Create Pitch from Sound Directly (Autocorrelation) - Full Parameters

Create Pitch analysis using autocorrelation method with full control over all voicing parameters. Returns a raw external pointer instead of an R6 object.

**NEW in v4.0.1:** Exposes all voicing parameters that were previously only available in Tier 1 (Standard) API.

Usage

to_pitch_ac_direct(
  sound,
  time_step = 0,
  pitch_floor = 75,
  pitch_ceiling = 600,
  max_candidates = 15,
  very_accurate = FALSE,
  silence_threshold = 0.03,
  voicing_threshold = 0.45,
  octave_cost = 0.01,
  octave_jump_cost = 0.35,
  voiced_unvoiced_cost = 0.14,
  max_number_of_candidates = NULL
)

Arguments

sound: Sound object or external pointer
time_step: Time step (0 = auto, typically 0.75/pitch_floor)
pitch_floor: Minimum pitch (Hz, default 75)
pitch_ceiling: Maximum pitch (Hz, default 600)
max_candidates: Maximum number of pitch candidates (default 15)
very_accurate: Use accurate but slower method (default FALSE)
silence_threshold: Frames below this relative intensity are unvoiced (default 0.03)
voicing_threshold: Strength required for voiced decision (default 0.45)
octave_cost: Cost per octave in path finding (default 0.01)
octave_jump_cost: Cost for octave jumps (default 0.35)
voiced_unvoiced_cost: Cost for voicing transitions (default 0.14)
max_number_of_candidates: Maximum number of pitch candidates per frame

Value

External pointer to Pitch (NOT R6 object)

Examples

sound <- Sound$create_tone(frequency = 200, duration = 0.5)

# With custom voicing threshold (stricter voicing detection)
pitch_ptr <- to_pitch_ac_direct(sound, voicing_threshold = 0.6)

# Use with query functions
f0 <- get_pitch_value_direct(pitch_ptr, 0.25, "hertz", TRUE)