libflo-audio 0.1.3

The official encoder/decoder for the flo audio format
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

/// Maximum positive value for 16-bit signed integer (2^15 - 1)
pub const I16_MAX_F32: f32 = 32767.0;

/// Minimum value for 16-bit signed integer (-2^15)
pub const I16_MIN_F32: f32 = -32768.0;

/// Maximum absolute value for 16-bit signed integer as f64
pub const I16_MAX_F64: f64 = 32767.0;

/// Number of discrete levels in a 16-bit integer codec (2^15)
/// The integer grid IS the PCM grid: 1.0 maps to 32768 (clamped to 32767) and
/// -1.0 maps to -32768. Keeping this identical to the normalized PCM/float
/// mapping is what makes lossless encode/decode round-trips bit-exact.
pub const I16_SCALE_F32: f32 = 32768.0;

/// Inverse of I16_SCALE_F32, used for int -> float conversion (1/32768)
pub const I16_TO_F32_SCALE: f32 = 1.0 / 32768.0;

/// Convert f32 sample to i32 for processing
/// Rounds to nearest so inverse conversion recovers the grid value exactly.
#[inline]
pub fn f32_to_i32(sample: f32) -> i32 {
    libm::roundf(sample * I16_SCALE_F32).clamp(I16_MIN_F32, I16_MAX_F32) as i32
}

/// Convert i32 sample to f32
#[inline]
pub fn i32_to_f32(sample: i32) -> f32 {
    sample as f32 * I16_TO_F32_SCALE
}

/// Convert f32 sample to i16
#[inline]
pub fn f32_to_i16(sample: f32) -> i16 {
    libm::roundf(sample * I16_SCALE_F32).clamp(I16_MIN_F32, I16_MAX_F32) as i16
}

/// Convert i16 sample to f32
#[inline]
pub fn i16_to_f32(sample: i16) -> f32 {
    sample as f32 * I16_TO_F32_SCALE
}