Sie sind auf Seite 1von 67

preamble

Historical Overview of Audio Spectral Modeling


Julius Smith
CCRMA, Stanford University
Music 421 Applications Lecture
January 7, 2014

Julius Smith

Music 421 Applications Lecture 1 / 53

Milestones in Audio Spectral Modeling


Outline
Telharmonium
Voder
Channel Vocoder
Phase Vocoder
Additive Synthesis
FM Synthesis
Sinusoidal Modeling
Future

Julius Smith

Fouriers theorem (1822)


Telharmonium (1898)
Voder (1920s)
Vocoder (1920s)
Hammond Organ (1930s)
Phase Vocoder (1966)
Digital Organ (1968)
Additive Synthesis (1969)
FM Brass Synthesis (1970)
Synclavier 8-bit FM/Additive synthesizer (1975)
FM singing voice (1978)
Sinusoidal Modeling (1985)
Sines + Noise (1988)
Sines + Noise + Transients (1988,1996,1998,2000)
Inverse FFT synthesis (1992)
Future Directions

Music 421 Applications Lecture 2 / 53

Outline
Telharmonium
Voder
Channel Vocoder
Phase Vocoder
Additive Synthesis
FM Synthesis
Sinusoidal Modeling
Future

Telharmonium (1898)

Julius Smith

Music 421 Applications Lecture 3 / 53

Telharmonium (Cahill 1898)


U.S. patent 580,035:
Art of and Apparatus for Generating and Distributing Music Electrically

Julius Smith

Music 421 Applications Lecture 4 / 53

Telharmonium Rheotomes
Forerunner of the Hammond Organ Tone Wheels

Julius Smith

Music 421 Applications Lecture 5 / 53

Telharmonium Rotor: Forerunner of Hammond Organ Tone


Wheels
Outline
Telharmonium

Telharmonium
Voder
Channel Vocoder
Phase Vocoder
Additive Synthesis
FM Synthesis
Sinusoidal Modeling
Future

Julius Smith

Music 421 Applications Lecture 6 / 53

Outline
Telharmonium
Voder
Channel Vocoder
Phase Vocoder
Additive Synthesis
FM Synthesis
Sinusoidal Modeling
Future

The Voder (1939)

Julius Smith

Music 421 Applications Lecture 7 / 53

The Voder (Homer Dudley 1939 Worlds Fair)

[<

http://davidszondy.com/future/robot/voder.htm
Julius Smith

Music 421 Applications Lecture 8 / 53

Voder Keyboard

http://www.acoustics.hut.fi/publications/files/theses/
lemmetty mst/chap2.html (from Klatt 1987)
Julius Smith

Music 421 Applications Lecture 9 / 53

Voder Schematic

http://ptolemy.eecs.berkeley.edu/~eal/audio/voder.html

Julius Smith

Music 421 Applications Lecture 10 / 53

Voder Demos
Outline

Video

Telharmonium
Voder

Audio

Voder Keyboard
Voder Schematic
Voder Demos
Channel Vocoder
Phase Vocoder
Additive Synthesis
FM Synthesis
Sinusoidal Modeling
Future

Julius Smith

Music 421 Applications Lecture 11 / 53

Outline
Telharmonium
Voder
Channel Vocoder
Phase Vocoder
Additive Synthesis
FM Synthesis
Sinusoidal Modeling
Future

The Channel Vocoder (1928)


(Voice Coder)

Julius Smith

Music 421 Applications Lecture 12 / 53

Vocoder Analysis & Resynthesis (Dudley 1928)


Analysis:
Outline
Telharmonium
Voder

Ten analog bandpass filters between 250 and 3000 Hz:


Bandpass rectifier lowpass filter amplitude envelope

Channel Vocoder

Vocoder Examples
Phase Vocoder
Additive Synthesis

Voiced/Unvoiced decision made


Fundamental frequency F0 measured for voiced case

FM Synthesis
Sinusoidal Modeling

Synthesis:

Future

Ten matching bandpass filters driven by a


buzz source (voiced), or
hiss source (unvoiced)
Bands were scaled by amplitude envelopes and summed
Said to have an unpleasant electrical accent
Julius Smith

Related Speech Models:

Music 421 Applications Lecture 13 / 53

Vocoder Filter Bank Analysis/Resynthesis


magnitude, or
magnitude and
phase extraction

Outline
Telharmonium
Voder
A

x0 (t)

Channel Vocoder

Vocoder Examples

Phase Vocoder
Additive Synthesis

x
0 (t)

x(t)

FM Synthesis

x1 (t)
f

Data Compression,
Transmission,
Storage,
Manipulation,
Noise reduction, ...

x
1 (t)

x
(t)

Sinusoidal Modeling
Future

xN 1

x
N 1

Analysis

Julius Smith

Processing

Synthesis

Music 421 Applications Lecture 14 / 53

Channel Vocoder Sound Examples


Outline
Telharmonium

Original

Voder
Channel Vocoder

Vocoder Examples
Phase Vocoder
Additive Synthesis

10 channels, sine carriers


10 channels, narrowband-noise carriers

FM Synthesis
Sinusoidal Modeling
Future

26 channels, sine carriers


26 channels, narrowband-noise carriers
26 channels, narrowband-noise carriers, channels reversed

Phase Vocoder: Identity system in absence of modifications


Julius Smith

Music 421 Applications Lecture 15 / 53

Outline
Telharmonium
Voder
Channel Vocoder
Phase Vocoder
Additive Synthesis
FM Synthesis
Sinusoidal Modeling
Future

The Phase Vocoder (1966)

Julius Smith

Music 421 Applications Lecture 16 / 53

Phase Vocoder Analysis for Additive Synthesis (1976)


Analysis Model

Outline
Telharmonium

ak(t)

Out

Additive Synthesis
FM Synthesis
Sinusoidal Modeling

k+k(t)

A
F
Sine Osc

Channel Vocoder
Phase Vocoder

ak

Channel Filter
Response

Voder

Synthesis Model

Early channel vocoder implementations (hardware) only


measured amplitude ak (t) (Dudley 1939)

Future

The phase vocoder (Flanagan and Golden 1966) added


phase tracking in each channel

Portnoff (1976) developed the FFT phase vocoder,


which replaced the heterodyne comb in computer-music
additive-synthesis analysis (James A. Moorer)

Inverse FFT synthesis (Rodet and Depalle 1992) gave faster


sinusoidal oscillator banks
Julius Smith

Music 421 Applications Lecture 17 / 53

Amplitude and Frequency Envelopes


Outline

ak (t)

Telharmonium
Voder
Channel Vocoder
Phase Vocoder
Additive Synthesis
FM Synthesis
Sinusoidal Modeling
Future

k (t) = k (t)

Julius Smith

Music 421 Applications Lecture 18 / 53

Outline
Telharmonium
Voder
Channel Vocoder
Phase Vocoder
Additive Synthesis
FM Synthesis
Sinusoidal Modeling
Future

Additive Synthesis (1969)

Julius Smith

Music 421 Applications Lecture 19 / 53

Classic Additive-Synthesis Analysis (Heterodyne Comb)


Outline
Telharmonium
Voder
Channel Vocoder
Phase Vocoder
Additive Synthesis

Additive Analysis
Additive Synthesis
FM Synthesis
Sinusoidal Modeling
Future

John Grey 1975 CCRMA Tech. Reports 1 & 2


(CCRMA STANM reports available online)
Julius Smith

Music 421 Applications Lecture 20 / 53

Classic Additive-Synthesis (Sinusoidal Oscillator Envelopes)


Outline
Telharmonium
Voder
Channel Vocoder
Phase Vocoder
Additive Synthesis

Additive Analysis
Additive Synthesis
FM Synthesis
Sinusoidal Modeling
Future

John Grey 1975 CCRMA Tech. Reports 1 & 2


(CCRMA STANM reports available online)
Julius Smith

Music 421 Applications Lecture 21 / 53

Classic Additive Synthesis Diagram (Computer Music, 1960s)


Outline

A1 (t) f1 (t) A2 (t) f2 (t) A3 (t) f3 (t) A4 (t) f4 (t)

Telharmonium
Voder
Channel Vocoder
Phase Vocoder
Additive Synthesis

Additive Analysis
Additive Synthesis
FM Synthesis
Sinusoidal Modeling
Future

noise

FIR

y(t) =

4
X
i=1

Julius Smith

Ai (t) sin

Z

i (t)dt + i (0)
0

Music 421 Applications Lecture 22 / 53

Classic Additive-Synthesis Examples


Outline
Telharmonium
Voder
Channel Vocoder
Phase Vocoder
Additive Synthesis

Additive Analysis
Additive Synthesis
FM Synthesis
Sinusoidal Modeling
Future

Bb Clarinet
Eb Clarinet
Oboe
Bassoon
Tenor Saxophone
Trumpet
English Horn
French Horn
Flute

All of the above


Independently synthesized set
(Synthesized from original John Grey data)

Julius Smith

Music 421 Applications Lecture 23 / 53

Outline
Telharmonium
Voder
Channel Vocoder
Phase Vocoder
Additive Synthesis
FM Synthesis
Sinusoidal Modeling
Future

Frequency Modulation Synthesis


(1973)

Julius Smith

Music 421 Applications Lecture 24 / 53

Frequency Modulation (FM) Synthesis


Outline

FM synthesis is normally used as a spectral modeling technique.

Telharmonium
Voder
Channel Vocoder

Discovered and developed (1970s) by John M. Chowning


(CCRMA Founding Director)

Phase Vocoder
Additive Synthesis

Key paper: JAES 1973 (vol. 21, no. 7)

FM Synthesis

FM Synthesis
FM Formula
FM Patch
FM Spectra
FM Examples
FM Voice
Sinusoidal Modeling

Commercialized by Yamaha Corporation:


DX-7 synthesizer (1983)
OPL chipset (SoundBlaster PC sound card)
Cell phone ring tones

Future

On the physical modeling front, synthesis of vibrating-string


waveforms using finite differences started around this time:
Hiller & Ruiz, JAES 1971 (vol. 19, no. 6)
Julius Smith

Music 421 Applications Lecture 25 / 53

Frequency Modulation (FM) Synthesis


Outline

FM synthesis is normally used as a spectral modeling technique.

Telharmonium
Voder
Channel Vocoder

Discovered and developed (1970s) by John M. Chowning


(CCRMA Founding Director)

Phase Vocoder
Additive Synthesis

Key paper: JAES 1973 (vol. 21, no. 7)

FM Synthesis

FM Synthesis
FM Formula
FM Patch
FM Spectra
FM Examples
FM Voice
Sinusoidal Modeling

Commercialized by Yamaha Corporation:


DX-7 synthesizer (1983)
OPL chipset (SoundBlaster PC sound card)
Cell phone ring tones

Future

On the physical modeling front, synthesis of vibrating-string


waveforms using finite differences started around this time:
Hiller & Ruiz, JAES 1971 (vol. 19, no. 6)
Julius Smith

Music 421 Applications Lecture 25 / 53

Frequency Modulation (FM) Synthesis


Outline

FM synthesis is normally used as a spectral modeling technique.

Telharmonium
Voder
Channel Vocoder

Discovered and developed (1970s) by John M. Chowning


(CCRMA Founding Director)

Phase Vocoder
Additive Synthesis

Key paper: JAES 1973 (vol. 21, no. 7)

FM Synthesis

FM Synthesis
FM Formula
FM Patch
FM Spectra
FM Examples
FM Voice
Sinusoidal Modeling

Commercialized by Yamaha Corporation:


DX-7 synthesizer (1983)
OPL chipset (SoundBlaster PC sound card)
Cell phone ring tones

Future

On the physical modeling front, synthesis of vibrating-string


waveforms using finite differences started around this time:
Hiller & Ruiz, JAES 1971 (vol. 19, no. 6)
Julius Smith

Music 421 Applications Lecture 25 / 53

Frequency Modulation (FM) Synthesis


Outline

FM synthesis is normally used as a spectral modeling technique.

Telharmonium
Voder
Channel Vocoder

Discovered and developed (1970s) by John M. Chowning


(CCRMA Founding Director)

Phase Vocoder
Additive Synthesis

Key paper: JAES 1973 (vol. 21, no. 7)

FM Synthesis

FM Synthesis
FM Formula
FM Patch
FM Spectra
FM Examples
FM Voice
Sinusoidal Modeling

Commercialized by Yamaha Corporation:


DX-7 synthesizer (1983)
OPL chipset (SoundBlaster PC sound card)
Cell phone ring tones

Future

On the physical modeling front, synthesis of vibrating-string


waveforms using finite differences started around this time:
Hiller & Ruiz, JAES 1971 (vol. 19, no. 6)
Julius Smith

Music 421 Applications Lecture 25 / 53

Frequency Modulation (FM) Synthesis


Outline

FM synthesis is normally used as a spectral modeling technique.

Telharmonium
Voder
Channel Vocoder

Discovered and developed (1970s) by John M. Chowning


(CCRMA Founding Director)

Phase Vocoder
Additive Synthesis

Key paper: JAES 1973 (vol. 21, no. 7)

FM Synthesis

FM Synthesis
FM Formula
FM Patch
FM Spectra
FM Examples
FM Voice
Sinusoidal Modeling

Commercialized by Yamaha Corporation:


DX-7 synthesizer (1983)
OPL chipset (SoundBlaster PC sound card)
Cell phone ring tones

Future

On the physical modeling front, synthesis of vibrating-string


waveforms using finite differences started around this time:
Hiller & Ruiz, JAES 1971 (vol. 19, no. 6)
Julius Smith

Music 421 Applications Lecture 25 / 53

Frequency Modulation (FM) Synthesis


Outline

FM synthesis is normally used as a spectral modeling technique.

Telharmonium
Voder
Channel Vocoder

Discovered and developed (1970s) by John M. Chowning


(CCRMA Founding Director)

Phase Vocoder
Additive Synthesis

Key paper: JAES 1973 (vol. 21, no. 7)

FM Synthesis

FM Synthesis
FM Formula
FM Patch
FM Spectra
FM Examples
FM Voice
Sinusoidal Modeling

Commercialized by Yamaha Corporation:


DX-7 synthesizer (1983)
OPL chipset (SoundBlaster PC sound card)
Cell phone ring tones

Future

On the physical modeling front, synthesis of vibrating-string


waveforms using finite differences started around this time:
Hiller & Ruiz, JAES 1971 (vol. 19, no. 6)
Julius Smith

Music 421 Applications Lecture 25 / 53

Frequency Modulation (FM) Synthesis


Outline

FM synthesis is normally used as a spectral modeling technique.

Telharmonium
Voder
Channel Vocoder

Discovered and developed (1970s) by John M. Chowning


(CCRMA Founding Director)

Phase Vocoder
Additive Synthesis

Key paper: JAES 1973 (vol. 21, no. 7)

FM Synthesis

FM Synthesis
FM Formula
FM Patch
FM Spectra
FM Examples
FM Voice
Sinusoidal Modeling

Commercialized by Yamaha Corporation:


DX-7 synthesizer (1983)
OPL chipset (SoundBlaster PC sound card)
Cell phone ring tones

Future

On the physical modeling front, synthesis of vibrating-string


waveforms using finite differences started around this time:
Hiller & Ruiz, JAES 1971 (vol. 19, no. 6)
Julius Smith

Music 421 Applications Lecture 25 / 53

Frequency Modulation (FM) Synthesis


Outline

FM synthesis is normally used as a spectral modeling technique.

Telharmonium
Voder
Channel Vocoder

Discovered and developed (1970s) by John M. Chowning


(CCRMA Founding Director)

Phase Vocoder
Additive Synthesis

Key paper: JAES 1973 (vol. 21, no. 7)

FM Synthesis

FM Synthesis
FM Formula
FM Patch
FM Spectra
FM Examples
FM Voice
Sinusoidal Modeling

Commercialized by Yamaha Corporation:


DX-7 synthesizer (1983)
OPL chipset (SoundBlaster PC sound card)
Cell phone ring tones

Future

On the physical modeling front, synthesis of vibrating-string


waveforms using finite differences started around this time:
Hiller & Ruiz, JAES 1971 (vol. 19, no. 6)
Julius Smith

Music 421 Applications Lecture 25 / 53

FM Formula
Outline

x(t) = Ac sin[c t + c + Am sin(m t + m )]

Telharmonium
Voder

where

Channel Vocoder
Phase Vocoder

(Ac , c , c ) specify the carrier sinusoid

Additive Synthesis
FM Synthesis

FM Synthesis
FM Formula
FM Patch
FM Spectra
FM Examples
FM Voice

(Am , m , m ) specify the modulator sinusoid


Can also be called phase modulation

Sinusoidal Modeling
Future

Julius Smith

Music 421 Applications Lecture 26 / 53

Simple FM Brass Patch (Chowning 1970)


Outline
Telharmonium

Jean-Claude Risset observation (19641969):


Brass bandwidth amplitude

Voder

fm = f0

Channel Vocoder
Phase Vocoder

g
A

Additive Synthesis

FM Synthesis

FM Synthesis
FM Formula
FM Patch
FM Spectra
FM Examples
FM Voice
Sinusoidal Modeling

fc = f0
A

Future

Out

Julius Smith

Music 421 Applications Lecture 27 / 53

FM Harmonic Amplitudes (Bessel Function of First Kind)


Outline

Harmonic number k , FM index :

Telharmonium
Voder
Channel Vocoder
Phase Vocoder
Additive Synthesis
FM Synthesis

0.5

Jk()

FM Synthesis
FM Formula
FM Patch
FM Spectra
FM Examples
FM Voice

Sinusoidal Modeling

Future

2
-0.5
0

4
5

10
15

20
25
30

10

Order k

Argument

Julius Smith

Music 421 Applications Lecture 28 / 53

Frequency Modulation (FM) Examples


All examples by John Chowning unless otherwise noted:
Outline
Telharmonium

FM brass synthesis

Voder
Channel Vocoder
Phase Vocoder
Additive Synthesis
FM Synthesis

FM Synthesis
FM Formula
FM Patch
FM Spectra
FM Examples
FM Voice
Sinusoidal Modeling
Future

Low Brass example


Dexter Morrils FM Trumpet
FM singing voice (1978)
Each formant synthesized using an FM operator pair
(two sinusoidal oscillators)

Chorus
Voices
Basso Profundo
Other early FM synthesis
Clicks and Drums
Big Bell
String Canon

Julius Smith

Music 421 Applications Lecture 29 / 53

FM Voice
Outline
Telharmonium

FM voice synthesis can be viewed as compressed modeling of


spectral formants

Voder
Channel Vocoder

Carrier 1

Additive Synthesis
FM Synthesis

FM Synthesis
FM Formula
FM Patch
FM Spectra
FM Examples
FM Voice
Sinusoidal Modeling
Future

Julius Smith

Carrier 2
Carrier 3

Magnitude

Phase Vocoder

0
Frequency
Modulation Frequency (all three)

Music 421 Applications Lecture 30 / 53

Outline
Telharmonium
Voder
Channel Vocoder
Phase Vocoder
Additive Synthesis
FM Synthesis
Sinusoidal Modeling
Future

Sinusoidal Modeling Synthesis


(1988)

Julius Smith

Music 421 Applications Lecture 31 / 53

Tracking Spectral Peaks in the Short-Time Fourier Transform


Outline

atan

Telharmonium
Voder
Channel Vocoder

s(t)

Phases

FFT

Frequencies

Phase Vocoder
Additive Synthesis
FM Synthesis

dB mag

Peak
tracking

Quadratic
Peak
Interpolation

window w(n)
Amplitudes

Sinusoidal Modeling

Sinusoidal Modeling
Spectral Trajectories
Sines + Noise
S+N Examples
S+N FX
S+N XSynth
Sines + Transients
S + N + Transients
S+N+T TSM
S+N+T Freq Map
S+N+T Windows
HF Noise Modeling
HF Noise Band
S+N+T Examples

STFT peak tracking at CCRMA: mid-1980s (PARSHL program)


Motivated by vocoder analysis of piano tones
Influences: STFT (Allen and Rabiner 1977),
ADEC (1977), MAPLE (1979)
Independently developed for speech coding by McAulay and
Quatieri at Lincoln Labs (1985)

Future

Julius Smith

Music 421 Applications Lecture 32 / 53

Example Spectral Trajectories


Outline

Telharmonium
Voder
Channel Vocoder
Phase Vocoder
Additive Synthesis
FM Synthesis
Sinusoidal Modeling

Sinusoidal Modeling
Spectral Trajectories
Sines + Noise
S+N Examples
S+N FX
S+N XSynth
Sines + Transients
S + N + Transients
S+N+T TSM
S+N+T Freq Map
S+N+T Windows
HF Noise Modeling
HF Noise Band
S+N+T Examples

Future

Julius Smith

Music 421 Applications Lecture 33 / 53

Parametric Spectral Modeling


Outline

A1 (t) f1 (t) A2 (t) f2 (t) A3 (t) f3 (t) A4 (t) f4 (t)

Telharmonium
Voder
Channel Vocoder
Phase Vocoder
Additive Synthesis
FM Synthesis
Sinusoidal Modeling

Sinusoidal Modeling
Spectral Trajectories
Sines + Noise
S+N Examples
S+N FX
S+N XSynth
Sines + Transients
S + N + Transients
S+N+T TSM
S+N+T Freq Map
S+N+T Windows
HF Noise Modeling
HF Noise Band
S+N+T Examples

white noise

u(t)

y(t) =

filter(t)

ht ( )
4
P

i=1

Ai (t) cos

hR
t
0

i (t)dt + i (0) + (ht u)(t)

Future

Julius Smith

Music 421 Applications Lecture 34 / 53

Sines + Noise Sound Examples


Outline

Xavier Serra thesis demos (Sines + Noise signal modeling)

Telharmonium
Voder

Piano

Channel Vocoder
Phase Vocoder
Additive Synthesis

Original
Sinusoids alone

FM Synthesis
Sinusoidal Modeling

Sinusoidal Modeling
Spectral Trajectories
Sines + Noise
S+N Examples
S+N FX
S+N XSynth
Sines + Transients
S + N + Transients
S+N+T TSM
S+N+T Freq Map
S+N+T Windows
HF Noise Modeling
HF Noise Band
S+N+T Examples

Residual after sinusoids removed


Sines + noise model
Voice
Original
Sinusoids
Residual
Synthesis

Future

Julius Smith

Music 421 Applications Lecture 35 / 53

Musical Effects with Sines+Noise Models


Outline
Telharmonium
Voder

Piano Effects
Pitch downshift one octave

Channel Vocoder
Phase Vocoder
Additive Synthesis

Pitch flattened
Varying partial stretching

FM Synthesis
Sinusoidal Modeling

Sinusoidal Modeling
Spectral Trajectories
Sines + Noise
S+N Examples
S+N FX
S+N XSynth
Sines + Transients
S + N + Transients
S+N+T TSM
S+N+T Freq Map
S+N+T Windows
HF Noise Modeling
HF Noise Band
S+N+T Examples

Voice Effects
Frequency-scale by 0.6
Frequency-scale by 0.4 and stretch partials
Variable time-scaling, deterministic to stochastic

Future

Julius Smith

Music 421 Applications Lecture 36 / 53

Cross-Synthesis with Sines+Noise Models


Outline
Telharmonium

Voice modulator

Voder
Channel Vocoder
Phase Vocoder

Creaking ships mast carrier

Additive Synthesis
FM Synthesis

Voice-modulated creaking mast

Sinusoidal Modeling

Sinusoidal Modeling
Spectral Trajectories
Sines + Noise
S+N Examples
S+N FX
S+N XSynth
Sines + Transients
S + N + Transients
S+N+T TSM
S+N+T Freq Map
S+N+T Windows
HF Noise Modeling
HF Noise Band
S+N+T Examples

Same with modified spectral envelopes

Future

Julius Smith

Music 421 Applications Lecture 37 / 53

Sines + Transients Sound Examples


Outline
Telharmonium

In this technique, the sinusoidal sum is phase-matched at the


cross-over point only (with no cross-fade).

Voder
Channel Vocoder

Marimba

Phase Vocoder
Additive Synthesis
FM Synthesis
Sinusoidal Modeling

Sinusoidal Modeling
Spectral Trajectories
Sines + Noise
S+N Examples
S+N FX
S+N XSynth
Sines + Transients
S + N + Transients
S+N+T TSM
S+N+T Freq Map
S+N+T Windows
HF Noise Modeling
HF Noise Band
S+N+T Examples

Original
Sinusoidal model
Original attack, followed by sinusoidal model
Piano
Original
Sinusoidal model
Original attack, followed by sinusoidal model

Future

Julius Smith

Music 421 Applications Lecture 38 / 53

Multiresolution Sines + Noise + Transients


Outline

Why Model Transients Separately?

Telharmonium
Voder
Channel Vocoder
Phase Vocoder
Additive Synthesis
FM Synthesis
Sinusoidal Modeling

Sinusoidal Modeling
Spectral Trajectories
Sines + Noise
S+N Examples
S+N FX
S+N XSynth
Sines + Transients
S + N + Transients
S+N+T TSM
S+N+T Freq Map
S+N+T Windows
HF Noise Modeling
HF Noise Band
S+N+T Examples

Sinusoids efficiently model spectral peaks over time


Filtered noise efficiently models spectral residual vs. t
Neither is good for abrupt transients in the waveform
Phase-matched oscillators are expensive
More efficient to switch to a transient model during transients
Need sinusoidal phase matching at the switching times

Transient models:

Original waveform slice (1988)


Wavelet expansion (Ali 1996)
MPEG-2 AAC (with short window) (Levine 1998)
Frequency-domain LPC
(time-domain amplitude envelope) (Verma 2000)

Future

Julius Smith

Music 421 Applications Lecture 39 / 53

Time Scale Modification of Sines + Noise + Transients Models

original signal

Voder

sines +
noise

sines +
noise

transients

Telharmonium

transients

Outline

sines +
noise

Channel Vocoder
Phase Vocoder

time-scaled
signal

sines +
noise

sines +
noise

transients

FM Synthesis

transients

Additive Synthesis

sines +
noise

Sinusoidal Modeling

Sinusoidal Modeling
Spectral Trajectories
Sines + Noise
S+N Examples
S+N FX
S+N XSynth
Sines + Transients
S + N + Transients
S+N+T TSM
S+N+T Freq Map
S+N+T Windows
HF Noise Modeling
HF Noise Band
S+N+T Examples

time

Time-Scale Modification (TSM) becomes well defined:

Transients are translated in time


Sinusoidal envelopes are scaled in time
Noise-filter envelopes also scaled in time
Dual of TSM is frequency scaling

Future

Julius Smith

Music 421 Applications Lecture 40 / 53

Sines + Noise + Transients Time-Frequency Map


16

Outline
14

Telharmonium

Channel Vocoder
Phase Vocoder
Additive Synthesis

12

frequency [kHz]

Voder

10
8
6

FM Synthesis

Sinusoidal Modeling

2
0

50

100

150

200

250

50

100

150

200

250

amplitude

Sinusoidal Modeling
Spectral Trajectories
Sines + Noise
S+N Examples
S+N FX
S+N XSynth
Sines + Transients
S + N + Transients
S+N+T TSM
S+N+T Freq Map
S+N+T Windows
HF Noise Modeling
HF Noise Band
S+N+T Examples

0.5
0
-0.5
-1

time [milliseconds]

(Levine 1998)

Future

Julius Smith

Music 421 Applications Lecture 41 / 53

Outline
Telharmonium

transient

Corresponding Analysis Windows

Additive Synthesis
FM Synthesis
Sinusoidal Modeling

Sinusoidal Modeling
Spectral Trajectories
Sines + Noise
S+N Examples
S+N FX
S+N XSynth
Sines + Transients
S + N + Transients
S+N+T TSM
S+N+T Freq Map
S+N+T Windows
HF Noise Modeling
HF Noise Band
S+N+T Examples

low octave

Phase Vocoder

amplitude

Channel Vocoder

middle octave high octave

Voder

50

100

150

200

250

time [milliseconds]

Future

Julius Smith

Music 421 Applications Lecture 42 / 53

Quasi-Constant-Q (Wavelet) Time-Frequency Map


16

Outline

14

Telharmonium

Channel Vocoder
Phase Vocoder
Additive Synthesis

12

frequency [kHz]

Voder

FM Synthesis

8
6
4

Sinusoidal Modeling

2
0

0x 104

50

100

150

200

250

50

100

150

200

250

amplitude

Sinusoidal Modeling
Spectral Trajectories
Sines + Noise
S+N Examples
S+N FX
S+N XSynth
Sines + Transients
S + N + Transients
S+N+T TSM
S+N+T Freq Map
S+N+T Windows
HF Noise Modeling
HF Noise Band
S+N+T Examples

10

1
0
-1
-2

time [milliseconds]

Future

Julius Smith

Music 421 Applications Lecture 43 / 53

Bark-Band Noise Modeling at High Frequencies (Levine 1998)


Outline

85

Telharmonium

80
Voder

75

Channel Vocoder
Phase Vocoder

70

Additive Synthesis

Sinusoidal Modeling

Sinusoidal Modeling
Spectral Trajectories
Sines + Noise
S+N Examples
S+N FX
S+N XSynth
Sines + Transients
S + N + Transients
S+N+T TSM
S+N+T Freq Map
S+N+T Windows
HF Noise Modeling
HF Noise Band
S+N+T Examples

65

magnitude [dB]

FM Synthesis

60
55
50
45
40
35
30

500

1000

1500

2000

2500

3000

3500

4000

4500

frequency [Hz]

Future

Julius Smith

Music 421 Applications Lecture 44 / 53

Amplitude Envelope for One Noise Band


80

Telharmonium
Voder
Channel Vocoder
Phase Vocoder

Original Mag. [dB]

Outline

70

60

50

Additive Synthesis
40

FM Synthesis

50

100

50

100

150

200

250

300

350

150

200

250

300

350

Sinusoidal Modeling
80
75

LSA Mag. [dB]

Sinusoidal Modeling
Spectral Trajectories
Sines + Noise
S+N Examples
S+N FX
S+N XSynth
Sines + Transients
S + N + Transients
S+N+T TSM
S+N+T Freq Map
S+N+T Windows
HF Noise Modeling
HF Noise Band
S+N+T Examples

70
65
60
55
50
45

time [milliseconds]

Future

Julius Smith

Music 421 Applications Lecture 45 / 53

Sines + Noise + Transients Sound Examples


Outline
Telharmonium

Scott Levine Thesis Demos (Sines + Noise + Transients at 32 kbps)


(http://ccrma.stanford.edu/~scottl/thesis.html)

Voder
Channel Vocoder
Phase Vocoder
Additive Synthesis

It Takes Two by Rob Base & DJ E-Z Rock

FM Synthesis
Sinusoidal Modeling

Sinusoidal Modeling
Spectral Trajectories
Sines + Noise
S+N Examples
S+N FX
S+N XSynth
Sines + Transients
S + N + Transients
S+N+T TSM
S+N+T Freq Map
S+N+T Windows
HF Noise Modeling
HF Noise Band
S+N+T Examples

Original
MPEG-AAC at 32 kbps
Sines+transients+noise at 32 kbps

Multiresolution sinusoids
Residual Bark-band noise
Transform-coded transients (AAC)
Bark-band noise above 5 kHz

Future

Julius Smith

Music 421 Applications Lecture 46 / 53

Time Scale Modification using Sines + Noise + Transients


Outline
Telharmonium

Scott Levine Thesis Demos (Sines + Noise + Transients at 32 kbps)


(http://ccrma.stanford.edu/~scottl/thesis.html)

Voder
Channel Vocoder
Phase Vocoder
Additive Synthesis

Time-Scale Modification (pitch unchanged)

FM Synthesis
Sinusoidal Modeling

Sinusoidal Modeling
Spectral Trajectories
Sines + Noise
S+N Examples
S+N FX
S+N XSynth
Sines + Transients
S + N + Transients
S+N+T TSM
S+N+T Freq Map
S+N+T Windows
HF Noise Modeling
HF Noise Band
S+N+T Examples

S+N+T time-scale factors [2.0, 1.6, 1.2, 1.0, 0.8, 0.6, 0.5]

S+N+T Pitch Shifting (timing unchanged)

Pitch-scale factors [0.89, 0.94, 1.00, 1.06, 1.12]

Future

Julius Smith

Music 421 Applications Lecture 47 / 53

Outline
Telharmonium
Voder
Channel Vocoder
Phase Vocoder
Additive Synthesis
FM Synthesis
Sinusoidal Modeling
Future

Summary and Future Prospects

Julius Smith

Music 421 Applications Lecture 48 / 53

Spectral Modeling History Highlights

Bernoullis modal sums (1733)


Fouriers initial theorem (1822)
Telharmonium (1906)
Hammond organ (1930s)
Channel Vocoder (1939)
Phase Vocoder (1966)
Additive Synthesis (1969)
FFT Phase Vocoder (1976)
Sinusoidal Modeling
(1977,1979,1985)
Sines+Noise (1989)
Sines+Transients (1989)
TF Reassignment (1995)
Sines+Noise+Transients
(1998)

Julius Smith

Perceptual audio coding:


Princen-Bradley filterbank
(1986)
K. Brandenburg thesis (1989)
Auditory masking usage
Dolby AC2
Musicam
ASPEC
MPEG-I,II,IV
(S+N+T parametric sounds)

Music 421 Applications Lecture 49 / 53

Telharmonium (Cahill 1898)


U.S. patent 580,035:
Art of and Apparatus for Generating and Distributing Music Electrically

Julius Smith

Music 421 Applications Lecture 50 / 53

Telharmonium Rheotomes
Forerunner of the Hammond Organ Tone Wheels

Julius Smith

Music 421 Applications Lecture 51 / 53

Telharmonium Rotor: Forerunner of Hammond Organ Tone


Wheels
Outline
Telharmonium
Voder
Channel Vocoder
Phase Vocoder
Additive Synthesis
FM Synthesis
Sinusoidal Modeling
Future

SM Summary
Telharmonium
Spec Future

Julius Smith

Music 421 Applications Lecture 52 / 53

Future Prospects
Observations:

Sinusoidal modeling of sound is Unreasonably Effective


Basic auditory masking discards 90% information
Interesting neuroscience observation:
... most neurons in the primary auditory cortex A1 are silent
most of the time ...
(from Sparse Time-Frequency Representations, Gardner and
Magnesco, PNAS:103(16), April 2006)

What is the right psychospectral model for sound?


The cochlea of the ear is a real-time spectrum analyzer
How is the ears spectrogram represented at higher levels?
Julius Smith

Music 421 Applications Lecture 53 / 53

Future Prospects
Observations:

Sinusoidal modeling of sound is Unreasonably Effective


Basic auditory masking discards 90% information
Interesting neuroscience observation:
... most neurons in the primary auditory cortex A1 are silent
most of the time ...
(from Sparse Time-Frequency Representations, Gardner and
Magnesco, PNAS:103(16), April 2006)

What is the right psychospectral model for sound?


The cochlea of the ear is a real-time spectrum analyzer
How is the ears spectrogram represented at higher levels?
Julius Smith

Music 421 Applications Lecture 53 / 53

Future Prospects
Observations:

Sinusoidal modeling of sound is Unreasonably Effective


Basic auditory masking discards 90% information
Interesting neuroscience observation:
... most neurons in the primary auditory cortex A1 are silent
most of the time ...
(from Sparse Time-Frequency Representations, Gardner and
Magnesco, PNAS:103(16), April 2006)

What is the right psychospectral model for sound?


The cochlea of the ear is a real-time spectrum analyzer
How is the ears spectrogram represented at higher levels?
Julius Smith

Music 421 Applications Lecture 53 / 53

Future Prospects
Observations:

Sinusoidal modeling of sound is Unreasonably Effective


Basic auditory masking discards 90% information
Interesting neuroscience observation:
... most neurons in the primary auditory cortex A1 are silent
most of the time ...
(from Sparse Time-Frequency Representations, Gardner and
Magnesco, PNAS:103(16), April 2006)

What is the right psychospectral model for sound?


The cochlea of the ear is a real-time spectrum analyzer
How is the ears spectrogram represented at higher levels?
Julius Smith

Music 421 Applications Lecture 53 / 53

Future Prospects
Observations:

Sinusoidal modeling of sound is Unreasonably Effective


Basic auditory masking discards 90% information
Interesting neuroscience observation:
... most neurons in the primary auditory cortex A1 are silent
most of the time ...
(from Sparse Time-Frequency Representations, Gardner and
Magnesco, PNAS:103(16), April 2006)

What is the right psychospectral model for sound?


The cochlea of the ear is a real-time spectrum analyzer
How is the ears spectrogram represented at higher levels?
Julius Smith

Music 421 Applications Lecture 53 / 53

Future Prospects
Observations:

Sinusoidal modeling of sound is Unreasonably Effective


Basic auditory masking discards 90% information
Interesting neuroscience observation:
... most neurons in the primary auditory cortex A1 are silent
most of the time ...
(from Sparse Time-Frequency Representations, Gardner and
Magnesco, PNAS:103(16), April 2006)

What is the right psychospectral model for sound?


The cochlea of the ear is a real-time spectrum analyzer
How is the ears spectrogram represented at higher levels?
Julius Smith

Music 421 Applications Lecture 53 / 53

Future Prospects
Observations:

Sinusoidal modeling of sound is Unreasonably Effective


Basic auditory masking discards 90% information
Interesting neuroscience observation:
... most neurons in the primary auditory cortex A1 are silent
most of the time ...
(from Sparse Time-Frequency Representations, Gardner and
Magnesco, PNAS:103(16), April 2006)

What is the right psychospectral model for sound?


The cochlea of the ear is a real-time spectrum analyzer
How is the ears spectrogram represented at higher levels?
Julius Smith

Music 421 Applications Lecture 53 / 53

Das könnte Ihnen auch gefallen