surge-math 0.1.36-alpha.0

surge synthesizer -- dsp math used in synthesis
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
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147

ix!();

// Snabba sin(x) substitut
// work in progress
#[inline] pub fn sine_float_nowrap(x: f32) -> f32 
{
   // http://www.devmaster.net/forums/showthread.php?t=5784
    let four_over_pi:         f32 = (4.0 / PI) as f32;
    let minus_four_over_pi_2: f32 = (-4.0 / (PI * PI)) as f32;

    let y: f32 = four_over_pi * x + minus_four_over_pi_2 * x * x.abs();

    // EXTRA_PRECISION
    //  const float Q = 0.775;
    let p: f32 = 0.225;

    // Q * y + P * y * abs(y)
    p * (y * y.abs() - y) + y 
}

#[inline] pub fn sine_ps_nowrap(x: __m128) -> __m128 
{
    unsafe {
        let four_over_pi: __m128 = _mm_set1_ps((4.0 / PI) as f32);
        let minus_four_over_pi_2: __m128 = _mm_set1_ps((-4.0 / (PI * PI)) as f32);

        // todo wrap x [0..1] ?

        // y = B * x + C * x * abs(x);
        let y: __m128 = _mm_add_ps(
            _mm_mul_ps(four_over_pi, x), 
            _mm_mul_ps(
                _mm_mul_ps(minus_four_over_pi_2, x), 
                _mm_and_ps(m128_mask_absval![], x)));

        // EXTRA_PRECISION
        //  const float Q = 0.775;
        let p: __m128 = _mm_set1_ps(0.2250);

        _mm_add_ps(
            _mm_mul_ps(
                _mm_sub_ps(
                    _mm_mul_ps(
                        _mm_and_ps(
                            m128_mask_absval![], 
                            y), 
                        y), 
                    y), 
                p), 
            y)
    }
}

/// sin(x*pi)
#[inline] pub fn sine_xpi_ps_sse2(mut x: __m128) -> __m128 
{
    unsafe {
        let four:    __m128  = _mm_set1_ps(4.0);
        let mask:    __m128i = _mm_set1_epi32(0x01ffffff);
        let offset:  __m128i = _mm_set1_epi32(0x01000000);

        // wrap x
        x = _mm_cvtepi32_ps(
            _mm_sub_epi32(
                _mm_and_si128(
                    _mm_add_epi32(
                        offset, 
                        _mm_cvttps_epi32(x)), 
                    mask), 
                offset));

        let y: __m128 = _mm_mul_ps(
            four, 
            _mm_sub_ps(
                x, 
                _mm_mul_ps(
                    x, 
                    _mm_and_ps(
                        m128_mask_absval![], 
                        x))));


        let p: __m128 = _mm_set1_ps(0.2250);

        _mm_add_ps(
            _mm_mul_ps(
                _mm_sub_ps(
                    _mm_mul_ps(
                        _mm_and_ps(
                            m128_mask_absval![], 
                            y), 
                        y), 
                    y), 
                p), 
            y)
    }
}

#[cfg(target_argh = "x86")]
pub fn sine(x: __m64) -> __m64 
{
    unsafe {

        let xabs: __m64 = _mm_xor_si64(
            x, 
            _mm_srai_pi16(x, 15));

        let mut y: __m64 = _mm_subs_pi16(
            _mm_srai_pi16(x, 1), 
            _mm_mulhi_pi16(x, xabs));

        y = _mm_slli_pi16(y, 2);
        y = _mm_adds_pi16(y, y);

        let q: __m64 = _mm_set1_pi16(0x6333);
        let p: __m64 = _mm_set1_pi16(0x1CCD);

        let yabs: __m64 = _mm_xor_si64(
            y, 
            _mm_srai_pi16(y, 15));

        let y1: __m64 = _mm_mulhi_pi16(q, y);

        let y2: __m64 = _mm_mulhi_pi16(
            p, 
            _mm_slli_pi16(
                _mm_mulhi_pi16(y, yabs), 
                1));

        y = _mm_add_pi16(y1, y2);
        _mm_adds_pi16(y, y)
    }
}

/// 16-bit sine
#[cfg(target_argh = "x86_64")]
pub fn sine(x: i32) -> i32 
{
   x = ((x + 0x8000) & 0xffff) - 0x8000;

   let y: i32 = ((x << 2) - ((abs(x >> 1) * (x >> 1)) >> 11));
   let q: i32 = (0.7750 * 65536.0);
   let p: i32 = (0.2250 * 32768.0);

   ((q * y) >> 16) + (((((y >> 2) * abs(y >> 2)) >> 11) * p) >> 15)
}