Unified Track Reading Interface — read

Reads either SSFF or JSTF files transparently based on file extension.

Usage

read_track(
  file,
  begin = 0,
  end = 0,
  samples = FALSE,
  validate = TRUE,
  zero_to_na = TRUE,
  tracks = NULL,
  threads = 1L,
  snap = c("none", "nearest")
)

Arguments

file: Path to track file (.f0, .fms, .vat, .vsj, etc.)
begin: Start of region to read (seconds, or samples if samples = TRUE). Default 0 = file start. For JSTF files, slices whose begin_time is before this value are excluded.
end: End of region to read (seconds, or samples if samples = TRUE). Default 0 = file end. For JSTF files, slices whose begin_time is after this value are excluded.
samples: Logical. If TRUE, begin/end are in samples (divided by sample_rate in the JSTF header); otherwise in seconds. Default FALSE.
validate: Logical, validate after reading (default: TRUE, JSTF only)
zero_to_na: Logical, SSFF only. If TRUE (default) stored values that are exactly 0 are returned as NA for every track that is not sampled audio. SSFF has no NULL/NA encoding and 0 is its substitute, so this is what makes "no value" distinguishable from a measured zero. Use read_ssff for the values exactly as stored.
tracks: Optional character vector of track names to read (SSFF only). NULL (default) reads every track in the file; unselected tracks are skipped without being converted.
threads: Number of threads used to convert large SSFF files (default 1, serial). Results are identical regardless of the value.
snap: SSFF only; ignored for JSTF files. One of "none" (default) or "nearest" — see read_ssff. Controls whether a single off-grid time point (begin == end, non-zero) errors (matching wrassp) or is rounded to the nearest frame.

Value

AsspDataObj (for SSFF) or JsonTrackObj (for JSTF)

Examples

if (FALSE) { # \dontrun{
# Read SSFF pitch track
pitch <- read_track("audio.f0")

# Read SSFF with time windowing
pitch <- read_track("audio.f0", begin = 1.0, end = 2.5)

# Read JSTF voice quality track
vq <- read_track("audio.vat")

# Both can be converted to data.frame
df1 <- as.data.frame(pitch)
df2 <- as.data.frame(vq)
} # }