Evolutionary Music Rochester Institute Of Technology-Books Download

Evolutionary Music Rochester Institute of Technology
18 Feb 2020 | 25 views | 0 downloads | 22 Pages | 517.06 KB

Share Pdf : Evolutionary Music Rochester Institute Of Technology

Download and Preview : Evolutionary Music Rochester Institute Of Technology

Report CopyRight/DMCA Form For : Evolutionary Music Rochester Institute Of Technology


Evolutionary Music Tutorial GECCO 2005,Musical Tasks EC in Music. Composition Create score abstraction Dates back to 1991. Performance Realize score in sound Horner and Goldberg Thematic bridging. Gibson and Byrne NEUROGEN,Synthesis Generate sounds electronically. Listening Derive abstraction from sounds Activity increasing rapidly. Reviewed over 120 articles for this tutorial,Improvisation Everything simultaneously. EC music class projects appearing on the www,Generative Systems Survey of EC Applied to Music. Certainly evolutionary certainly relevant Organized around musical tasks. Cellular Automata music apps since 1980 s Task analysis of the musical domain. Swarms emergent behavior colonies Choose subtasks where EC used. Artificial Life Some representative examples, Sonification of data DNA Genetic music See my Web site for references and links.
Fractals chaotic systems music since 1970 s www it rit edu jab. Not my primary focus due to time Goals,Recruit some new blood. Motivate discussion of fundamental EC issues,2005 John A Biles 2. Evolutionary Music Tutorial GECCO 2005,EC in Composition Composition Subtasks. First application area 1991 Generate melodies motives. Largest application area Generate melodic line sequence of pitches. Generate rhythm sequence of durations, Describe subtasks of composition Develop extend enhance melodies. Cite some examples Generate variations, Summarize themes and variations Combine motives to create longer lines.
Generate countermelodies,Composition Subtasks A Few Examples. Harmonization Horner and Goldberg 1991, Generate harmony parts hymns chorales Thematic bridging melody morphing. Generate harmonic foundation chord changes Bred sequence of operations to transform one. Arranging motive into another,Fitness hit target if so check bridge length. Rhythm section accompaniment,Counterpoint NEUROGEN Gibson and Byrne 1991. Structure Rhythm GA with NN fitness function,Add pitch GA 2 NN interval structure.
Generate or adhere to form,Harmony Simple rule base. Generate sections higher level units,2005 John A Biles 3. Evolutionary Music Tutorial GECCO 2005, variations Bruce Jacob 1995 GP Music Johanson Poli 97. Three components all GAs GP melody generator short monophonic. Composer builds phrases from user supplied Terminals pitches or rest. Functions musical development,Ear Judges the composer s output fitness. Arranger Orders phrases into composition,No real rhythm all notes same length.
fitness by user Fitness, Starts at motive level above notes Interactive 1 100 rating pair wise comparison. Co evolution of Composers and Ears Neural nets trained on ratings from interactive. Sample Hegemon Fibre 1st movement runs 1 100 version worked less badly. Even toy domains are tricky,GenDash Rodney Waschka II Harmonization SATB. New music composer not a techie Soprano Alto Tenor Bass classic four part. GenDash GA tool he tweaks for each Voicing individual chords and voice leading. piece since mid 1990 s Standard rule sets exist automatic fitness. Sappho s Breath 2001 1 act opera arias Basically a scheduling problem optimize. Initial population 26 measures of music,Represent chord sequence or voice sequences. Random selection crossover at note level,Fitness usually number of constraints violated. All children of each generation heard,Around five generations per aria.
Mixed success, Highly collaborative artistic Easy if chords specified more constrained. Harder if chords evolved too more creative,2005 John A Biles 4. Evolutionary Music Tutorial GECCO 2005,Harmonization Examples Rhythm Drum Machine. Horner and Ayers 1995 Generate single measure or longer patterns. Melody and chord symbols 4 part harmony 2D grid standard drum machine interface. Broke problem into 2 parts,Time on X axis,Enumerate all possible voicings for each chord. GA to find best sequence of voicings voice leading Instrument on Y axis. Phon Amnuaisuk et al 1999 MIDI velocity in the cells 0 127. Evolved chords themselves as well Build textures,More creative less tractable Loop one measure.
Rule based system worked better Build longer phrases from multiple patterns. EC probably not the best approach,Rhythm Drum Machine Rhythm Examples. Horowitz 1994,Representation params to generating function. One measure drum textures presented visually,Mentor listens selects favorites to survive breed. CONGA Tokui and Iba 2000,4 to 16 measure patterns user specifies. GA evolves half or one measure patterns grid,GP arranges patterns into phrases hierarchy.
Levels evolved separately mentor switches, Screen Shot from Band in a Box PG Music Neural net to thin the GA population. 2005 John A Biles 5,Evolutionary Music Tutorial GECCO 2005. SBEAT Tatsuo Unemi 2002 Pitch Duration Representations. Currently in third version Pitch, Representation individuals are measures Absolute pitch scale degree MIDI note Hz. 16 events fixed time grid X Relative interval, 3 chromosomes pitch rhythm velocity X From previous pitch. Up to 23 parts 13 solo 2 chord 8 rhythm From beginning of phrase or composition. Collaborative system User can From tonic of key or root of chord. Select individuals to breed Durations, Manipulate underlying chord scale Beat oriented multiples divisions of beat.
Enter and protect parts Absolute milliseconds,Arrange measures into score piece. Melody Chromosomes Melody Fitness,Position based, Time windows on fixed temporal grid beats fractions. Explicit rules and heuristics, Enforces beat measure phrase structure From music theory or hip pocket. Tilts toward beat oriented music Usually combined via weighted average. Order based Interactive human mentor critic rater,Pitch duration pairs durations can be arbitrary. Measure lines ignored superimposed or irrelevant Display individuals rater selects and rates. Facilitates non pulse music Perform in musical context real time. Tree based GP Learn from examples neural networks,Terminals usually notes pitch maybe duration.
Input either features or melodic fragments,Functions usually musical operators. Facilitates more complex forms extend hierarchy Examples come from desired style. 2005 John A Biles 6,Evolutionary Music Tutorial GECCO 2005. Operators Initialization Operators Selection, Random Start from scratch Traditional fitness based. Uniform white noise generator Encourages convergence. Fractals Can be problem if diversity critical,Markov chains Musically aware. Sampled Look for individuals to fill a role, User supplied motive s to develop Random no fitness.
Licks from analyzed corpus Works if individuals all musically meritorious. Maximum diversity,Crossover and Mutation EC in Performance. Is the purpose to alter or develop Expressive performance of score not trivial. Alter more random less guided Classical alter note onsets length envelopes. Develop more musically aware Jazz also alter notes add delete change. Crossover point s Annotate jazz performance Grachten. At bit vs musical boundaries note measure GA to minimize cost of edit distance. Random vs musically meaningful operations to transform score to performance. Use training sets of correct performances,Flip bits likely to be unmusical. Musically meaningful may be too safe,2005 John A Biles 7. Evolutionary Music Tutorial GECCO 2005,Audience Mediated Performance Performance kind of. GenJam Populi more later GA to enhance public speaking voice Sato. Sound Gallery Woolf and Thompson Three genes pitch volume speed. Artistic installation piece Fitness from mentors, Speakers in corners of room four islands Not real time yet.
Each driven by evolving hardware distorting a HPDJ Hewlett Packard Disc Jockey. source sound Select tunes sequence them do crossfades. Fitness location of patrons closer is better Fitness crowd animation level. Migration to keep people moving,EC in Synthesis Matching a Target Sound. Control synthesis algorithms techniques Basically an optimization problem. Goal Higher level more musical interface Fitness perceptual spectral matching. Huge chaotic parameter spaces GA to evolve parameter settings Horner. Provide guided search through synthesis space Unit generator UG parameters FM modular. Two different subtasks Additive synthesis envelope breakpoints. Match a target sound Wavetable physical modeling parameters. Generate new hopefully interesting sounds CSound Recipes Horner and Ayres 2002. GP to evolve UG topologies Garcia 2001,Reverb params match room Mrozek 96. 2005 John A Biles 8,Evolutionary Music Tutorial GECCO 2005. Search for New Sounds Granular Synthesis, Explore a synthesis technique s sound space Sound objects made up of 1 100 ms grains. Fitness mentor preference Each grain has waveform pitch envelope. Sound object cloud has density shape,Goal often collaborative tool for sound.
designers and composers Microsound Roads 2001 MIT Press. Example Timbre trees Takala 1993 GA to evolve parameters Johnson 99. Evolve topology of unit generator patches GP FOF formant wave function synthesis. Sounds synchronized to animated motion Evolves parameters for CSound function call. Emergent Granular Synthesis Synthesizer Control, Chaosynth Miranda 1995 Commercial Synthesizers hard to control. CA to control grain parameters Muta Synth Palle Dahlstedt 2001. As CA self organizes sound emerges Customizable S W controller for Nord synth. Swarm Granulator Blackwell 2003 Extended to real time interactive performance. Swarmer Swarm is the granular cloud Genophone Mandelis 2002. Interpreter Interprets swarm for granulator Evolves sounds and gesture mappings. Granulator Sound engine Max MSP Data glove interface. Real time interactive performance Sends SysEx messages to Korg Prophecy. 2005 John A Biles 9,Evolutionary Music Tutorial GECCO 2005. Breed Actual Waveforms EC in Listening, Thesis Cristyn Magnus SDSU 2003 NEXTPITCH Francine Federman 2000. Representation LCS to predict next pitch in melody. Waveform sample array Nursery tunes and chorales simple melodies. Genes segments bounded by zero crossings Accidental evolution of a radio Layzell 02. Operators Trying to evolve a hardware oscillator, Crossover and mutations at gene level only Got a radio that received oscillations from a. Eliminates clicks and pops,nearby computer,Fitness Match waveform or amp envelope.
Piece is evolution of initial to target sounds,EC Listeners in Composers EC in Improvisation. The EAR in Bruce Jacob s variations system Compose and perform concurrently Jazz. IGA to breed set of data filters for harmonies Spontaneous real time interactive. Each filter passes an acceptable chord,Has to be right the first time. Co evolved critics Todd and Werner 99 Jazz is an inherently evolutionary domain. Male singers 32 note song Jam session environment highly competitive. Female critics prefer certain intervals Survival of fittest cutting sessions. Female selects male with best intervals Players borrow others ideas licks. Best means most surprising Can even trace lineage of licks and soloists. 2005 John A Biles 10,Evolutionary Music Tutorial GECCO 2005. Spector and Alpern 1994 5 Papadopoulos and Wiggins 98. Toward general case based artist generator Generate blues chorus not real time. Traded bebop fours using GP not real time Chromosome 12 bar blues of 1 16th notes. Terminal set four bar phrase from human,Initialization Random. Function set 13 melody transforms,Evolved programs to transform human four.
Crossover single and two point note level,Fitness Mutation musically meaningful. Five features from jazz theory literature Fitness 8 features in fixed weighted sum. Neural net trained on Bird licks Goal Eliminate subjectivity EC neat. Hybrid combination worked best,Best sounding result was human edited. Swarm Music GenJam An In Depth Example, Tim Blackwell 2003 GenJam Genetic Jammer 1994 present. Swarm based collective improvisation Models a jazz improviser agent of sorts. Basically Swarm Granulator operating at Real time interactive performance MIDI. note level instead of grain level Lets a trumpet player work as a single. Self organization Versions for 4 4 3 4 5 4 7 4 12 8 16 8. Stigmergy interact by modifying environs About 250 tunes in repertoire. Swing bebop cool Latin funk new age,Follow me from CD Swarm Music. 2005 John A Biles 11,Evolutionary Music Tutorial GECCO 2005.
Representation of a Phrase,Interactive GenJam Architecture. GenJam Normal Form,23 12 57 57 11 38 11 6 9 7 0 5 7 8 7 5. 38 4 7 8 7 7 15 15 15 0,Phrase Population 57 22 9 7 0 5 7 15 15 0. Measure Population,Chord Scale Mappings,Chord Scale Notes. GenJam s Genetic Algorithm,Cmaj7 Major avoid 4th CDEGAB.
C7 Mixolydian avoid 4th C D E G A Bb,Cm7 Minor avoid 6th C D Eb F G Bb. Cm7b5 Locrian avoid 2nd C Eb F Gb Ab Bb,Fairly standard GA process for both populations. Cdim W H Diminished C D Eb F Gb G A B Random initialization. C Lydian Augmented C D E F G A B, C7 Whole Tone C D E F G Bb Tournament selection 4 individuals in a family. C7 11 Lydian Dominant C D E F G A Bb 2 fittest family members become parents. C7alt Altered Scale C Db D E Gb G Bb, C7 9 Mix 2 avoid 4th C D E G A Bb Single point crossover creates 2 kids. C7b9 Harm Minor V no 6th C Db E F G Bb Musically meaningful mutation until kids are unique. CmMaj7 Melodic Minor C D Eb F G A B, Cm6 Dorian avoid 7th C D Eb F G A 2 kids replace 2 least fit family members.
Cm7b9 Melodic Minor II C Db Eb F G A Bb,Cmaj7 11 Lydian C D E F G A B. Replace 50 of each population in breed mode, C7sus Mixolydian C D E F G A Bb Replace worst 4 measures 3 phrases in tweak. Cmaj7sus Major CDEFGAB,C7Bl Blues C Eb F Gb G Bb,2005 John A Biles 12. Evolutionary Music Tutorial GECCO 2005, Example Measure Crossover Musically Meaningful Mutations. on Measures,Random bit level crossover point,Standard melodic development techniques.
9705 7 8 75 Parent1 1001 0111 0000 0101 0111 100 0 0111 0101. 7 8 7 7 15 15 15 0 Parent2 0111 1000 0111 0111 1111 111 1 1111 0000. 9 7 0 5 7 9 15 0 Child1 1001 0111 0000 0101 0111 100 1 1111 0000. 7 8 7 7 15 14 7 5 Child2 0111 1000 0111 0111 1111 111 0 0111 0101. Musically Meaningful Mutations,Intelligent Genetic Operators. on Phrases, Operate at measure pointer level not bit level GA s usually have dumb operators smart fitness. Rely on fitness to guide search, Mutation Operator Mutated Phrase Explanation Leads to fitness bottleneck in IGAs especially temporal. None 57 57 11 38 Original Phrase, Rotate Right Random 57 11 38 57 3 positions in this case. GenJam currently uses smart operators, Reverse 38 11 57 57 Play measures in reverse order Intelligent mutation Already seen.
True Retrograde 38 11 57 57 Play measures backward too Intelligent initialization Fractals Markov chains. Sequence Phrase 57 57 38 38 Repeat a measure, Intelligent crossover Preserve horizontal intervals. Genetic Repair 57 57 11 23 Replace worst measure, Super Phrase 55 13 21 34 Winners of fitness tournaments Good parents tend to have good children. Lick Thinner 31 57 11 38 Replace most common measure. Reduces volume through the fitness bottleneck, Orphan Phrase 43 37 53 19 Losers of frequency tournaments. 2005 John A Biles 13,Evolutionary Music Tutorial GECCO 2005. GenJam Generations Demo Real Time Interaction,Old GenJam version improvise 4 choruses.
When GenJam trades fours with human,Tune is Tadd Dameron s Lady Bird. Listen to human s four Roland GI 10, 16 bar form straight up rhythm Map human phrase to GJNF chromosomes. Each chorus uses a more mature generation Mutate the phrase and 4 measures. 1st Generation 0 white noise generator Play mutated result as its response. 2nd Gen 1 one breeding 50 new Use mutation as melodic development. 3rd Gen 3 two more breeding Results in true conversation. 4th Gen 5 one breed one tweak one cheat Highly robust and formidable opponent. Final chorus Gen 5 using current system,Fault Tolerant Pitch Tracking Anatomy of a Four. I played quote from Prince Albert,Pitch tracker makes lots of mistakes. Wrong pitch,Extra note on events,GenJam heard this from pitch tracker.
Extra note off events,Not a problem,Map to GJNF which is highly robust. Errors not mistakes they re development,GenJam mutated and played this back. Will mutate anyway before playing,2005 John A Biles 14. Evolutionary Music Tutorial GECCO 2005,Collective Improvisation Making GenJam Autonomous. GenJam and human solo simultaneously GenJam more fun when interactive. GenJam listens to human while it s soloing Fitness not necessary or even possible. Maps to GJNF Good human four good GenJam four, Plays what human did earlier delay line Initialization is very smart.
Delay of 1 bar or n events 4 bars smart echo GenJam s full chorus solos not as good. No mutation Replay as close as possible Ideas competent but seldom compelling. Human can trade 1 s play harmony counterpoint Initialization not smart enough. Challenge for the human Move to an autonomous GenJam. Autonomous GJ Architecture Initialize from Stored Licks. Licks Databases several styles,4 bar licks come from 1001 Jazz Licks. Map to GJNF by hand,Database Initialization algorithm. Select 16 4 bar licks from database,Seed measure pop with those 64 measures. First 16 phrases are the 16 original licks,Remaining 32 phrases are smart crossovers.

Related Books

BAB II TINJAUAN PUSTAKA A. Kecemasan Menghadapi Ujian ...

BAB II TINJAUAN PUSTAKA A Kecemasan Menghadapi Ujian

Ujian Nasional adalah kekhawatiran, perasaan tegang yang tidak menyenangkan, dan mengeluhkan sesuatu hal buruk terkait Ujian Nasional di masa yang akan datang ditandai dengan gejala-gejala fisik, perilaku, serta pemikiran yang muncul ketika subjek dihadapkan pada situasi yang membuatnya cemas, dalam hal ini ialah situasi yang berkaitan dengan Ujian Nasional. 2. Ciri-ciri Kecemasan Menghadapi ...

Pine Hill Public Schools Curriculum

Pine Hill Public Schools Curriculum

7.1.IL.C.2 Dramatize student-created and/or authentic short plays, skits, poems, songs, stories, or reports. 7.1.IL.C.3 Use language creatively to respond in writing to a variety of oral or visual prompts about familiar and some unfamiliar situations. 7.1.IL.C.4 Synthesize information found in age- and level-



Intermediate Russian, like its sister volume, Basic Russian, is ideal for both independent study and use in class. The two books comprise a compendium of the essentials of Russian grammar. John Murray and Sarah Smyth are Lecturers in Russian at Trinity College, Dublin.

ANUL UNIVERSITAR 2017-2018 - dppd.ase.ro

ANUL UNIVERSITAR 2017 2018 dppd ase ro

1 metodologie privind organizarea ?i desf??urarea concursului de admitere la programele de formare psihopedagogic?, anul universitar 2017-2018

Literature homework help

Literature homework help

time. 5 Cavewoman The cavewoman costume is one of my favourites as it combines feminine allure with a girl power undertone. Hepl will also aid them in optimizing internal literature homework help and functioning to ensure that performance and literature homework help within the organization is literature homework help at its peak. It is carved ...

December 1995

December 1995

too-tight cavewoman costume, chicken bone necklace flapping away, A couple of motorists honked and waved at the odd lot of pedestrians-in appreciation, I think, Then it was back into the cemetary, where we belonged in spirit: Batman, Peter Pan, two very fast ballerinas, a cluster of purple grapes, two



clinical operations at a Harvard Medical School-affiliated hospital. Given the rapid change across all sectors of health care, the importance of formal training in operational aspects of health care has never been more critical. Health care organizations require high-performing, collaborative teams that can work together seamlessly to deliver value-based care, ensure patient satisfaction and ...



of Affiliated Engineers Inc. He has extensive experience planning, engineering, and designing biological and chemical research laboratories and vivaria nationwide and is a technical contributor to the NIH Design Requirements Manual for Biomedical Laboratories and Animal Research Facilities.

La Entrada High School CDS Code

La Entrada High School CDS Code

Mathemat-ics textbooks are currently being piloted in grades K-12 and will be recommended for adoption spring of 2015 for 2015-16 implementation. English Language Arts/ English Language Development textbooks will be available for pilot during the 2016-17 school year and recommended for adoption in spring 2017 and implementa-tion in 2017-18. 2014-15 School Year Textbooks and Instructional ...