- Activation function, 370
- AD conversion, 97
- delta‐sigma, 104
- oversampling, 100
- AD converter, 113
- characteristics, 113
- counter, 118
- delta‐sigma, 120
- half‐flash, 116–117
- parallel, 116
- subranging, 117
- successive approximation, 117
- Adagrad, 373
- Adam, 373
- Adaptive filters, 18
- autocorrelation method, 18–20
- LMS method, 20
- Additive synthesis, 346
- AES/EBU interface, 134
- Aliasing distortion, 350, 351, 353
- Aliasing suppression, 351, 354
- Allpass decomposition, 164, 166, 167
- Anti‐imaging filter, 68
- Antialiasing filter, 69
- Antiderivative antialiasing, 352
- Aperture delay, 113
- Aperture jitter, 114
- Artificial Neural Network, 369
- Attack time, 273
- Attention module, 389, 391
- Audio coding, 291
- lossless, 291–293
- lossy, 293–294
- Audio denoising, 388
- Audio interfaces
- computer, 140–141
- internet‐based, 145–146
- layer‐1, 142–143
- layer‐2, 143–144
- layer‐3, 144–145
- network‐based, 141
- Averaging time, 273
- Backpropagation, 369, 371
- causal BP through time (CBPTT), 375
- instantaneous BP through time (IBPTT), 375, 377
- Band limiting, 77, 97, 120
- Bandwidth, 149, 155, 162, 167, 377
- Bilinear transform, 356
- Bit allocation, 309
- Boost/Cut, 156, 158, 164
- Center frequency, 155, 160, 162, 167, 377
- Chain rule of derivatives, 369, 372, 374
- CIPIC database, 383
- Clipping, 343
- Coding
- Intensity stereo, 320–321
- Mono/Side, 320
- Coding techniques, 291, 303
- Comb filter, 109
- Compressor, 266, 276
- Constrained Energy Lapped Transform, 327
- Conversion time, 113
- Convolution
- continuous‐time, 3
- discrete‐time, 8, 9
- Convolutional layer, 373
- Convolutional Neural Network, 367, 373
- Critical bands, 295
- Cutoff frequency, 8, 17, 149, 154, 156, 158, 378–380, 382
- DA conversion, 97
- delta‐sigma, 104
- oversampling, 100
- DA converter, 120
- characteristics, 121
- delta‐sigma, 127
- R‐‐2R networks, 126
- switched sources, 123–124
- weighted capacitors, 125–126
- weighted resistors, 125
- Datasets, 391
- Decimation, 72, 109, 204, 210, 350
- Deep Learning, 367
- Deep Neural Network, 367
- Deglitcher, 121
- Delta modulation, 103
- Delta‐sigma modulation, 100
- decimation filter, 109–112
- first‐order, 102–105
- higher‐order, 108–109
- multistage, 106–108
- second‐order, 105–106
- signal‐to‐noise ratio, 108–109
- DFT, 191
- Difference equation, 12, 13
- Differential nonlinearity, 114
- Dilation factor, 390
- Discrete Fourier Transform, 11
- Distortion, 343
- Dither, 40, 50
- Downsampling, 69
- DSP, 132, 133
- Dynamic range, 265
- Dynamic range control, 265, 266
- Early reflections, 225, 235
- Echo density, 245
- Eigenfrequency, 226, 242, 244
- Energy decay curve, 387
- Entropy coding, 293
- Equalizers, 149
- design of non‐recursive, 201
- design of recursive, 153
- non‐recursive, 190
- recursive, 149, 162
- Expander, 266, 276
- Fast convolution, 191, 194
- Feedback delay network (FDN), 375
- Feedback systems, 249
- FFT, 298
- Filter
- allpass, 18, 165–167, 170, 243
- bandpass, 15, 149, 154, 170
- bandstop, 149, 154
- bilinear transform, 160
- constant‐Q, 155
- decimation, 204, 210, 211
- highpass, 15, 149, 153, 169, 216
- interpolation, 204, 210, 211
- kernel, 204, 205, 211
- lowpass, 15, 149, 153, 168, 216
- octave, 150
- one‐third octave, 152
- peak, 152, 158, 166, 169, 170, 216, 375
- Q‐factor, 160
- shelving, 152, 155–158, 164, 167–170, 216, 375
- Filter bank, 303
- analysis, 257
- multi‐complementary, 202
- octave‐band, 203
- synthesis, 257
- Filter structures
- coefficient quantization, 172
- limit cycles, 190
- noise behavior of recursive, 177–184
- noise shaping, 184–188
- non‐recursive, 194, 202
- parametric, 162–172
- recursive, 172
- scaling, 188–189
- FIR filter, 12
- Fourier transform
- continuous‐time, 4
- discrete‐time, 9, 10
- inverse continuous‐time, 4
- Frequency density, 244
- Frequency response, 4, 9
- Frequency sampling, 201
- Gain, 162, 167, 377
- Gain error, 114
- Global average pooling, 390
- Gradient, 369, 372, 374
- Gradient descent, 369, 373
- Hard clipping, 345, 350
- Harmonic distortion, 342
- Harmonics, 342
- HDMI, 139
- Head related impulse response, 382
- Head related transfer functions, 375
- Hidden layers, 369
- Huffmann coding, 293
- IDFT, 191
- IIR filter, 12
- Image model, 227
- Impulse response, 3, 4, 9
- Inception module, 389, 391
- Integral nonlinearity, 115
- Interpolation, 72, 74, 78, 100, 109, 204, 210, 350
- Lagrange, 85–87
- polynomial, 83–85
- Spline, 87–94
- Inverse Discrete Fourier Transform, 11
- ISO‐MPEG1, 303–309
- Laplace transform, 6
- Latency time, 112, 213, 293
- Limit cycles, 190
- Limiter, 266, 273
- Linear prediction, 20
- Log‐spectral distance, 383
- Loss function, 371
- Machine Learning, 367
- MADI interface, 135
- Magnitude response, 4, 10
- Masking, 297, 299, 302
- Masking index, 308
- Masking threshold, 297, 299, 300, 302, 307
- Monotonicity, 115
- MPEG‐2, 310
- MPEG‐2 AAC, 310–321
- MPEG‐4, 321–325
- Multilayer Perceptrons, 370
- Nesterov momentum, 373
- Neurons, 369
- Nodal DK method, 360
- Noise gate, 265, 266, 276
- Noise shaping, 46, 51, 178, 184
- Nonlinear filters, 347
- Nonlinear processing, 341
- Nonlinear stateful filters, 345
- Number representation, 51
- fixed‐point, 52–57
- floating‐point, 57–60
- format conversion, 60–62
- Nyquist sampling, 97
- Offset error, 114
- Overdrive, 343
- Oversampling, 98, 350
- signal‐to‐noise ratio, 108
- Padding, 390
- Peak factor, 27
- PEAK measurement, 153, 269
- Perceptual evaluation
- Phase response, 4, 10
- Polyphase representation, 71
- Prediction, 293
- Pseudo‐random sequence, 230, 257
- Psychoacoustic models, 305
- Pyramid Vector Quantization, 330
- Q‐factor, 154, 155
- Quantization, 6
- Quantization error
- correlation with signal, 39
- first‐order statistics, 34
- noise shaping, 46
- power, 98
- probability density function, 26
- second‐order statistics, 37
- Quantization model, 25
- Quantization noise, 302
- Quantization step, 26, 27, 74, 75, 98
- Quantization theorem, 25, 28
- Range coding, 331
- Ray tracing model, 227
- Reconstruction, 6
- Recurrent Neural Network, 367
- Release time, 273
- ReLU, 370
- Resolution, 113
- Reverberation time, 225, 235, 246, 387
- RMS measurement, 271
- Rmsprop, 373
- Room impulse response, 225, 375
- approximation, 256
- measurement of, 230
- Room simulation, 225, 383
- Root‐mean‐square measurement, 153
- Sample‐and‐hold, 113
- Sampling, 6
- Sampling interval, 6, 97
- Sampling rate, 6, 97
- Sampling rate conversion, 67
- asynchronous, 74–75
- multistage, 78–80
- single‐stage, 76–78
- synchronous, 70–73
- Sampling theorem, 6, 97
- Scale factor, 321
- Scaling factor, 307
- Settling time, 121
- Short‐time Fourier Transform, 389
- Sigmoid, 370
- Signal
- continuous‐time, 2
- discrete‐time, 8
- Signal processor
- fixed‐point, 132–133
- floating‐point, 133
- Signal quantization, 25
- Signal‐to‐mask ratio, 302, 305, 308, 309
- Signal‐to‐noise ratio, 27, 57, 59, 60
- Sinc distortion, 98
- Soft‐clipping, 344
- Softmax, 391
- Source‐filter processing, 20
- SPDIF interface, 135
- Spectral Band Replication SBR, 325
- Spreading function, 300
- Stability
- State‐space modeling, 354
- State‐space representation, 347, 359
- Static curve, 266
- Static nonlinear system, 341
- Steepest descent, 369
- Subsequent reverberation, 225, 241
- Supervised learning, 368
- System
- continuous‐time, 3
- discrete‐time, 9, 10
- transfer function, 13
- Tanh, 370
- THD, 342
- Threshold of hearing, 51, 296, 298, 307
- Time constants, 272
- Tonality index, 298, 302
- Total harmonic distortion, 115, 342
- Trapezoidal rule, 356
- Unsupervised learning, 368
- Upsampling, 68
- Virtual Analog Modeling, 354
- Wave Digital Filters, 355–359
- adaptors, 357
- connection tree, 358
- Wavefolding, 345
- Weighting filter
- Wirtinger calculus, 383
- Word length, 27, 172
..................Content has been hidden....................
You can't read the all page of ebook, please click
here login for view all page.