-
Notifications
You must be signed in to change notification settings - Fork 15
/
buffer.go
254 lines (220 loc) · 6.79 KB
/
buffer.go
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
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
package beep
import (
"fmt"
"math"
"time"
"github.com/gopxl/beep/v2/internal/util"
)
// SampleRate is the number of samples per second.
type SampleRate int
// D returns the duration of n samples.
func (sr SampleRate) D(n int) time.Duration {
return time.Second * time.Duration(n) / time.Duration(sr)
}
// N returns the number of samples that last for d duration.
func (sr SampleRate) N(d time.Duration) int {
return int(d * time.Duration(sr) / time.Second)
}
// Format is the format of a Buffer or another audio source.
type Format struct {
// SampleRate is the number of samples per second.
SampleRate SampleRate
// NumChannels is the number of channels. The value of 1 is mono, the value of 2 is stereo.
// The samples should always be interleaved.
NumChannels int
// Precision is the number of bytes used to encode a single sample. Only values up to 6 work
// well, higher values lose precision due to floating point numbers.
Precision int
}
// Width returns the number of bytes per one frame (samples in all channels).
//
// This is equal to f.NumChannels * f.Precision.
func (f Format) Width() int {
return f.NumChannels * f.Precision
}
// EncodeSigned encodes a single sample in f.Width() bytes to p in signed format.
func (f Format) EncodeSigned(p []byte, sample [2]float64) (n int) {
return f.encode(true, p, sample)
}
// EncodeUnsigned encodes a single sample in f.Width() bytes to p in unsigned format.
func (f Format) EncodeUnsigned(p []byte, sample [2]float64) (n int) {
return f.encode(false, p, sample)
}
// DecodeSigned decodes a single sample encoded in f.Width() bytes from p in signed format.
func (f Format) DecodeSigned(p []byte) (sample [2]float64, n int) {
return f.decode(true, p)
}
// DecodeUnsigned decodes a single sample encoded in f.Width() bytes from p in unsigned format.
func (f Format) DecodeUnsigned(p []byte) (sample [2]float64, n int) {
return f.decode(false, p)
}
func (f Format) encode(signed bool, p []byte, sample [2]float64) (n int) {
switch {
case f.NumChannels == 1:
x := util.Clamp((sample[0]+sample[1])/2, -1, 1)
p = p[encodeFloat(signed, f.Precision, p, x):]
case f.NumChannels >= 2:
for c := range sample {
x := util.Clamp(sample[c], -1, 1)
p = p[encodeFloat(signed, f.Precision, p, x):]
}
for c := len(sample); c < f.NumChannels; c++ {
p = p[encodeFloat(signed, f.Precision, p, 0):]
}
default:
panic(fmt.Errorf("format: encode: invalid number of channels: %d", f.NumChannels))
}
return f.Width()
}
func (f Format) decode(signed bool, p []byte) (sample [2]float64, n int) {
switch {
case f.NumChannels == 1:
x, _ := decodeFloat(signed, f.Precision, p)
return [2]float64{x, x}, f.Width()
case f.NumChannels >= 2:
for c := range sample {
x, n := decodeFloat(signed, f.Precision, p)
sample[c] = x
p = p[n:]
}
for c := len(sample); c < f.NumChannels; c++ {
_, n := decodeFloat(signed, f.Precision, p)
p = p[n:]
}
return sample, f.Width()
default:
panic(fmt.Errorf("format: decode: invalid number of channels: %d", f.NumChannels))
}
}
func encodeFloat(signed bool, precision int, p []byte, x float64) (n int) {
var xUint64 uint64
if signed {
xUint64 = floatToSigned(precision, x)
} else {
xUint64 = floatToUnsigned(precision, x)
}
for i := 0; i < precision; i++ {
p[i] = byte(xUint64)
xUint64 >>= 8
}
return precision
}
func decodeFloat(signed bool, precision int, p []byte) (x float64, n int) {
var xUint64 uint64
for i := precision - 1; i >= 0; i-- {
xUint64 <<= 8
xUint64 += uint64(p[i])
}
if signed {
return signedToFloat(precision, xUint64), precision
}
return unsignedToFloat(precision, xUint64), precision
}
func floatToSigned(precision int, x float64) uint64 {
if x < 0 {
compl := uint64(-x * math.Exp2(float64(precision)*8-1))
return uint64(1<<uint(precision*8)) - compl
}
return uint64(math.Min(x*math.Exp2(float64(precision)*8-1), math.Exp2(float64(precision)*8-1)-1))
}
func floatToUnsigned(precision int, x float64) uint64 {
return uint64(math.Min((x+1)/2*math.Exp2(float64(precision)*8), math.Exp2(float64(precision)*8)-1))
}
func signedToFloat(precision int, xUint64 uint64) float64 {
if xUint64 >= 1<<uint(precision*8-1) {
compl := 1<<uint(precision*8) - xUint64
return -float64(int64(compl)) / math.Exp2(float64(precision)*8-1)
}
return float64(int64(xUint64)) / math.Exp2(float64(precision)*8-1)
}
func unsignedToFloat(precision int, xUint64 uint64) float64 {
return float64(xUint64)/(math.Exp2(float64(precision)*8))*2 - 1
}
// Buffer is a storage for audio data. You can think of it as a bytes.Buffer for audio samples.
type Buffer struct {
f Format
data []byte
tmp []byte
}
// NewBuffer creates a new empty Buffer which stores samples in the provided format.
func NewBuffer(f Format) *Buffer {
return &Buffer{f: f, tmp: make([]byte, f.Width())}
}
// Format returns the format of the Buffer.
func (b *Buffer) Format() Format {
return b.f
}
// Len returns the number of samples currently in the Buffer.
func (b *Buffer) Len() int {
return len(b.data) / b.f.Width()
}
// Pop removes n samples from the beginning of the Buffer.
//
// Existing Streamers are not affected.
func (b *Buffer) Pop(n int) {
b.data = b.data[n*b.f.Width():]
}
// Append adds all audio data from the given Streamer to the end of the Buffer.
//
// The Streamer will be drained when this method finishes.
func (b *Buffer) Append(s Streamer) {
var samples [512][2]float64
for {
n, ok := s.Stream(samples[:])
if !ok {
break
}
for _, sample := range samples[:n] {
b.f.EncodeSigned(b.tmp, sample)
b.data = append(b.data, b.tmp...)
}
}
}
// Streamer returns a StreamSeeker which streams samples in the given interval (including from,
// excluding to). If from<0 or to>b.Len() or to<from, this method panics.
//
// When using multiple goroutines, synchronization of Streamers with the Buffer is not required,
// as Buffer is persistent (but efficient and garbage collected).
func (b *Buffer) Streamer(from, to int) StreamSeeker {
return &bufferStreamer{
f: b.f,
data: b.data[from*b.f.Width() : to*b.f.Width()],
pos: 0,
}
}
type bufferStreamer struct {
f Format
data []byte
pos int
}
func (bs *bufferStreamer) Stream(samples [][2]float64) (n int, ok bool) {
if bs.pos >= len(bs.data) {
return 0, false
}
for i := range samples {
if bs.pos >= len(bs.data) {
break
}
sample, advance := bs.f.DecodeSigned(bs.data[bs.pos:])
samples[i] = sample
bs.pos += advance
n++
}
return n, true
}
func (bs *bufferStreamer) Err() error {
return nil
}
func (bs *bufferStreamer) Len() int {
return len(bs.data) / bs.f.Width()
}
func (bs *bufferStreamer) Position() int {
return bs.pos / bs.f.Width()
}
func (bs *bufferStreamer) Seek(p int) error {
if p < 0 || bs.Len() < p {
return fmt.Errorf("buffer: seek position %v out of range [%v, %v]", p, 0, bs.Len())
}
bs.pos = p * bs.f.Width()
return nil
}