Skip to content

Source files: 6 | Classes: 11 | Methods: 6 | Enums: 0


GTOS.NaturalLanguage

GTPhoneme

struct

GTPhoneme - Universal phoneme primitive represented as SDF in 3D acoustic space.
This is THE fundamental unit for speech recognition and synthesis.
SDF Distance = Phonetic similarity:
- Negative: Inside phoneme region (close match)
- Zero: Exact phoneme center
- Positive: Outside phoneme region (poor match)
MIL-SPEC: Struct-based, zero allocation, 64-byte cache line aligned.
GPU-Ready: Can be uploaded to compute shader for parallel ray marching.

Source: GTPhoneme.cs

Constants and Fields

Category

PhonemeCategory

Manner

MannerOfArticulation

Place

PlaceOfArticulation

Voicing

Voicing

GTPhoneme3DVocoder

static class

3DVocoder - The Speaker for GTPhoneme SDF fields.
Renders PhonemeAudioEnvelope SDF fields to audio samples via GPU-accelerated
formant synthesis. This is THE component that converts phoneme primitives
to actual sound waves.
Architecture:
Input: PhonemeAudioEnvelope (3D SDF field + temporal/voicing/intensity envelopes)
Processing: GPU ray marching through SDF field at each time step
Output: Audio samples (float[] PCM)
MIL-SPEC compliant: Zero allocation, real-time synthesis, GPU-ready.

Source: GTPhoneme.cs

Methods

SpeakPhoneme

float[] SpeakPhoneme ( PhonemeAudioEnvelope envelope, float sampleRate = 48000.0f )

Synthesize audio from single PhonemeAudioEnvelope via formant synthesis.
Uses SDF field to guide formant frequencies over time.
This is the core "Speaker" function - converts SDF field to sound waves.
PhonemeAudioEnvelope to render.
Audio sample rate (Hz).
Audio samples (float[] PCM).

GTPhonemeDataset

static class

GTPhonemeDataset - Complete IPA phoneme coverage.
Every possible human vocalization represented as GTPhoneme SDF primitive.
Categories:
- Pulmonic Consonants (p, b, t, d, k, g, m, n, ŋ, f, v, s, z, ʃ, ʒ, h, l, r, w, j, etc.)
- Non-Pulmonic Consonants (Clicks: ʘ, ǀ, ǃ, ǂ, ǁ; Implosives: ɓ, ɗ, ɠ, ʛ)
- Vowels (i, e, ɛ, a, ɑ, ɔ, o, u, ɪ, ʊ, ə, ʌ, æ, etc.)
- Diphthongs (eɪ, aɪ, ɔɪ, aʊ, oʊ, ɪə, ɛə, ʊə)
- Tones (High, Mid, Low, Rising, Falling, Dipping, Peaking)
- Affricates (tʃ, dʒ, ts, dz, etc.)
Source: International Phonetic Alphabet (IPA) 2015 revision + common variants.

Source: GTPhoneme.cs

NaturalLanguageCoreAtomics

static class

Natural Language Core Atomics - computational primitives for linguistic analysis.
MIL-SPEC compliant: Zero allocation, struct-based, self-documenting.
Foundation for Speech Recognition, Speech Synthesis, Universal Translation.

Source: NaturalLanguageCoreAtomics.cs

Methods

CharToLetter

EnglishLetter CharToLetter ( char c )

Convert character to EnglishLetter enum.

NaturalLanguageGeneration

static class

Natural Language Generation - Speech synthesis via SDF envelope interpolation.
Core API for converting UNLOCK glyphs → phonemes → audio.

Source: NaturalLanguageGeneration.cs

Methods

MapUNLOCKGlyphToPhoneme

GTPhoneme? MapUNLOCKGlyphToPhoneme ( UNLOCK.UNLOCKGlyph glyph )

Map UNLOCK glyph to GTPhoneme for speech synthesis.
Links linguistic energetics to acoustic phonetics.
Examples:
A (ANIMATE) → /eɪ/ (diphthong)
E (ENERGY) → /iː/ (long i)
C (CENTER) → /siː/ (see)
This is the bridge between UNLOCK linguistic analysis and
physics-based speech synthesis.
UNLOCK glyph to map.
Corresponding GTPhoneme (or null if no mapping).

NaturalLanguageUnderstanding

static class

Natural Language Understanding - Speech recognition via SDF phoneme matching.
Core API for converting audio → phonemes → UNLOCK glyphs → linguistic meaning.

Source: NaturalLanguageUnderstanding.cs

Methods

RecognizePhonemeFromFormants

GTPhoneme? RecognizePhonemeFromFormants ( float f1_Hz, float f2_Hz, float f3_Hz, GTPhoneme[] phonemeDatabase, float maxDistance = 2.0f )

Recognize phoneme from formant frequencies via SDF ray marching.
Input: F1, F2, F3 formants extracted from audio (FFT window).
Output: Closest matching GTPhoneme (or null if no match within threshold).
This is GPU-parallelizable - all phonemes checked simultaneously.
Sub-millisecond recognition on typical hardware.
First formant frequency (Hz).
Second formant frequency (Hz).
Third formant frequency (Hz).
Array of GTPhoneme SDF fields to search.
Maximum SDF distance for valid match (default 2.0).
Closest matching GTPhoneme, or null if no match.

Phoneme

struct

Phoneme - single unit of sound in a language.
MIL-SPEC: Struct-based, zero allocation, cache-friendly.
Use: Phonetic transcription, speech recognition, pronunciation analysis.

Source: NaturalLanguageCoreAtomics.cs

Constants and Fields

Backness

VowelBackness

Description

string

DurationMs

float

Height

VowelHeight

IsConsonant

bool

Manner

MannerOfArticulation

PhonemeID

int

Place

PlaceOfArticulation

Roundedness

VowelRoundedness

Symbol

string

Voicing

Voicing

PhonemeAudioEnvelope

struct

PhonemeAudioEnvelope - SDF field representation for 3D vocoder network.
This is the elegant, GPU-friendly phoneme representation for synthesis.
Instead of raw waveforms or spectrograms, we represent phonemes as:
- 3D SDF field (formant space: F1 × F2 × F3)
- Temporal envelope (amplitude over time)
- Voicing envelope (voiced/voiceless transition)
- Intensity envelope (loudness over time)
This can be fed directly to a 3DVocoder Network for real-time synthesis.

Source: GTPhoneme.cs

Constants and Fields

DurationMs

float

Phoneme

GTPhoneme

SampleRateHz

float

PhoneticTranscription

struct

Phonetic transcription - sequence of phonemes representing pronunciation.
MIL-SPEC: Fixed-size array, zero allocation during use.

Source: NaturalLanguageCoreAtomics.cs

Constants and Fields

OrthographicForm

string

PhonemeCount

int

Phonemes

Phoneme[]

TotalDurationMs

float

GTOS.NaturalLanguage.Examples

GTPhoneme3DVocoderUsageExamples

static class

Usage examples for GTPhoneme3DVocoder - The Speaker for GTPhoneme.

Source: GTPhoneme3DVocoderUsageExamples.cs

Methods

Example1_SpeakSinglePhoneme

string Example1_SpeakSinglePhoneme ( )

Example 1: Speak a single phoneme.
Demonstrates the most basic use case - render a single GTPhoneme to audio.

NaturalLanguageUsageExamples

static class

Source: NaturalLanguageUsageExamples.cs

Methods

Example_AlphabetOperations

string Example_AlphabetOperations ( )

Example 1: English alphabet operations


Generated from GTOS Savants source -- 2026-03-22

SILVIA is a registered Trademark of Cognitive Code Corp.