Struct SsirConfig 

pub struct SsirConfig {

    pub sample_rate: f64,
    pub direct_sound_window_ms: (f64, f64),
    pub ler_window_ms: f64,
    pub energy_threshold: f64,
    pub doa_threshold_deg: f64,
    pub toa_threshold_ms: f64,
    pub min_segment_ms: f64,
    pub mixing_time_ms: Option<f64>,
    pub onset_window_ms: f64,
    pub final_segment_ms: f64,
    pub min_peak_distance_ms: f64,
    pub doa_bandpass_hz: (f64, f64),
    pub doa_bandpass_order: u32,
}

Configuration for SSIR (Spatial Segmentation of Impulse Response) analysis.

Default values correspond to the SSIR-Mk2 configuration from Pawlak & Lee (Applied Acoustics 249, 2026), Table 1.

Fields

sample_rate: f64

Sample rate in Hz

direct_sound_window_ms: (f64, f64)

Direct sound window: (pre, post) in ms relative to detected onset. Reflections within this window are excluded from detection. Default: (0.5, 3.5) — the direct sound typically occupies ~4ms.

ler_window_ms: f64

Local Energy Ratio analysis window length in ms. The RIR is divided into consecutive windows of this length. Default: 1.0 ms (48 samples @ 48kHz).

energy_threshold: f64

Energy threshold as a multiple of the per-window median energy. A sample is considered a reflection candidate if its energy exceeds this multiple of the window’s median energy. Default: 3.0

doa_threshold_deg: f64

Minimum angular distance (degrees) between consecutive reflections for them to be considered distinct events. Pairs below this threshold are merged. Default: 9.0 degrees. Only used with multi-channel (SRIR) input.

toa_threshold_ms: f64

Minimum time-of-arrival difference (ms) between consecutive reflections. Pairs closer than this are merged regardless of DOA. Default: 0.5 ms.

min_segment_ms: f64

Minimum segment duration (ms) for early reflections. Segments shorter than this are merged with the preceding segment. Default: 0.5 ms.

mixing_time_ms: Option<f64>

Mixing time in ms (boundary between early reflections and reverberant tail). If None, estimated automatically from the Schroeder decay curve. Default: None (auto-estimate, typical values: 30-50ms for small rooms).

onset_window_ms: f64

Pre-onset window length (ms) for refining segment boundaries. For each detected reflection, the onset is searched within [TOA - onset_window_ms, TOA]. Default: 0.5 ms.

final_segment_ms: f64

Duration (ms) of the optional final segment after the last detected event. Default: 2.0 ms.

min_peak_distance_ms: f64

Minimum peak distance (ms) for direct sound onset detection. Peaks closer than this are suppressed when searching for the direct sound. Default: 0.1 ms (5 samples @ 48kHz).

doa_bandpass_hz: (f64, f64)

Band-limiting frequency range (Hz) for DOA estimation from B-format channels.

The pseudo-intensity vector method is most reliable within a frequency band where spatial aliasing is low and wavelengths are short enough for directional resolution. Low frequencies have poor spatial resolution; high frequencies may alias depending on the microphone array.

Default: (500.0, 4000.0) — a commonly used range for first-order Ambisonics.

doa_bandpass_order: u32

Butterworth filter order for DOA band-limiting.

Applied as a zero-phase (filtfilt) bandpass, so the effective order is doubled. Default: 4 (effective 8th-order after forward-reverse filtering).

Implementations

impl SsirConfig

pub fn new(sample_rate: f64) -> Self

Create a config with the given sample rate and all other values at defaults.

Trait Implementations

impl Clone for SsirConfig

fn clone(&self) -> SsirConfig

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for SsirConfig

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Default for SsirConfig

fn default() -> Self

Returns the “default value” for a type.