transcribe-rs 0.3.9

A simple library to help you transcribe audio
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109

//! Audio processing utilities for transcription.
//!
//! This module provides functions for reading and processing audio files
//! to prepare them for transcription engines.

use std::path::Path;

use crate::TranscribeError;

/// Number of samples per millisecond at 16 kHz.
pub const SAMPLES_PER_MS: usize = 16;

/// Read WAV file samples and convert them to the required format.
///
/// This function reads a WAV file and converts it to the format expected by
/// transcription engines: 16kHz sample rate, 16-bit samples, mono channel.
///
/// # Arguments
///
/// * `wav_path` - Path to the WAV file to read
///
/// # Returns
///
/// Returns a vector of f32 samples normalized to the range [-1.0, 1.0].
///
/// # Errors
///
/// This function will return an error if:
/// - The file cannot be opened or read
/// - The WAV format is incorrect (not 16kHz, 16-bit, mono)
/// - The samples cannot be converted to the expected format
///
/// # Examples
///
/// ```rust,no_run
/// use transcribe_rs::audio::read_wav_samples;
/// use std::path::Path;
///
/// let samples = read_wav_samples(Path::new("audio.wav"))?;
/// println!("Loaded {} samples", samples.len());
/// # Ok::<(), Box<dyn std::error::Error>>(())
/// ```
///
/// # Audio Requirements
///
/// The input WAV file must have:
/// - Sample rate: 16,000 Hz
/// - Bit depth: 16 bits per sample
/// - Channels: 1 (mono)
/// - Format: PCM integer samples
pub fn read_wav_samples(wav_path: &Path) -> Result<Vec<f32>, TranscribeError> {
    let mut reader = hound::WavReader::open(wav_path).map_err(|e| {
        TranscribeError::Audio(format!("failed to open {}: {}", wav_path.display(), e))
    })?;
    let spec = reader.spec();

    let expected_spec = hound::WavSpec {
        channels: 1,
        sample_rate: 16000,
        bits_per_sample: 16,
        sample_format: hound::SampleFormat::Int,
    };

    if spec.channels != expected_spec.channels {
        return Err(TranscribeError::Audio(format!(
            "Expected {} channels, found {}",
            expected_spec.channels, spec.channels
        )));
    }

    if spec.sample_rate != expected_spec.sample_rate {
        return Err(TranscribeError::Audio(format!(
            "Expected {} Hz sample rate, found {} Hz",
            expected_spec.sample_rate, spec.sample_rate
        )));
    }

    if spec.bits_per_sample != expected_spec.bits_per_sample {
        return Err(TranscribeError::Audio(format!(
            "Expected {} bits per sample, found {}",
            expected_spec.bits_per_sample, spec.bits_per_sample
        )));
    }

    if spec.sample_format != expected_spec.sample_format {
        return Err(TranscribeError::Audio(format!(
            "Expected Int sample format, found {:?}",
            spec.sample_format
        )));
    }

    let samples: Result<Vec<f32>, _> = reader
        .samples::<i16>()
        .map(|sample| sample.map(|s| s as f32 / i16::MAX as f32))
        .collect();

    Ok(samples?)
}

/// Prepend silence to audio samples.
///
/// Returns a new buffer with `silence_ms` milliseconds of zeros
/// followed by the original samples. Assumes 16 kHz sample rate.
pub fn prepend_silence(samples: &[f32], silence_ms: u32) -> Vec<f32> {
    let silence_len = silence_ms as usize * SAMPLES_PER_MS;
    let mut padded = vec![0.0f32; silence_len];
    padded.extend_from_slice(samples);
    padded
}