Struct AmbeDecoder 

pub struct AmbeDecoder { /* private fields */ }

Stateful AMBE 3600×2400 voice frame decoder.

The AMBE codec uses inter-frame prediction: each frame’s gain and spectral magnitudes are delta-coded against the previous frame. This decoder maintains three parameter snapshots to support that:

• cur — parameters decoded from the current frame
• prev — previous frame’s parameters (before enhancement), used as the prediction reference for delta decoding
• prev_enhanced — previous frame’s parameters (after spectral enhancement), used as the cross-fade source during synthesis

Invariants

• Create one AmbeDecoder per voice stream (per D-STAR StreamId).
• Feed frames sequentially in receive order.
• Discard the decoder when the stream ends (VoiceEnd event).
• The decoder is deterministic: same input sequence always produces the same output.

Implementations

impl AmbeDecoder

pub const fn new() -> Self

Creates a new decoder with zeroed initial state.

The first decoded frame will use silence as its prediction reference, which may produce a brief transient. This matches the behavior of hardware DVSI vocoders.

pub fn decode_frame(&mut self, ambe: &[u8; 9]) -> [i16; 160]

Decodes a single 9-byte AMBE frame into 160 PCM samples.

Returns 160 signed 16-bit samples at 8000 Hz (20 ms of audio). A gain factor of 7.0 is applied and samples are clamped to ±32767 × 0.95 to match JMBE soft-clipping semantics.

If the frame contains excessive bit errors (more than the FEC can correct) or the decoder has hit the maximum repeat count, comfort noise is output instead of synthesized speech.

Trait Implementations

impl Clone for AmbeDecoder

fn clone(&self) -> AmbeDecoder

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for AmbeDecoder

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

Formats the value using the given formatter.

impl Default for AmbeDecoder

fn default() -> Self

Returns the “default value” for a type.