DAVID A JAFFE Ten Criteria for Evaluating synthesis Techniques part1

gpud
Mind Map by gpud, updated more than 1 year ago
gpud
Created by gpud over 5 years ago
4
0

Description

David@Jaffe.com the first grouping out of two sets of five criteria each. The first relates to the parameters that define the synthesis type - whether additive synthesis, physical modelling, waveguide physical modelling, FM synthesis, AM synthesis with parameters such as filtering, envelope shaping, metaparameters, to build HYBRID METHODS - linear interpolation, from Karplus-Strong to multiple complex sampling with filtering, interpolation, delays, comb (all-pass) filters, scaling, noise filtes
Tags

Resource summary

DAVID A JAFFE Ten Criteria for Evaluating synthesis Techniques part1
1 10 EVALUATION CRITERIA: PARAMETRIC 1 - 5
1.1 how INTUITIVE are the parameters
1.1.1 STRONG -MAPPING control parameters to musical parameters
1.1.1.1 EGS STRONG
1.1.1.1.1 dynamics
1.1.1.1.2 articulation
1.1.1.1.3 ANALOG SYNTHESIS LPF maps BW / AMP less brightness with softer dynamics
1.1.1.1.3.1 LOW VALUES = PP
1.1.1.1.3.1.1 softer notes = lower AMP + less high freq
1.1.1.1.3.2 HI VALUES = FF
1.1.1.1.3.2.1 louder notes = higher AMP + more high freq
1.1.1.1.3.2.1.1 physical nonlinearities become more prominant as instr are played louder
1.1.1.1.3.3 ADJUSTABLE ON note-by-note basis: DECOUPLES
1.1.1.1.3.3.1 DISTANCE - FUNCTION OF LOUDNESS
1.1.1.1.3.3.2 DYNAMICS - FUNCTION OF BRIGHTNESS
1.1.2 WEAK UNMAPPED mathematical variables
1.1.2.1 EGS WEAK
1.1.2.1.1 Changing FM INDEX OF A CASCADE MODULATOR SLIGHTLY
1.1.2.1.1.1 DRASTIC, DIFFICULT TO PREDICT CHANGE IN TONAL QUALITY EG FROM DRUYM TO WOODBLOCK
1.1.2.1.1.1.1 Annoying for composers
1.1.2.1.1.1.2 Problematic for performer - in live situ
1.2 how PERCEPTIBLE are parameter changes
1.2.1 more parameters - weaker the impact of each
1.2.2 STRONG
1.2.2.1 sig. ch. = obvious audible effect
1.2.2.2 too strong = uncontrollable
1.2.2.3 e.g. a low pass filter proportional to cutoff freq (expon or linear)
1.2.3 WEAK
1.2.3.1 sig. ch. = barely audible
1.2.3.2 too weak = fruitless arbitrary alterations
1.2.3.3 METAPARAMETERS
1.2.3.3.1 EG1a
1.2.3.3.1.1 Used to enable more intuitive and perceptible changes to synthesis methods - e.g. brightness to each array element freq, AMP array, brightness can be applied across each element of the array
1.2.3.3.1.1.1 ActualAmps[i]=Amps[i]*MAX(brightness, 0.001)^log2(MAX(Freqs[i],100.0)/50.0)
1.2.3.3.1.2 addSyn used as inv FFT the LPF - log2 and MAX() rtns max of its args,
1.2.3.3.2 EG1b
1.2.3.3.2.1 addSyn * LineSeg - multidimentional timbral space
1.2.3.3.2.1.1 values added to scale addSyn partials producing complex interpolations (mcNabb 8f1)
1.2.3.3.3 EG2
1.2.3.3.3.1 Digital Sampling (*ID synth tech). edit 1 sample no effect used wtih HYBRID PARAMETERIZED PROCESSING MODULES eg filters, amp envs etc
1.2.3.4 EG1 addSyn - ch. amp of each partial could be weak for hi harmonics
1.3 how PHYSICAL are parameters
1.3.1 cntrls synthetic instr. in same way as real-world instr. is cntrled
1.3.1.1 EG1
1.3.1.1.1 Violin - waveguide model bow pressure/velocity (smith 92)
1.3.1.2 EG2
1.3.1.2.1 Clarinet mouth breath pressure parameter (cook 88)
1.3.2 ADV
1.3.2.1 1 good at: Complx behaviours at unstable moments in a tones evolution eg attack and transistions
1.3.2.2 2 no need for const. analytical modelling/database player/composer/implemntor neednt understand what happens in output wav during attack/trams te physical nature of parameter cuases the correct result
1.3.3 DIS
1.3.3.1 For nonexpert players - nonphysical INTUITIVE cntrls - more appropr.
1.3.4 for non-phys. synth parameters (eg. amp of addSyn partial -
1.3.4.1 SOLUTION - MAP WITH META- PARAMETER
1.3.4.2 PROB - need thorough ACOUSTICAL KNOWLEDGE to defne mapping adquately
1.4 how WELL BEHAVED are the parameters
1.4.1 EGS - LINEAR TECHNIQUES
1.4.1.1 when CNTRL RATE = SLOW RELATIVE TO AUDIO SR
1.4.1.1.1 STRONG
1.4.1.1.1.1 STRONG - PROPORTIONAL change to sound wth cntrler
1.4.1.1.1.2 FILTERS - GOOD DURING TRANSITIONS
1.4.1.1.1.2.1 PERRY COOK WAVEGUIDE-BASED VOCAL SYNTH MODEL 90
1.4.1.1.1.2.1.1 MODELS vocal tract - series waveguide filters
1.4.1.1.1.2.1.1.1 Each waveguide filter = xsect. of vocal tract
1.4.1.1.1.2.1.1.1.1 each vowel = physical interpretation - 1 or sev section diameters: shrink / incr.
1.4.1.1.1.2.1.1.2 interpolating between vowels corres. dir. to real-world spatial interpolation
1.4.1.1.1.3 Each intermediate val. have INTUITIVE PHYSICAL interpretation = WELLBEHAVED
1.4.1.2 CNTRL parameters change RAPIDLY - ENERGY IS INECTED INTO TH SSYSTEM WTH POSS OF UNPREDICTABLE UNEXPECTED RESULTS
1.4.1.2.1 WEAK
1.4.1.2.1.1 WEAK - small parameter ch - WILD, UNPREDICTABLE sonic outcome
1.4.1.2.1.2 EG 3 NONLINEAR FB TECH (Mcintyre/ woodhouse 83
1.4.1.2.1.2.1 nonlin. FB in phys. mod. wind instr.
1.4.1.2.1.2.2 EXTREME SENS. to INITIAL COND.
1.4.1.2.1.2.2.1 DEF. OF CHAOTIC SYS.
1.4.1.2.1.2.2.1.1 can be modelled in REGIONS of its STATE SPACE for PREDICABLE EFFECTIVE HAEVIOUR
1.4.1.2.2 E.G. AUDIO RATE AMPLITUDE SCALING
1.4.1.2.2.1 E.G. SINUSOIDAL
1.4.1.2.2.1.1 AM - 2 sidebands are produced for each componente in original signal = NONLINEAR - SIDEBANDS ARE AT FREQ CORRESONDING TO THE FREQ OF EACH COMPONENT PLUS OR MINUS THE MOD FREQ
1.4.1.2.2.2 E.G. NON- SINUSOIDAL
1.4.1.2.2.2.1 EVEN MORE CMPLX than above
1.4.1.2.3 EG2 FILTER STRUCTURES THAT BECOME UNSTABLE DURING TRANSITIONS FROM 1 SET OF COEFFICIENTS TO ANOTHER
1.5 how ROBUST is the sound's IDENTITY
1.5.1 HOW WELL SOUND RETAINS ID IN CONTEXT OF VARIATION
1.5.1.1 True rep. of expres.. instr. = FAMILY OF RELATED SOUNDS req.
1.5.1.2 not repr. of 1 note
1.5.2 Inst. Def. wth grt variety in context of particular sounrce ID
1.5.3 synthesis NOT a snapshot = expressive OVER TIME
1.5.4 EG Good
1.5.4.1 PHYS MOD
1.5.4.1.1 PhM Waveguide
1.5.4.1.1.1 Karplus- Strong String
1.5.4.1.1.1.1 rich expr. vocab. not loose ID
1.5.4.1.1.1.1.1 e.g. pick pos, string: flex, thickness, dyamics, decay char
1.5.4.1.1.1.2 ks 83 / jaffe J smith 83
1.5.4.2 SMPLING Power / flexib. CAN BE grtly enhan.
1.5.4.2.1 Multiple rec. xros
1.5.4.2.1.1 BW
1.5.4.2.1.2 dynamic contours
1.5.4.2.1.3 attacks
1.5.4.2.1.3.1 Vibrato
1.5.4.2.1.4 max multiple guises
1.5.4.2.2 filters = dynamics
1.5.4.2.3 linear interpolation for timbral variety eg thru BW n.b. take care to MATCH PITCH AND PHASE
1.5.4.2.4 as many samples per inst. as memory allows - more guises covered more expressive result
1.5.5 EG POOR
1.5.5.1 Pure Sampling
1.5.5.1.1 add var. to one sample = 2D cardboard sound
1.5.5.1.1.1 e.g. articulation, dynamics, timbre falls over if pused lightly in any direction
1.5.5.2 1 pitch
1.5.6 sampling = challenge + opportunity
2 identity synthesis technique = DIGITAL SAMPLING. Each SAMPLE is a PARAMETER itself
Show full summary Hide full summary

Similar

Evaluation of Explanations of Schizophrenia
Charlotte97
Evaluation of Conformity
littlestephie
Evaluation: Psychological Formulation
Katie Greensted
Evaluation: Social Causation Hypothesis as an Explanation for Schizophrenia
Katie Greensted
Evaluation: Drug Treatment for Anorexia
Katie Greensted
Evaluation: Genetic Explanation of Anorexia Nervosa
Katie Greensted
How does my media product represent particular social group?
rutendomsonzams
Sperry Evaluation
mitchcharlie
psychology: stress in everyday life
m.siddika98
Genetic explanation of depression
bro-bro-bro123
Psychology memory studies and evaluations
eharveyhudl