct, highly pronounced chest bias, generating a deep, soothing resonance that creates a sympathetic vibration in the listener’s own chest. The low-frequency harmonics are heavily emphasized, producing the classic, highly attractive warmth of a late-night radio DJ. The voice feels physically large, deeply comfortable, and highly authoritative.
5	The audio is dominated by massive physical projection and a heavily booming, intensely thick chest resonance that completely saturates the low-end spectrum. The sound wave feels physically heavy and deeply robust, effortlessly commanding the acoustic space with immense, deeply grounded power. It mirrors the larger-than-life, incredibly resonant depth of a classic theatrical hero.
6	The fundamental tone is driven by an exaggerated, floor-shaking, sub-bass seismic depth that relies almost entirely on extreme chest and thoracic vibration. The voice is an overwhelming, low-frequency rumble that actively sacrifices high-end definition for pure, monstrous physical mass. The acoustic output mimics the impossible, hyper-resonant growl of a cinematic dragon or giant.

R_THRT -- Throat Resonance

Throat Resonance describes sound trapped in the pharynx and neck, creating darkness, a swallowed quality, or guttural textures.

Samples: 175 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=25)

Value	Description
0	The biological vocal tract is completely open, highly relaxed, and entirely free of any pharyngeal constriction or 'swallowed' acoustic qualities. The sound wave effortlessly bypasses the throat, projecting forward with perfect clarity and absolutely zero guttural interference. The physical pipeline is flawlessly unobstructed, allowing for a totally pure, unchoked emission.
1	The voice exhibits only the absolute faintest, microscopic trace of coloring in the back of the throat, perhaps indicating very mild vocal fatigue. The primary resonance remains highly forward and clear, with the pharyngeal cavity barely influencing the overall acoustic geometry. The subtle darkness is almost entirely negligible.
2	A distinctly noticeable pharyngeal resonance pulls the voice slightly backward, preventing the acoustic signal from achieving full, forward oral projection. The fundamental tone sounds subtly 'thick' and slightly muffled, as if the speaker is holding the sound too far back in their neck. The resulting audio is functionally intelligible but slightly physically trapped.
3	The acoustic profile is clearly centered in the neck, characterized by a highly guttural, distinctly dark, and physically 'thick' vocal texture. The sound wave feels slightly obstructed by the root of the tongue, preventing bright, forward harmonics from escaping the vocal tract. The delivery has a heavy, slightly rough, and distinctly 'swallowed' quality.
4	The vocal tract suffers from significant pharyngeal constriction, creating a highly choked, extremely dark acoustic signal that struggles to project outward. The speaker sounds exactly as if they are artificially puffing out their neck to mimic a brute or a large amphibian. The sound is heavily trapped behind the tongue, resulting in a deeply muffled, highly unnatural resonance.
5	The acoustic output is intensely strangled and heavily trapped within a tightly constricted throat, producing a highly distorted, deeply unnatural vocal tone. The voice strongly resembles a classic 'zombie' groan or a heavily exaggerated cartoon character, completely lacking any clean, forward articulation. The physical effort required to force sound out of the neck is blatantly audible.
6	The fundamental tone is completely imploded and entirely swallowed by the throat, resulting in deeply unintelligible, heavily gargling, and violently grunting acoustic textures. The mouth and forward resonating cavities are rendered completely irrelevant as the sound is violently trapped in the deep pharynx. It is the absolute extreme of a physically choked, completely non-communicative biological noise.

R_ORAL -- Oral Resonance

Oral Resonance describes sound centered in the middle of the mouth, representing the neutral human standard for clear articulation.

Samples: 152 total (0=18, 1=25, 2=25, 3=25, 4=25, 5=25, 6=9)

Value	Description
0	The acoustic signal is entirely displaced from the mouth, completely trapped in the extreme nasal, deep throat, or heavy chest cavities. There is absolutely no neutral vowel formation occurring in the oral space, severely compromising standard human phonetic legibility. The sound wave entirely bypasses the primary articulatory chamber of the lips and tongue.
1	The vocal placement is highly vague and acoustically unfocused, making it difficult to pinpoint but definitely not centered cleanly within the oral cavity. The speech lacks the sharp, front-facing clarity required for highly intelligible communication, instead wandering into adjacent resonant spaces. The mouth is underutilized as the primary acoustic amplifier.
2	The resonance is mostly oral but is heavily colored and distinctly influenced by competing nasal twang or heavy chest tones. While the speaker is clearly using their mouth to shape the vowels, the acoustic center of gravity is audibly split, preventing a perfectly clean, neutral delivery. The resulting sound is functional but acoustically impure.
3	The voice perfectly exemplifies the 'everyman' oral placement, utilizing the mouth as the highly balanced, primary acoustic resonating chamber. The vowels are cleanly shaped, the consonants are appropriately crisp, and the overall signal is exceptionally clear and easy to understand. It is the absolute, perfectly neutral baseline of standard human speech.
4	The acoustic delivery features very clear, highly forward articulation where the sound feels exactly as if it is sitting directly on the speaker's lips. The oral cavity is highly active, precisely shaping the phonemes to ensure maximum forward projection and pristine phonetic legibility. The voice 'bites' through the air cleanly without needing to shout.
5	The speaker utilizes heavily exaggerated, wide-open mouth shapes, producing a highly round, distinctly over-enunciated acoustic profile. The voice sounds exactly as if the speaker is trying to communicate with a lip-reader, overly prioritizing massive oral space over natural conversational flow. The resulting audio is highly clear but physically unnatural and overly 'large.'
6	The mouth functions as an overwhelming, absolute megaphone, serving as the sole resonating cavity while completely abandoning all chest depth or nasal nuance. The sound is violently projected entirely through a stiff, wide-open oral posture, creating a highly utilitarian, deeply un-nuanced acoustic blast. The delivery is aggressively loud, totally flat, and heavily forward.

R_MASK -- Mask Resonance

Mask Resonance describes sound vibrating in the hard palate, cheekbones, and teeth, creating ping, buzz, and intense forward projection.

Samples: 157 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=7)

Value	Description
0	The acoustic signal is completely dull and physically dead, entirely lacking the high-frequency 'ring' or 'buzz' required to project across a space. The sound wave literally falls out of the mouth and dies instantly, possessing zero forward momentum or hard-palate reflection. It is an incredibly flat, intensely soft, and deeply un-projected vocal emission.
1	The vocal delivery contains only a highly microscopic amount of forward focus, making the voice incredibly soft and remarkably difficult to hear in a noisy environment. The high frequencies fail to properly engage the hard structures of the face, resulting in a heavily muffled, deeply recessive acoustic profile. The sound lacks any distinct 'edge' or physical carry.
2	The voice features a standard, highly present level of acoustic clarity, utilizing just enough mask resonance to easily carry a conversation across a small table. The fundamental tone is supported by a gentle, highly natural forward buzz that ensures standard intelligibility without sounding aggressive. The projection is completely comfortable and functionally appropriate.
3	The acoustic output is characterized by a highly clear, distinctly 'pingy' quality that easily and effortlessly cuts right through heavy background noise. The speaker expertly utilizes the hard palate to create a bright, highly focused buzz, perfectly mirroring the highly efficient projection of an experienced classroom teacher. The audio is bright, sharp, and intensely legible.
4	The vocal signal boasts a strong, highly metallic ring that indicates a heavily projected, highly efficient theatrical 'belt.' The speaker is effortlessly 'calling out' to someone across a large street without actually resorting to the physical strain of shouting. The mask resonance is highly active, turning the face into a powerful acoustic amplifier.
5	The intense forward focus creates a highly piercing, deeply aggressive acoustic signal that essentially drills directly into the listener's ear. The voice is violently bright and heavily concentrated in the cheekbones and teeth, exactly like a stage actor fiercely projecting to the absolute back row of a theater. The sound is incredibly loud, sharp, and physically demanding.
6	The fundamental tone is a pure, laser-focused, siren-like vibration located entirely in the extreme front of the face and hard palate. The acoustic output is uncomfortably bright, violently loud, and completely devoid of any softening chest warmth, mimicking the extreme projection of a town crier or an opera singer hitting a peak note. It is an overwhelming, highly piercing acoustic blast.

R_NASL -- Nasal Resonance

Nasal Resonance describes sound vibrating in the nasal cavity (velum lowered), creating twang, whine, or a pinched acoustic quality.

Samples: 104 total (0=3, 1=25, 2=25, 3=25, 4=16, 5=10)

Value	Description
0	The nasal passage is completely blocked, forcing all air entirely through the mouth and violently turning 'M' and 'N' consonants into 'B' and 'D' sounds. The acoustic signal sounds exactly like a speaker suffering from a severe, highly congested head cold, completely lacking any upper-cavity airflow. The voice is deeply denasal, flat, and physically obstructed.
1	The vocal delivery is completely clean and entirely devoid of any audible nasality, representing the absolute gold standard for clear, highly resonant broadcasting tone. The velum is perfectly raised, expertly directing the entire acoustic signal smoothly through the oral cavity without any high-frequency nasal leakage. The sound is rich, pure, and flawlessly balanced.
2	A highly slight, highly natural acoustic 'twang' subtly colors the voice, introducing a gentle brightness commonly found in American Midwest or Australian dialects. This minor nasal involvement adds a distinct, highly recognizable character to the vowels without compromising the overall clarity or warmth of the speech. The resonance remains deeply functional and culturally standard.
3	The acoustic profile features a distinctly obvious nasal coloring, making the fundamental tone sound significantly 'higher' and 'thinner' than its actual pitch. The voice strongly resembles a stereotypical, slightly pinched 'nerd' caricature, where a significant portion of the sound wave is constantly escaping through the nose. The overall delivery is noticeably bright and mildly squeezed.
4	The vocal tract suffers from strong, highly deliberate nasal constriction, creating a deeply pinched, aggressively whiny acoustic signal. The speaker sounds exactly as if they are physically pinching their nostrils shut while speaking, forcing the sound into a highly tight, uncomfortable frequency band. The audio is distinctly abrasive and lacks all bodily depth.
5	The nasal resonance is intensely unpleasant and highly piercing, heavily dominating the acoustic spectrum with a deeply complaining, aggressively nagging tone. The sound wave is violently squeezed into the upper nasal cavity, completely stripping away any soothing low-mid frequencies and replacing them with a harsh, highly irritating buzz. The voice is physically exhausting to listen to.
6	The acoustic output completely loses all oral resonance, resulting in a hypernasal, violently bleating audio signal that sounds exactly like a cartoon witch or a distressed sheep. All consonants and vowels are heavily distorted by massive amounts of air aggressively escaping through the nose. It is an extreme, highly un-human, and deeply abrasive acoustic caricature.

R_HEAD -- Head Resonance

Head Resonance describes sound vibrating in the upper sinuses and skull, creating lightness, loft, and acoustic air.

Samples: 165 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=15)

Value	Description
0	The fundamental tone is entirely stuck deep in the physical body, completely failing to access any upper harmonics, lightness, or high-frequency lift. The voice is heavily grounded and entirely incapable of producing a soft, floating, or delicate acoustic texture. The sound wave is a dense, heavily weighted block of pure, low-mid frequency resonance.
1	The vocal delivery contains only a microscopic amount of acoustic lift, resulting in a highly heavy, distinctly 'meat and potatoes' everyday voice. The higher frequencies are severely muted, preventing the signal from achieving any true sense of airy brilliance or delicate upper-skull vibration. The voice is functionally solid but entirely lacks high-end grace.
2	The acoustic profile features a highly balanced, gentle upper harmonic ring that beautifully complements a voice still firmly rooted in the chest. The speaker successfully accesses the lower head cavities to add a pleasing, highly natural brightness to their speech without losing their physical grounding. The resonance is slightly lifted but deeply stable.
3	The audio is characterized by a distinct, highly gentle, 'floating' speaking voice that naturally rests in the upper resonant cavities of the skull. This highly soothing, physically light acoustic texture is typical of soft-spoken librarians or therapists aiming to project deep calm. The voice is entirely devoid of aggressive low-end rumble, favoring a highly delicate airiness.
4	The vocal signal is highly airy and deeply lofty, creating a distinct psychoacoustic illusion that the voice is literally floating slightly above the speaker's head. The fundamental tone is heavily dominated by upper sinus resonance, perfectly mimicking the breathy, highly lifted delivery of Marilyn Monroe or a gentle choirboy. The sound is incredibly pure, soft, and highly elevated.
5	The speaker utilizes their distinct 'head voice' or falsetto register almost exclusively, producing an incredibly hollow, highly pure, and deeply elevated acoustic signal. The entire lower frequency spectrum is completely abandoned, leaving only a highly delicate, heavily ringing upper harmonic structure. The voice completely lacks heavy physical weight, feeling highly ethereal and distinctly fragile.
6	The fundamental frequency is pushed to the absolute extreme upper limits of human head resonance, leaving only the highest possible, deeply disembodied harmonic overtones. The acoustic signal sounds exactly like a tiny, high-frequency whistle, a delicate fairy, or a completely weightless ghost. It entirely lacks all human biological weight, existing purely as a microscopic, piercing vibration.

R_MIXD -- Mixed Resonance

Mixed Resonance describes the technical ability to seamlessly blend Chest and Mask/Head resonance, indicating vocal mastery and a connected range.

Samples: 150 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25)

Value	Description
0	The vocal delivery is entirely unbalanced, violently breaking or flipping wildly between isolated chest resonance and isolated head resonance. The speaker is completely incapable of blending the two cavities, resulting in an acoustic signal that is either entirely heavy or entirely thin, never both simultaneously. The voice sounds physically fractured and highly uncoordinated.
1	The acoustic profile reveals a highly weak, deeply struggling attempt to connect the lower and upper vocal registers. The speaker occasionally manages to slightly blend the resonances, but the overall sound remains distinctly uneven, highly fragile, and acoustically unsupported. The transition between high and low frequencies is heavily labored and physically clumsy.
2	The listener can clearly hear both the chest and head resonances operating simultaneously, but they remain audibly split and highly un-integrated. The acoustic signal lacks a smooth, unified center, sounding like two distinct voices attempting to speak at the exact same time without actually blending together. The vocal mix is functional but noticeably disjointed.
3	The voice demonstrates a highly connected, perfectly standard level of speech-level mixing, successfully uniting the chest and head into one cohesive biological instrument. The acoustic transition between low and high frequencies is highly smooth, completely hiding the mechanical 'break' in the vocal folds. It is a highly reliable, completely healthy, and functionally unified vocal tone.
4	The acoustic delivery is incredibly rich and beautifully blended, effortlessly combining the deep, warm gravity of the chest with the bright, highly focused ring of the mask. The speaker possesses excellent, highly developed vocal control, creating a highly complex, deeply satisfying harmonic footprint that is both powerful and brilliant. The voice sounds highly expensive and deeply resonant.
5	The vocal signal boasts a highly powerful, full-spectrum Broadway or professional-grade sound that perfectly maxes out the potential of the human vocal tract. The speaker seamlessly integrates massive chest depth with intense, highly piercing forward mask resonance to create the absolute 'ideal' voice for high-end narration or acting. The acoustic output is immensely strong, flawless, and deeply commanding.
6	The resonance achieves an impossibly huge, super-resonant 'wall of sound' where every single biological cavity is vibrating at maximum, perfect harmonic capacity. The acoustic signal perfectly mirrors the legendary, overwhelming power of an elite operatic tenor or a historically massive orator. The voice is an absolute, perfectly unified force of nature that effortlessly dominates the entire frequency spectrum.

S_CASU -- Casual Style

Casual Speaking Style describes informal, low-effort, 'lazy' speech where the priority is purely comfort, not acoustic clarity.

Samples: 141 total (0=1, 1=25, 2=25, 3=25, 4=25, 5=25, 6=15)

Value	Description
0	The vocal delivery is entirely robotic, violently rigid, and deeply artificial, perfectly mimicking a machine reading a complex legal disclaimer. The speaker hits every single consonant with mathematically perfect, highly unnatural precision, completely avoiding any organic slurring or relaxed pacing. The acoustic signal is the absolute, unyielding antithesis of human comfort and casual interaction.
1	The speech is distinctly stiff and highly rehearsed, sounding exactly like an individual nervously attempting to appear relaxed during a high-stakes job interview. The speaker tries to adopt a casual tone, but underlying tension forces the articulation to remain overly precise and distinctly rigid. The resulting audio feels highly forced, deeply uncomfortable, and entirely unnatural.
2	The acoustic delivery is highly polite, distinctly friendly, but heavily guarded, perfectly capturing a standard, low-stakes interaction with a complete stranger. The speaker maintains a very safe, highly clear articulatory baseline without fully relaxing their physical posture or dropping their professional facade. The voice is functionally warm but completely lacks true, intimate looseness.
3	The audio features the highly relaxed, deeply comfortable speech of a person casually talking with a close friend in a completely safe environment. The speaker's physical shoulders are audibly down, resulting in minor, highly natural slurring of words like 'gonna' or 'wanna' without losing overall legibility. The acoustic rhythm is highly organic, completely unforced, and deeply human.
4	The vocal style is incredibly loose and highly informal, actively dropping final 'g' consonants and heavily blending words together into a continuous, deeply relaxed stream. The speaker completely abandons formal posture, projecting a deeply comfortable, 'feet on the table' acoustic energy that prioritizes sheer ease over clear diction. The audio is incredibly laid-back and heavily casual.
5	The speech is distinctly lazy and heavily mumbled, with the speaker barely opening their mouth to physically shape the acoustic vowels. Words bleed heavily into one another, creating the highly indistinct, deeply low-effort acoustic texture typical of modern 'mumble-core' acting. The vocal apparatus is operating at the absolute minimum required physical effort for basic phonation.
6	The vocal delivery perfectly embodies the 'slacker' or 'drunk' archetype, featuring extreme, highly exaggerated vowel elongation and severely slurred, completely mushy consonants. The speaker completely lacks any sense of urgency, producing a deeply dragging, physically exhausted acoustic signal that barely qualifies as structured language. It is the absolute, hyper-relaxed vocal equivalent of wearing heavy sweatpants.

S_CONV -- Conversational Style

Conversational Style describes interactive, social, feedback-oriented speech designed to actively invite a response from a partner.

Samples: 159 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=9)

Value	Description
0	The delivery is a completely rigid, heavily aggressive monologue that functions exactly like a massive, uninterrupted wall of acoustic sound. The speaker is violently talking 'at' the listener, completely failing to leave even a single micro-second gap for a response, agreement, or natural feedback. The audio is the absolute antithesis of a two-way social interaction.
1	The speech is highly detached and completely un-interactive, sounding exactly like a person coldly reading a script into an empty void without visualizing a partner. The vocal tone entirely lacks the subtle, upward-inflected 'checking' markers that normally invite a listener into the acoustic space. The delivery is distinctly solitary, highly flat, and completely non-social.
2	The acoustic output strongly resembles a theatrical soliloquy; the speaker is clearly thinking out loud, but is completely unconcerned with whether another person is actually listening. The intonation is highly internal and slightly wandering, lacking the sharp, outward-facing projection required to actively engage a conversational partner. The audio feels highly personal but distinctly disconnected.
3	The audio captures a perfectly standard, highly functional two-way conversational flow where the speaker actively and naturally shares the acoustic space. The delivery features highly subtle, perfectly timed micro-gaps and gentle intonation lifts designed specifically to allow the listener to nod, agree, or interject. The social rhythm is highly balanced, deeply respectful, and effortlessly interactive.
4	The vocal style is incredibly engaged and highly social, heavily utilizing distinct 'active listening' markers such as a deep smile in the voice and highly inviting, rising pitch contours. The speaker is highly focused on maintaining a strong, deeply warm interpersonal connection, actively pulling the listener into the acoustic momentum of the story. The audio is highly vibrant and deeply collaborative.
5	The speech is intensely interactive, with the speaker constantly and actively checking in with the listener by appending rapid questions like 'You know?' or 'Right?' to the ends of phrases. The vocal delivery radiates high warmth, deep connection, and a highly eager desire to ensure the partner is completely following the acoustic narrative. The social energy is highly forward and deeply inclusive.
6	The delivery is an over-the-top, hyper-social caricature of extreme extroversion, featuring rapid-fire turn-taking and a desperate, highly aggressive need to maintain the interpersonal bond. The speaker constantly interrupts their own laughter to connect, using massive pitch swings to violently force the listener's continued engagement. It is an overwhelming, highly manic display of absolute conversational desperation.

S_FORM -- Formal Style

Formal Style describes professional, structured, 'high' register speech focused heavily on precision, protocol, and gravity.

Samples: 167 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=17)

Value	Description
0	The speech is intensely vulgar, highly aggressive, and deeply rooted in entirely informal, unfiltered street slang. The speaker heavily utilizes severe grammatical errors, harsh swearing, and a highly relaxed, slouching acoustic posture that violently rejects any concept of polite society. The audio is the absolute, hostile antithesis of structured, formal protocol.
1	The vocal delivery is entirely casual and completely everyday, heavily relying on standard linguistic contractions like 'can't' and 'won't' to maintain a highly relaxed flow. The speaker completely lacks the stiff, upright acoustic posture required for formal address, favoring a highly comfortable, deeply unpretentious vocal tone. The speech is highly functional but entirely unpolished.
2	The acoustic profile captures a highly polite, distinctly 'customer service' voice that is very nice, deeply non-threatening, but entirely lacks rigid, high-level formality. The speaker uses correct grammar and a gentle tone, but the overall delivery remains highly accessible, deeply soft, and distinctly un-authoritative. The vocal posture is functional, highly friendly, and slightly subservient.
3	The speech perfectly embodies a highly serious, distinctly boardroom-professional tone that completely eliminates all slang and strictly utilizes standard, correct grammar. The delivery is highly clear, deeply focused, and acoustically crisp, projecting a highly competent, deeply reliable corporate or academic authority. The voice sounds highly structured, deeply organized, and strictly business-oriented.
4	The vocal style is highly official and deeply rigid, perfectly mimicking the intense, highly controlled delivery of a major press statement or a high-stakes news briefing. The speaker completely avoids all contractions, utilizing 'cannot' and 'will not' to enforce a highly commanding, deeply upright acoustic posture. The sheer physical discipline of the speaker is blatantly audible in the highly precise audio wave.
5	The delivery is intensely ceremonial and deeply grave, sounding exactly like a high-ranking judge or a priest delivering a highly important, ancient address. The speaker utilizes archaic, highly elevated vocabulary, an incredibly slow, deliberate tempo, and a deeply resonant, highly serious acoustic gravity. The audio feels highly momentous, deeply traditional, and entirely unapproachable.
6	The speech is an absolute, hyper-rigid caricature of royal protocol, perfectly mimicking the flawlessly stiff, deeply transatlantic or aristocratic affect of a monarch's address. The speaker completely abandons all natural human warmth, transforming their voice into the highly sterile, perfectly polished acoustic manifestation of an ancient, unyielding institution. The delivery is completely flawless, totally cold, and deeply imperial.

S_DRAM -- Dramatic Style

Dramatic Style describes theatrical, emotional acting, measuring the degree to which the speaker is 'putting on a show.'

Samples: 175 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=25)

Value	Description
0	The delivery is incredibly deadpan, heavily bored, and completely devoid of any emotional affect, sounding exactly like a person falling asleep while reading a mundane grocery list. The speaker actively suppresses any dramatic instinct, maintaining a highly flat, violently uninteresting acoustic profile regardless of the actual words being spoken. The audio is the absolute, unyielding antithesis of theatrical performance.
1	The acoustic signal captures distinctly bad, highly flat acting where the speaker is mechanically reading the lines without genuinely feeling or projecting the underlying emotion. The vocal tone is highly wooden, deeply unconvincing, and completely fails to physically elevate the semantic meaning of the text. The dramatic attempt is highly visible but completely acoustically unsuccessful.
2	The speech perfectly embodies highly natural, deeply realistic cinematic acting where the emotional truth is distinctly present but highly subtle and entirely internalized. The speaker does not 'push' the audio for the microphone, instead allowing genuine, highly grounded human feelings to gently color the acoustic resonance. The dramatic performance feels deeply authentic, highly intimate, and completely unforced.
3	The vocal delivery utilizes a highly emotive, distinctly 'stage' acting style where the underlying emotion is actively and physically projected so the back row of a theater can clearly hear it. The acoustic performance is highly distinct, deeply felt, and entirely deliberate, making the speaker's emotional state completely undeniable to the listener. The audio is highly vibrant, distinctly performed, and slightly larger than life.
4	The speech is characterized by high drama and intense, deeply Shakespearean theatricality, utilizing massive, sweeping dynamic shifts from a highly tense whisper to a full-throated, ringing shout. The speaker completely commands the acoustic space, delivering the text with a highly stylized, deeply powerful, and intensely passionate vocal commitment. The dramatic arc is highly epic, totally uninhibited, and physically demanding.
5	The audio is heavily melodramatic and deeply exaggerated, perfectly mimicking the highly saturated, over-the-top acting style of a classic daytime soap opera. The speaker heavily relies on loud, highly audible gasps, a deeply trembling voice, and an exaggerated, highly artificial vocal vibrato placed specifically on emotional trigger words. The performance is highly entertaining but deeply unrealistic and heavily stylized.
6	The vocal delivery is an absolute, hyper-operatic caricature of 'chewing the scenery,' featuring a speaker violently screaming to the heavens and shattering glass with pure, unadulterated, raw emotion. The dramatic choices are completely over-the-top, deeply hyperbolic, and wildly unrestrained, completely abandoning all human realism in favor of maximum, explosive acoustic impact. The audio is a highly chaotic, wildly entertaining theatrical meltdown.

S_NARR -- Narrator Style

Narrator Style describes storytelling intended for an audience, measuring the objective, descriptive, and engaging qualities of the read.

Samples: 154 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=4)

Value	Description
0	The speaker is heavily stumbling and deeply confused, sounding exactly like a person painfully reading a highly complex text out loud for the very first time. The delivery is completely ruined by severe mispronunciations, highly awkward, completely unnatural pauses, and a total lack of forward narrative momentum. The acoustic flow is deeply broken, completely failing to guide the listener through the story.
1	The acoustic signal captures the highly monotonous, deeply focused delivery of a young child reading aloud in a classroom setting. The speaker's entire cognitive capacity is completely dedicated to simply pronouncing the words, resulting in a highly flat, totally un-nuanced vocal tone that completely ignores the emotional arc of the text. The audio is functionally legible but entirely lacks narrative engagement.
2	The speech perfectly mimics a highly dry, deeply academic lecture or the incredibly boring, strictly informative reading of a complex technical manual. The vocal delivery is completely objective, highly steady, and entirely devoid of any entertaining or dramatic flair. While the acoustic information is successfully transmitted, the deeply sterile tone actively discourages active listening.
3	The vocal style is highly warm, deeply clear, and perfectly engaging, perfectly matching the highly comforting tone of a parent reading a beloved story to a child. The speaker utilizes gentle, highly effective character differentiation and a highly pleasing, easily followable acoustic rhythm to keep the listener deeply invested. The narrative delivery is highly intimate, deeply organic, and beautifully paced.
4	The audio features the highly flawless, deeply polished delivery of an elite, professional audiobook voice actor operating at the absolute peak of their craft. The speaker utilizes perfect, highly calculated pacing, a highly subtle, deeply engaging vocal 'melody,' and a deeply warm, highly resonant acoustic baseline. The narrative performance is highly immersive, completely effortless, and acoustically beautiful.
5	The delivery is highly epic and deeply resonant, perfectly embodying the massive, deeply grave acoustic weight of a classic, high-budget movie trailer narration. The speaker utilizes a highly pushed, deeply chest-heavy 'In a world...' vocal placement, projecting immense importance, deep tension, and highly stylized dramatic authority. The audio is incredibly thick, highly commanding, and deeply cinematic.
6	The vocal style represents the absolute, omniscient 'Voice of God' archetype, perfectly mimicking the infinitely wise, deeply smoothed, and perfectly resonant delivery of a legendary narrator like Morgan Freeman. The speaker's tone is completely devoid of all human anxiety, projecting an absolute, completely unshakeable acoustic authority over the nature of reality itself. The audio is a deeply hypnotic, perfectly flawless, and overwhelmingly comforting narrative masterpiece.

S_NEWS -- Newsreader Style

Newsreader Style describes the specific cadence of broadcast journalism, measuring punchy, objective, and rhythmic delivery.

Samples: 165 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=15)

Value	Description
0	The speech is an entirely unorganized, highly rambling stream of consciousness completely dominated by confused 'ums' and highly erratic, deeply broken thought processes. The speaker entirely lacks any ability to structure their acoustic output, resulting in a highly messy, deeply unprofessional delivery that completely fails to convey hard information. It is the absolute, chaotic antithesis of polished broadcast journalism.
1	The delivery is completely conversational and highly casual, sounding exactly like a person informally relaying a funny story to a close friend over drinks. The vocal tone entirely lacks the highly structured, deeply projected, and sharply objective acoustic markers required for a formal news read. The audio is highly relaxed, deeply subjective, and completely un-journalistic.
2	The acoustic profile captures the highly opinionated, distinctly loose, and slightly aggressive delivery typical of a modern YouTube commentator or amateur blogger. The speaker is highly energetic and completely legible, but heavily relies on a highly subjective, deeply informal vocal tone that entirely lacks traditional broadcast neutrality. The speech is highly engaging but strictly amateur in its journalistic posture.
3	The audio perfectly mimics the highly clear, distinctly projected, and slightly breathless urgency of a professional 'on the scene' field reporter. The speaker maintains a highly objective, deeply informative acoustic baseline while allowing the chaotic energy of their supposed environment to naturally elevate their vocal volume and tempo. The delivery is highly active, distinctly journalistic, and deeply present.
4	The speech features the highly polished, completely flawless delivery of a high-end studio news anchor operating with a perfectly 'neutral' broadcast accent. The speaker utilizes perfect, highly crisp diction and a deeply steady, highly predictable acoustic rhythm to effortlessly deliver complex information. The vocal tone is highly objective, deeply trustworthy, and perfectly structured.
5	The vocal style is incredibly heavy and deeply serious, radiating the massive, highly comforting acoustic gravitas of a legendary, 'Voice of Truth' historical anchorman. The speaker utilizes a deeply resonant, highly grounded vocal placement that projects immense, entirely unshakeable authority and deep journalistic integrity. The audio commands absolute, highly respectful attention from the listener.
6	The delivery is an absolute, highly exaggerated caricature of a 1970s local news anchorman, completely dominated by a highly artificial, forcibly deepened vocal tone. The speaker utilizes a highly rigid, deeply exaggerated 'sing-song' cadence that ends every single sentence with a massive, highly predictable downward pitch swoop. The acoustic performance is a highly stylized, deeply unnatural broadcast satire.

S_TECH -- Teacher/Didactic Style

Teacher/Didactic Style describes explaining complex concepts, measuring slow, clear delivery and the checking of listener understanding.

Samples: 147 total (0=15, 1=25, 2=25, 3=25, 4=25, 5=25, 6=7)

Value	Description
0	The speech is deeply obscure, highly rushed, and entirely composed of heavily mumbled, highly dense technical jargon that actively confuses the listener. The speaker makes absolutely zero attempt to clarify, slow down, or translate their highly complex thoughts into a digestible acoustic format. The delivery is completely hostile to the concept of teaching or shared understanding.
1	The delivery is highly unclear and completely un-paced, featuring a speaker talking at a completely normal, highly rapid conversational speed about deeply complex, highly difficult topics. The speaker entirely fails to utilize crucial acoustic pauses or vocal emphasis to help the listener process the dense flow of information. The audio is highly functional for the speaker but completely useless for a student.
2	The acoustic profile perfectly captures the highly dry, deeply monotone, and incredibly boring delivery of an uninspired, deeply detached university professor. The speaker executes a massive, highly flat acoustic data dump, completely ignoring the emotional engagement or active comprehension of the listener. The vocal tone is completely sterile, highly academic, and deeply un-engaging.
3	The speech features the highly clear, deeply upbeat, and distinctly helpful vocal tone of an excellent, highly engaging modern YouTube tutorial or digital explainer. The speaker utilizes a highly friendly, deeply accessible acoustic rhythm, breaking down complex concepts with a bright, highly encouraging vocal placement. The delivery is highly effective, deeply welcoming, and perfectly paced.
4	The vocal style is intensely instructional and highly precise, perfectly separating every single phrase and concept with highly deliberate, deeply clear acoustic pauses. The speaker sounds exactly like they are carefully reading a highly critical, step-by-step emergency manual, ensuring absolutely perfect phonetic legibility and total comprehension. The delivery is highly clinical, totally unambiguous, and deeply structured.
5	The delivery becomes highly patronizing and deeply condescending, featuring a speaker who is actively over-simplifying their acoustic output to 'mansplain' or talk down to the listener. The vocal tone utilizes an artificially sweet, highly elongated rhythm that clearly implies the listener is deeply unintelligent or highly incapable. The audio is technically clear but highly insulting and deeply frustrating.
6	The speech is an absolute, highly exaggerated caricature of an overly enthusiastic preschool teacher actively treating the listener like a literal toddler. The speaker utilizes an incredibly slow, deeply agonizing tempo, massive, highly unnatural pitch swings, and highly exaggerated, deeply forced high-valence 'checking' questions. The acoustic delivery is a totally overwhelming, highly infantilizing educational satire.

S_AUTH -- Authoritative Style

Authoritative Style describes commanding, dominant speech that actively requires obedience and compliance from the listener.

Samples: 163 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=13)

Value	Description
0	The vocal delivery is entirely submissive, highly trembling, and deeply apologetic, featuring an extremely high, highly fragile pitch that actively begs for the listener's approval. The speaker completely surrenders all acoustic and social power, utilizing a highly questioning, deeply unsure tone that entirely lacks any forward projection. The audio is the absolute, highly anxious antithesis of commanding authority.
1	The speech is distinctly unsure and deeply lacking in personal conviction, sounding exactly like a highly nervous speaker gently offering a weak suggestion. The vocal tone entirely fails to physically anchor the sentence, utilizing soft, completely un-projected acoustic endings that actively invite the listener to disagree or interrupt. The delivery is highly passive and completely non-dominant.
2	The acoustic profile captures a highly balanced, completely egalitarian interaction among absolute peers, where the speaker confidently offers a suggestion without demanding compliance. The vocal tone is completely firm and highly clear, but entirely lacks the heavy, downward-inflected acoustic weight required to issue a true command. The social dynamic is highly cooperative and strictly horizontal.
3	The delivery perfectly embodies the highly firm, deeply expected compliance of a competent corporate manager issuing a standard, highly clear directive. The speaker utilizes a highly grounded, deeply steady vocal placement that confidently assumes the listener will immediately obey the instruction without question. The audio is highly professional, distinctly authoritative, and completely devoid of emotional anger.
4	The vocal style is highly commanding and intensely projected, perfectly mimicking the loud, highly aggressive, and deeply forceful delivery of a police officer or military captain. The speaker utilizes highly shortened, deeply percussive sentences that physically strike the listener's ear, completely demanding immediate, highly absolute physical obedience. The acoustic posture is highly dominant, deeply intense, and entirely unyielding.
5	The delivery captures the highly aggressive, deeply overwhelming, and highly percussive vocal style of a powerful dictator or demagogue intensely hyping a massive rally. The speaker heavily utilizes violent, screaming projection, a highly repetitive, deeply hypnotic acoustic rhythm, and massive, completely intimidating volume to totally crush all opposition. The audio is a deeply terrifying, highly potent display of absolute vocal dominance.
6	The speech represents the absolute, terrifying peak of human interpersonal dominance, perfectly capturing a highly aggressive drill sergeant violently screaming directly into the listener's ear. The vocal folds are pushed to their absolute biological limit, producing a highly abrasive, deeply deafening acoustic blast where pure, unadulterated intimidation is the sole, overwhelming goal. The delivery is an absolute, highly traumatic vocal assault.

S_PLAY -- Playful Style

Playful Style describes fun, lighthearted, high valence, and joking speech, measuring the degree of mischief and joy in the delivery.

Samples: 175 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=25)

Value	Description
0	The delivery is incredibly grim, heavily darkened, and deeply serious, perfectly capturing the heavy, highly devastating acoustic weight of a funeral director or a terminal medical diagnosis. The vocal tone is completely stripped of all light, warmth, and upward pitch movement, projecting an absolute, highly unyielding aura of deep tragedy. The audio is the complete, totally devastating antithesis of playfulness.
1	The speech is strictly business, highly serious, and deeply unsmiling, perfectly mimicking the highly tense, totally humorless acoustic environment of a severe tax audit. The speaker utilizes a highly flat, deeply objective vocal placement that entirely refuses to engage in any lighthearted banter or subtle emotional warmth. The delivery is highly functional, completely sterile, and entirely un-playful.
2	The acoustic profile captures a highly dry, deeply deadpan sense of humor where the actual semantic words may be funny, but the vocal delivery remains completely flat and totally un-amused. The speaker actively suppresses any acoustic 'smile,' utilizing a highly monotone, distinctly un-energetic rhythm to deliver the punchline with zero emotional affect. The audio is highly ironic and completely devoid of bright joy.
3	The vocal delivery is distinctly light, highly friendly, and beautifully warm, featuring a highly audible, deeply genuine 'smile' naturally embedded within the acoustic resonance. The speaker utilizes gentle, upward-inflected pitch contours and a highly relaxed rhythm to create a deeply pleasant, highly inviting atmosphere of casual, friendly banter. The overall tone is highly positive and deeply comfortable.
4	The speech is highly joyful and clearly joking, utilizing a highly distinct, deeply rhythmic 'setup-and-punchline' acoustic cadence that actively invites the listener to laugh. The vocal tone is highly energetic, deeply bouncy, and completely saturated with bright, highly positive resonance and clear, audible chuckles. The delivery is actively entertaining, highly humorous, and deeply fun.
5	The delivery is intensely mischievous, highly secretive, and deeply energetic, perfectly capturing the highly giggly, entirely teasing vocal tone of a dedicated prankster. The speaker utilizes highly rapid, deeply unpredictable pitch shifts and a highly breathy, deeply excited acoustic placement to project a highly infectious sense of impending chaos. The audio is wildly fun, highly active, and deeply playful.
6	The vocal style is an absolute, hyper-whimsical caricature of pure, unadulterated joy, perfectly mimicking the completely magical, totally unpredictable delivery of a character like Willy Wonka. The speaker utilizes a wildly bouncing acoustic rhythm, impossible, highly exaggerated pitch jumps, and a deeply manic, completely explosive sense of pure delight. The audio is a totally overwhelming, deeply ecstatic manifestation of absolute play.

S_CART -- Cartoonish Style

Cartoonish Style describes animation voice acting, measuring exaggerated parameters and 'funny' voices that defy normal human physics.

Samples: 175 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=25)

Value	Description
0	The speech is completely hyper-real, deeply gritty, and entirely bound by the strict, highly mundane physical laws of a highly serious documentary or a stark, unpolished drama. The vocal delivery actively rejects any form of theatrical exaggeration, resulting in an uncomfortably genuine, completely un-stylized acoustic signal that feels entirely raw. It is the absolute, highly grounded antithesis of a cartoon voice.
1	The vocal delivery perfectly captures completely natural, entirely unremarkable standard human speech, entirely devoid of any artificial 'put-on' characterizations or highly exaggerated acoustic traits. The resonance, pitch, and rhythm are completely biologically standard, producing a highly functional, deeply typical audio signal. The voice sounds exactly like a completely normal person speaking in a completely normal environment.
2	The acoustic profile features a highly slight, distinctly noticeable 'put-on' character voice, perfectly mimicking a person casually performing a mild impression at a highly relaxed party. The speaker slightly alters their natural pitch or resonance, but the overall delivery remains heavily grounded in normal human physics and completely recognizable as an amateur attempt. The voice is slightly altered but deeply human.
3	The speech perfectly embodies a highly distinctive, deeply animated sitcom character voice that is entirely unique but still strictly bound by the acoustic physics of the human vocal tract. The speaker utilizes highly consistent, distinctly stylized vowel shapes or a highly specific, deeply recognizable pitch placement that instantly defines the character's persona. The delivery is highly professional, deeply colorful, and functionally grounded.
4	The vocal style is intensely wacky, highly energetic, and deeply exaggerated, heavily utilizing highly funny, completely unnatural biological noises and rapid, extreme pitch shifting. The speaker completely abandons standard human resonance, producing a highly bright, deeply squawking, or intensely rubbery acoustic signal typical of a high-energy children's cartoon. The audio is wildly colorful, highly elastic, and deeply un-natural.
5	The delivery becomes completely zany and deeply chaotic, featuring a voice that actively and violently defies all standard laws of vocal physics and biological elasticity. The speaker executes highly rapid, completely jarring shifts from a massive, deeply guttural growl to an impossibly high, deeply piercing squeak within a single sentence. The acoustic performance is wildly unpredictable, highly extreme, and deeply exhausting.
6	The speech is an absolute, hyper-exaggerated 'Looney Tunes' caricature where the human voice entirely transforms into a highly complex, completely impossible sound effect. The speaker utilizes completely impossible acoustic stretches, violent, highly percussive pops, deeply piercing whistles, and absolute, highly manic vocal caricatures that completely shatter the boundaries of human biology. The audio is a totally overwhelming, highly legendary masterclass in extreme vocal manipulation.

S_ASMR -- ASMR Style

ASMR Style describes Autonomous Sensory Meridian Response delivery, measuring extreme acoustic intimacy, texture focus, and deep quiet.

Samples: 158 total (0=8, 1=25, 2=25, 3=25, 4=25, 5=25, 6=25)

Value	Description
0	The delivery is completely explosive, violently loud, and deeply distant, perfectly mimicking a highly aggressive sports announcer screaming at the absolute top of their lungs across a massive, echoing stadium. The acoustic signal is entirely built on extreme volume, massive physical distance, and completely overwhelming, highly abrasive vocal projection. The audio is the absolute, highly hostile antithesis of quiet, intimate ASMR.
1	The speech is distinctly loud and completely un-intimate, utilizing the highly projected, deeply resonant normal speaking volume required to be heard clearly across a highly crowded, noisy room. The vocal tone is completely public, highly forward, and entirely lacks any soft, delicate, or deeply relaxing acoustic textures. The delivery is strictly functional, highly bright, and completely un-soothing.
2	The acoustic profile captures a highly quiet, deeply polite, and distinctly hushed conversation perfectly suited for a highly intimate, closely-seated dinner in a quiet restaurant. The speaker significantly lowers their overall decibel level, but the voice remains completely fully voiced and entirely lacks the hyper-focused, deeply textural mouth sounds of true ASMR. The audio is highly gentle, deeply respectful, and comfortably soft.
3	The vocal delivery perfectly embodies the highly soothing, deeply gentle 'Bob Ross' acoustic zone, utilizing a highly slow, completely relaxed, and deeply soft-spoken vocal placement. The speaker remains fully voiced, but completely rounds off all harsh transients and heavily emphasizes a deeply warm, highly comforting low-mid frequency resonance. The audio is deeply relaxing, highly meditative, and beautifully smooth.
4	The speech is a perfectly standard, highly clear, and deeply functional whisper, where the speaker completely disengages the vocal folds to produce a highly airy, entirely unvoiced acoustic signal. The delivery is highly close to the microphone, creating a distinctly intimate, deeply quiet audio texture that completely avoids any aggressive popping or highly harsh sibilance. The audio is a completely pure, highly traditional whisper.
5	The delivery is intensely focused on highly specific, deeply satisfying acoustic tingles, heavily prioritizing highly exaggerated, deeply wet mouth sounds, sharp 'sk' clicks, and deeply audible, rhythmic breathing. The speaker completely weaponizes the microphone's proximity effect to capture the absolute microscopic biological textures of the mouth and tongue. The audio is highly textural, deeply intentional, and intensely physically triggering.
6	The vocal style is an absolute, hyper-invasive 'brain massage,' utilizing completely overwhelming binaural ear-to-ear movement, deeply wet lip smacks, and impossibly close, highly intimate whispering. The acoustic signal is so intensely close and deeply detailed that the listener physically feels as though the speaker is literally inside their own head. The audio is a totally overwhelming, deeply profound manifestation of extreme, hyper-textural intimacy.

S_WHIS -- Whisper-Talk Style

Whisper-Talk Style describes the reduction of vocal fold vibration for secrecy, intimacy, or dramatic acoustic effect.

Samples: 175 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=25)

Value	Description
0	The vocal delivery is completely full-voiced, deeply ringing, and highly resonant, utilizing the absolute maximum, highly efficient vibration of the vocal folds to achieve massive, loud projection. The speaker completely fills the acoustic space with a completely uninhibited, deeply powerful fundamental tone that entirely lacks any breathy or unvoiced qualities. The audio is the absolute, highly robust antithesis of a quiet whisper.
1	The speech is distinctly hushed and highly subdued, perfectly capturing a highly polite 'indoor voice' that significantly drops in volume while remaining completely, 100% fully voiced. The speaker actively suppresses their overall acoustic projection, but the fundamental tone remains highly solid, entirely unbroken, and completely devoid of any breathy leakage. The audio is highly quiet, deeply controlled, but structurally totally normal.
2	The acoustic profile perfectly captures the highly intimate, deeply breathy 'bedroom voice,' utilizing an exact, highly deliberate half-tone, half-air biological mixture. The vocal folds fail to close completely, allowing a highly continuous, deeply soothing stream of audible air turbulence to perfectly wrap around the fundamental pitch. The audio is deeply soft, highly diffused, and intensely romantic or highly secretive.
3	The delivery is a highly forced, deeply aggressive 'stage whisper,' entirely designed to sound exactly like a highly secretive whisper while remaining loud enough to be clearly heard across a large room. The speaker utilizes massive amounts of air pressure and highly exaggerated, deeply sharp consonant articulation to violently push the unvoiced sound wave forward. The audio is highly piercing, deeply intense, and physically highly demanding.
4	The speech is a perfectly standard, highly functional 'library whisper,' where the speaker completely disengages the vocal folds to ensure they are entirely unheard by anyone outside their immediate physical vicinity. The acoustic signal is completely unvoiced, highly gentle, and deeply quiet, relying entirely on the highly soft friction of breath passing over the articulators. The audio is highly polite, deeply stealthy, and completely unobtrusive.
5	The vocal style is intensely raspy and highly aggressive, perfectly mimicking the deeply tense, highly iconic 'Batman' whisper characterized by massive, painful-sounding throat constriction. The speaker heavily grinds the unvoiced air against the false vocal folds, producing a deeply rough, highly abrasive, and intensely threatening acoustic signal. The audio is highly forceful, deeply dark, and physically exhausting to produce.
6	The delivery is an absolute, completely ghostly, and totally unvoiced acoustic phenomenon, featuring exactly zero vocal fold vibration and an incredibly low, deeply haunting volume. The speaker utilizes pure, highly delicate air articulation, where only the absolute faintest, highly microscopic friction of the consonants is barely audible to the human ear. The audio is a totally ethereal, deeply chilling, and almost entirely invisible sonic footprint.

S_RANT -- Ranting/Angry Style

Ranting/Angry Style describes uninterrupted grievance, high energy, and the acoustic manifestation of aggressive complaint.

Samples: 175 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=25)

Value	Description
0	The delivery is highly diplomatic, deeply soothing, and completely focused on highly effective, perfectly calm conflict resolution. The speaker utilizes a highly measured, deeply gentle vocal tone that actively absorbs all tension and entirely diffuses any potential aggression within the acoustic space. The audio is the absolute, highly peaceful antithesis of an angry, high-energy rant.
1	The speech is completely calm, highly rational, and deeply polite, featuring a speaker distinctly stating a clear disagreement without allowing any aggressive emotion to leak into the audio. The vocal tone remains highly flat, completely objective, and entirely devoid of any raised volume, sharp transients, or highly heated rhythmic momentum. The delivery is completely controlled, highly professional, and entirely un-angry.
2	The acoustic profile captures a highly annoyed, deeply passive-aggressive complaint, characterized by a highly tight, slightly pinched vocal tone and heavily suppressed, deeply frustrated sighs. The speaker actively avoids a full explosion, instead utilizing highly sharp, slightly clipped consonants to clearly signal their deep, highly simmering irritation. The audio is highly tense, deeply uncomfortable, and distinctly sour.
3	The vocal delivery perfectly embodies a highly energetic, deeply fast-paced venting session, sounding exactly like a person rapidly complaining to a close friend about a highly frustrating day. The speaker utilizes a highly continuous, deeply rhythmic acoustic flow to actively and effectively release massive amounts of pent-up psychological tension. The audio is highly active, deeply cathartic, and completely uninhibited, but not truly hostile.
4	The speech is a highly heated, deeply aggressive argument, characterized by massive spikes in volume, highly violent vocal interruptions, and a speaker actively attempting to physically talk over their opponent. The vocal tone is deeply harsh, highly percussive, and completely saturated with intense, highly combative acoustic energy that actively assaults the listener's ear. The delivery is highly stressful, deeply angry, and completely unrelenting.
5	The delivery is an intense, highly rhythmic, and deeply overwhelming political tirade, perfectly mimicking a highly aggressive radio host delivering a massive, high-volume monologue. The speaker utilizes an incredibly fast, deeply repetitive acoustic cadence, continuously building massive emotional steam and explosive volume with every single sentence. The audio is highly overwhelming, deeply grievance-driven, and intensely forceful.
6	The vocal style is an absolute, completely unhinged caricature of a highly manic street preacher or a totally out-of-control, red-faced screaming meltdown. The speaker completely abandons all biological vocal safety, utilizing deeply violent, table-slamming volume, highly terrifying pitch breaks, and massive, near-heart-attack levels of pure, unadulterated grievance energy. The audio is a totally chaotic, deeply frightening display of absolute, explosive vocal rage.

S_STRY -- Storytelling Style

Storytelling Style describes the specific lilt of oral tradition, measuring narrative arc, acoustic suspense, and character delivery.

Samples: 160 total (0=21, 1=25, 2=25, 3=25, 4=25, 5=25, 6=14)

Value	Description
0	The delivery is completely robotic, highly sterile, and deeply mundane, perfectly mimicking a deeply bored individual painfully reading raw data from a highly complex spreadsheet. The vocal tone is completely flat, entirely devoid of any emotional arc, and totally lacks even the absolute faintest hint of a forward-moving narrative momentum. The audio is the absolute, highly utilitarian antithesis of engaging oral storytelling.
1	The speech is highly objective, completely chronological, and deeply reporter-like, simply stating a highly dry sequence of historical facts without applying any acoustic color or emotional weight. The speaker entirely avoids utilizing dramatic pauses, pitch sweeps, or character voices, resulting in a highly functional but deeply un-entertaining vocal delivery. The audio conveys the information cleanly but entirely fails to build any narrative suspense.
2	The acoustic profile captures a highly casual, deeply everyday anecdote, sounding exactly like a person quickly relaying a highly funny, entirely unpolished story to a friend at a loud bar. The speaker utilizes a completely natural, highly conversational rhythm that is highly engaging but entirely lacks the structured, deeply theatrical pacing of a true, formal storytelling performance. The delivery is highly relatable, deeply organic, and totally informal.
3	The vocal delivery perfectly embodies the highly engaging, deeply hypnotic 'campfire' storytelling tone, heavily utilizing deeply deliberate, highly pregnant pauses to build massive acoustic suspense. The speaker's voice possesses a highly distinct, deeply rhythmic lilt, gently pulling the listener into the narrative with a highly warm, distinctly theatrical vocal placement. The audio is deeply captivating, highly structured, and beautifully paced.
4	The speech is a highly performed, deeply professional narrative masterpiece, featuring a speaker actively 'doing the voices' with highly distinct, instantly recognizable character differentiation. The vocal tone flawlessly shifts between a highly objective, deeply resonant narrator baseline and highly colorful, deeply unique acoustic profiles for each individual character within the story. The delivery is highly entertaining, deeply immersive, and technically flawless.
5	The delivery utilizes a highly epic, deeply grave 'High Fantasy' acoustic placement, projecting an immense, ancient gravity and a deeply magical realism tone. The speaker utilizes a highly slowed, deeply resonant, and heavily chest-anchored vocal tone to imbue the narrative with massive, highly mythical importance. The audio is incredibly heavy, highly stylized, and deeply cinematic, perfectly suited for a massive, world-building saga.
6	The vocal style is an absolute, highly hypnotic Bardic performance, utilizing a deeply musical, almost completely chanted acoustic cadence that actively puts the listener into a deep trance. The speaker completely embodies the highly archetypal, deeply ancient 'Once upon a time' energy, using massive, sweeping pitch contours and perfect rhythmic poetry to totally overwhelm the listener's senses. The audio is a totally immersive, deeply legendary manifestation of pure, ancient oral tradition.

S_MONO -- Monologue Style

Monologue Style describes soliloquy, measuring the degree to which inner thoughts are spoken aloud and completely detached from a listener.

Samples: 175 total (0=25, 1=25, 2=25, 3=25, 4=25, 5=25, 6=25)

Value	Description
0	The speech is an incredibly active, highly engaged, and deeply interactive dialogue, featuring a speaker distinctly talking directly to a partner, actively waiting for answers, and highly focused on reading facial cues. The vocal tone is completely outward-facing, highly social, and entirely dependent on the continuous acoustic and emotional feedback of the listener. The audio is the absolute, highly collaborative antithesis of an isolated monologue.
1	The delivery is a highly projected, deeply public formal address, sounding exactly like a highly confident politician or leader delivering a massive speech to a huge crowd. While the speaker is technically the only person talking, the vocal tone is entirely outward-facing, highly performative, and distinctly designed to actively manipulate the emotions of a massive audience. The audio is completely public, deeply loud, and totally un-intimate.
2	The acoustic profile captures a highly casual, completely unstructured moment of a person simply thinking out loud or gently reflecting while actively working on a physical task. The speaker's volume is slightly lowered, and the vocal tone is highly wandering, indicating that they are technically speaking to themselves but without any deep, highly emotional introspection. The delivery is highly natural, deeply mundane, and completely un-dramatic.
3	The vocal delivery perfectly embodies the highly intimate, completely honest, and deeply unperformative tone of a person quietly reading a highly personal diary entry aloud. The speaker utilizes a highly soft, deeply vulnerable acoustic placement that completely strips away all social masks and formal projection. The audio is incredibly raw, highly private, and completely unconcerned with entertaining an external audience.
4	The speech is a highly dramatic, deeply intense theatrical soliloquy, perfectly capturing a highly skilled actor painfully processing massive, heavy emotions completely alone on a dark stage. The speaker utilizes highly powerful, deeply sweeping dynamic shifts and a highly resonant, deeply emotional vocal tone to fully explore their deepest, darkest internal thoughts. The delivery is highly performed, deeply tragic, and totally isolated.
5	The delivery is highly cinematic and deeply stylized, perfectly mimicking the highly polished, deeply resonant voiceover narration of a major film explaining the plot directly to the audience. The speaker utilizes a highly detached, deeply omniscient acoustic baseline that exists completely outside the physical reality of the scene, smoothly guiding the listener through the character's internal feelings. The audio is highly professional, deeply informative, and entirely disconnected from the physical environment.
6	The vocal style is an absolute, highly exaggerated caricature of a gritty, deeply stylized 'Noir' detective monologue, characterized by a completely isolated, deeply gravelly, and highly depressed inner-voice narration. The speaker entirely embodies a deeply cynical, totally exhausted acoustic posture, heavily implying they are the absolute last, deeply broken person left in the entire universe. The audio is a totally immersive, deeply dark, and highly theatrical manifestation of absolute, brooding isolation.

---

Validation Pipeline

Each audio sample in this dataset was validated through a multi-stage process:

1. FAISS Candidate Mining: Embedding-based search (gte-base-en-v1.5) identified top-20 SAE feature candidates per taxonomy bucket
2. LLM Relevance Filtering: Claude Haiku filtered candidates based on feature descriptions matching bucket definitions
3. Audio-Based Validation: Gemini and Qwen3-Omni listened to each candidate audio and confirmed/rejected its match to the target bucket
4. Duration Filtering: All samples are at least 3 seconds long
5. Deduplication: MD5-based deduplication removed identical audio

Audio Sources

Audio samples come from two primary sources:

• Taxonomy Audio (~32%): Curated audio specifically collected for voice taxonomy annotation
• SAE Feature Audio (~68%): Audio extracted from Sparse Autoencoder feature visualizations across multiple models (sae_emb020, sae_emb030, sae_maestrino100, sae_bude_whisper, clap_sae)

License

This dataset is released for research purposes.