static void ANNFFNAccomp()
{
Databank db = new Databank("lib");
var cat = db.Load("Classical");
Composition inputComp = cat.Compositions[6];
MelodySequence inputSeq = inputComp.Tracks[0].GetMainSequence() as MelodySequence;
AccompanimentGeneratorANNFF gen = new AccompanimentGeneratorANNFF(cat, PatchNames.Acoustic_Grand);
//gen.Initialize();
gen.SetSequence(inputSeq);
gen.Train();
var outMel = gen.Generate();
MusicPlayer player = new MusicPlayer();
Composition comp = new Composition();
Track t = new Track(PatchNames.Orchestral_Strings, 6);
t.AddSequence(outMel);
comp.Add(t);
comp.Add(inputComp.Tracks[0]);
player.Play(comp);
comp.WriteToMidi("ann_ff_accomp.mid");
}
static void BrownTest()
{
var gen = new ReflectingBrownNoteGenerator(new NoteRangeRestrictor(48, 72, 2, 16, Scales.ScaleTypes[1]), new Random(), -1, 1, -1, 1);
MelodySequence seq = gen.Generate();
Console.WriteLine(seq.ToString());
MusicPlayer player = new MusicPlayer();
player.Play(seq);
Console.ReadLine();
}
static void ChordTest()
{
MusicPlayer player = new MusicPlayer();
Chord dminor = new Chord(new NoteNames[] { NoteNames.D, NoteNames.F, NoteNames.A }, 4, Durations.wn);
Chord dmajor = new Chord(new NoteNames[] { NoteNames.G, NoteNames.B, NoteNames.B }, 4, Durations.wn);
Chord cmajor = new Chord(new NoteNames[] { NoteNames.C, NoteNames.E, NoteNames.G }, 4, Durations.bn);
HarmonySequence seq = new HarmonySequence();
seq.AddChord(dminor); seq.AddChord(dmajor); seq.AddChord(cmajor); seq.AddChord(dminor); seq.AddChord(cmajor); seq.AddChord(dminor); seq.AddChord(dmajor); seq.AddChord(cmajor);
player.Play(seq);
player.Close();
}
static void Test9()
{
MusicPlayer player = new MusicPlayer();
Console.WriteLine("Harmony Test");
//ChromaticToneDuration, ChromaticTone, MelodicBigram, RhythmicBigram
MetricSimilarity cosine = new MetricSimilarity(GetMelodySequence(@"test\other\ff7tifa.mid"), new ChromaticToneDuration());
GeneticGenerator evolver = new GeneticGenerator(cosine);
var notes = evolver.Generate();
SimpleChord ch = new SimpleChord();
var melody = notes;
HarmonySequence harmony = ch.GetHarmonySequence(melody);
Track t1 = new Track(PatchNames.Acoustic_Grand, 2);
t1.AddSequence(melody);
Track t2 = new Track(PatchNames.Church_Organ, 1);
t2.AddSequence(harmony);
Composition comp = new Composition();
// comp.Add(t1);
comp.Add(t2);
player.Play(comp);
Console.ReadLine();
return;
}
static void Test8()
{
MusicPlayer player = new MusicPlayer();
Console.WriteLine("Hidden Markov Model");
var m1 = GetMelodySequence(@"test\other\frere.mid");
var m2 = GetMelodySequence(@"test\other\babaa.mid");
var m3 = GetMelodySequence(@"test\other\twinkle.mid");
var m4 = GetMelodySequence(@"test\hagrid.mid");
var m5 = GetMelodySequence(@"test\harry.mid");
var m6 = GetMelodySequence(@"test\harry2.mid");
MarkovGenerator markov = new MarkovGenerator(new MelodySequence[] { m6 });
while (true)
{
Composition comp = new Composition();
GeneticGenerator gen = new GeneticGenerator(new MetricSimilarity(m6, new IMetric[] { new Rhythm(), new RhythmicBigram(), new RhythmicInterval() }));
var notes2 = gen.Generate();
Track t2 = new Track((PatchNames)0, 10);
MelodySequence s = new MelodySequence();
s.AddPause((int)Durations.wn * 10);
foreach (Note n in notes2.Notes)
{
if (n.Pitch <= 0)
n.Pitch = 0;
else if (n.Pitch > 48)
n.Pitch = 40;
else if (n.Pitch > 70)
n.Pitch = 35;
else
{
n.Pitch = 49;
n.Duration *= 4;
n.Velocity = 50;
s.AddPause(n.Duration * 2);
}
s.AddNote(n);
}
t2.AddSequence(s);
var notes3 = markov.Generate().Notes as Note[];
MelodySequence baseseq = new MelodySequence();
int max_dur = 0;
int max_index = 0;
for (int i = (int)(notes3.Length * 0.7f); i < notes3.Length; i++)
{
if (notes3[i].Duration > max_dur)
{
max_dur = notes3[i].Duration;
max_index = i;
}
}
Note last_note = null;
for (int i = 0; i < max_index; i++)
{
last_note = notes3[i];
baseseq.AddNote(last_note);
}
baseseq.AddNote(new Note(last_note.Pitch - 12, last_note.Duration));
baseseq.AddNote(new Note(last_note.Pitch - 24, last_note.Duration * 2));
baseseq.AddPause(last_note.Duration * 32);
Track t = new Track(PatchNames.Vibraphone, 1);
var b1 = baseseq.Clone() as MelodySequence;
var b2 = baseseq.Clone() as MelodySequence;
var b3 = baseseq.Clone() as MelodySequence;
b1.Transpose(12);
b2.Transpose(6);
b3.Transpose(-12);
t.AddSequence(baseseq);
t.AddSequence(b1);
t.AddSequence(b2);
t.AddSequence(b3);
comp.Add(t);
comp.Add(t2);
Console.WriteLine("Press enter key to listen");
Console.ReadLine();
player.Play(comp);
}
}
static void Play(MelodySequence notes, PatchNames instrument = PatchNames.Vibraphone)
{
MusicPlayer player = new MusicPlayer();
player.SetPatch((int)instrument, 1);
player.Play(notes);
player.Close();
}
static void NeuralNetworkAccompanimentTest()
{
// initialize input and output values
double[][] input = new double[4][] {
new double[] {0, 0}, new double[] {0, 1},
new double[] {1, 0}, new double[] {1, 1}
};
double[][] output = new double[4][] {
new double[] {0}, new double[] {1},
new double[] {1}, new double[] {0}
};
SigmoidFunction sig = new SigmoidFunction();
Accord.Neuro.Networks.RestrictedBoltzmannMachine boltz = new Accord.Neuro.Networks.RestrictedBoltzmannMachine(200, 200);
// create neural network
ActivationNetwork network = new ActivationNetwork(
new SigmoidFunction(2),
200,
20,
200);
//BackPropagationLearning teacher = new BackPropagationLearning(network);
//LevenbergMarquardtLearning teacher = new LevenbergMarquardtLearning(network);
Accord.Neuro.Learning.ParallelResilientBackpropagationLearning teacher = new ParallelResilientBackpropagationLearning(network);
Accord.Neuro.Networks.DeepBeliefNetwork dpn = new Accord.Neuro.Networks.DeepBeliefNetwork(200, 20);
// teacher.IncreaseFactor = 1.01;
Composition c = Composition.LoadFromMIDI("test/other/ff7tifa.mid");
MusicPlayer player = new MusicPlayer();
//player.Play(c.Tracks[0]);
List<double[]> inputs = new List<double[]>(); List<double[]> outputs = new List<double[]>();
inputs.Add(GetDoublesFromNotes((c.Tracks[0].GetMainSequence() as MelodySequence).ToArray()));
outputs.Add(GetDoublesFromNotes((c.Tracks[1].GetMainSequence() as MelodySequence).ToArray()));
// inputs.Add(GetDoublesFromNotes((c.Tracks[1].GetMainSequence() as MelodySequence).ToArray()));
// outputs.Add(GetDoublesFromNotes((c.Tracks[2].GetMainSequence() as MelodySequence).ToArray()));
// inputs.Add(GetDoublesFromNotes((c.Tracks[0].GetMainSequence() as MelodySequence).ToArray()));
// outputs.Add(GetDoublesFromNotes((c.Tracks[3].GetMainSequence() as MelodySequence).ToArray()));
int its = 0;
while (its++ < 10000)
{
double error = teacher.RunEpoch(inputs.ToArray(), outputs.ToArray());
Console.WriteLine("{0}: Error - {1}", its, error);
}
var input_melody = (c.Tracks[0].GetMainSequence() as MelodySequence);
var new_notes = network.Compute(GetDoublesFromNotes(input_melody.ToArray()));
var new_mel = GetMelodyFromDoubles(new_notes);
player.Play(new_mel);
Console.ReadLine();
}
static void MetricTimingTest()
{
Databank db = new Databank("lib");
var cat = db.Load("Classical");
var fmets = new IMetric[]{new ChromaticTone(), new ChromaticToneDistance(), new ChromaticToneDuration(), new MelodicBigram(), new MelodicInterval()
, new Pitch(), new Rhythm(), new RhythmicBigram(), new RhythmicInterval()};
MusicPlayer player = new MusicPlayer();
foreach (var m in fmets)
{
Stopwatch watch = new Stopwatch();
watch.Start();
var metric = MetricSimilarity.GenerateMetricSimilarityMulti(cat.Compositions, new IMetric[] { m });
GeneticGenerator gen = new GeneticGenerator(metric);
gen.PrintProgress = false;
gen.MaxGenerations = 1000;
var mel = gen.Generate();
mel.Trim(20);
watch.Stop();
var className = ((object)m).GetType().Name;
Console.WriteLine("{0} - MF: {1}, AF {2}, T {3}", className, gen.MaxFitness, gen.AvgFitness, watch.ElapsedMilliseconds);
Console.ReadLine();
player.Play(mel);
}
Console.ReadLine();
}
static void MarkovAccompTest()
{
Databank db = new Databank("lib");
var cat = db.Load("Classical");
Composition inputComp = cat.Compositions[6];
MelodySequence inputSeq = inputComp.Tracks[0].GetMainSequence() as MelodySequence;
AccompanyGeneratorMarkov gen = new AccompanyGeneratorMarkov(cat, PatchNames.Orchestral_Strings);
var outMel = gen.Generate(inputSeq, 10);
MusicPlayer player = new MusicPlayer();
Composition comp = new Composition();
Track t = new Track(gen.Instrument, 6);
t.AddSequence(outMel);
comp.Add(t);
comp.Add(inputComp.Tracks[0]);
player.Play(comp);
comp.WriteToMidi("markov_model_accomp.mid");
}
static void Main(string[] args)
{
/*Feed forward neural network
Group notes into 3
Classify all groups in existing songs as 1, all other combinations as 0.
Use as fitness function
Use HMM decode as fitness function GA*/
/* Databank db = new Databank("lib");
List<MelodySequence> drumSeqs = new List<MelodySequence>();
var comps = new Composition[][] { db.Load("Classic Rock").Compositions, db.Load("Jazz").Compositions, Utils.LoadCompositionsParallel(@"D:\Sync\4th year\Midi\Library2\Drums") };
foreach(Composition[] catt in comps)
{
foreach(var c in catt)
{
foreach(var t in c.Tracks)
{
if(t.Channel == 10)
{
drumSeqs.Add(t.GetMainSequence() as MelodySequence);
}
}
}
}
List<Composition> drums = new List<Composition>();
foreach(var m in drumSeqs)
{
Composition c = new Composition();
Track t = new Track(PatchNames.Acoustic_Grand, 10);
t.AddSequence(m);
c.Add(t);
drums.Add(c);
}
var catdrum = new CompositionCategory("Drums", "lib/Drums", drums.ToArray());
catdrum.Save("lib/Drums");
Console.ReadLine();
*/
/*Composition comp = Composition.LoadFromMIDI(@"C:\Users\1gn1t0r\Documents\git\GeneticMIDI\GeneticMIDI\bin\Debug\test\other\twinkle.mid");
float time = Note.ToRealDuration(comp.GetLongestTrack().Duration);
Console.WriteLine("Total time: {0}", time);
MusicPlayer player = new MusicPlayer();
player.Play(comp);
Console.ReadLine();*/
var twink = Composition.LoadFromMIDI(@"C:\Users\1gn1t0r\Documents\git\GeneticMIDI\GeneticMIDI\bin\Debug\test\other\twinkle.mid");
Databank db = new Databank("lib");
var cat = db.Load("Jazz");
Codebook<Note> book = new Codebook<Note>();
List<int[]> sequences = new List<int[]>();
foreach(var comp in cat.Compositions)
{
foreach(var track in comp.Tracks)
{
if(track.Instrument == PatchNames.Trumpet)
{
var mel = track.GetMelodySequence();
mel.Trim(80);
mel.StandardizeDuration();
book.Add(mel.Notes);
sequences.Add(book.ToCodes(mel.Notes));
}
}
if (sequences.Count > 10)
break;
}
DotNetLearn.Markov.HiddenMarkovModel hmm = new DotNetLearn.Markov.HiddenMarkovModel(60, book.TotalUniqueSymbols);
hmm.UniformRandomPriors();
hmm.MaxIterations = 50;
hmm.Train(sequences.ToArray());
var o = hmm.PredictObservationSequence(80);
var newmel = new MelodySequence(book.Translate(o));
Track t = new Track(PatchNames.Trumpet, 2);
t.AddSequence(newmel);
Console.ReadLine();
MusicPlayer player = new MusicPlayer();
player.Play(t);
player.Stop();
Console.ReadLine();
/*
var gen = new SamplingWithReplacement(cat, PatchNames.Acoustic_Grand);
var mel = gen.Generate();
Console.WriteLine(mel.ToString());
MusicPlayer player = new MusicPlayer();
player.Play(mel);
Console.WriteLine("Press enter to save");
Console.ReadLine();
WriteMelodyToFile(gen.OriginalSequence, "sampling_in.mid");
WriteMelodyToFile(mel, "sampling_out.mid");
*/
/* MusicPlayer player = new MusicPlayer();
//db.LoadAll();
AccompanyGeneratorMarkov genMark = new AccompanyGeneratorMarkov(cat);
var targetSeq = cat.Compositions[2].GetLongestTrack().GetMainSequence() as MelodySequence;
var seqTest = genMark.Generate(targetSeq,5);
Track trackTest = new Track(PatchNames.Orchestral_Strings, 2); trackTest.AddSequence(seqTest);
Track targetTrack = new Track(PatchNames.Acoustic_Grand, 3); targetTrack.AddSequence(targetSeq);
Composition comp = new Composition();
comp.Add(trackTest);
comp.Add(targetTrack);
player.Play(comp);
return;*/
/* Databank db = new Databank("lib");
var cat = db.Load("Classical");
//AccompanimentGenerator2 Test
AccompanimentGenerator2 gen = new AccompanimentGenerator2(cat, PatchNames.Orchestral_Strings);
gen.Train();
//
Composition comp = Composition.LoadFromMIDI(@"D:\Sync\4th year\Midi\Library2\Classical\Mixed\dvorak.mid");
//var comp = Composition.LoadFromMIDI(@"C:\Users\1gn1t0r\Documents\git\GeneticMIDI\GeneticMIDI\bin\Debug\test\ff7tifa.mid");
gen.SetSequence(comp.Tracks[0].GetMainSequence() as MelodySequence);
MusicPlayer player = new MusicPlayer();
Console.WriteLine("Press enter to listen");
Console.ReadLine();
var mel = gen.Generate();
Composition newComp = new Composition();
Track newTrack = new Track(PatchNames.Orchestral_Strings , 2);
newTrack.AddSequence(mel);
newComp.Add(newTrack);
/*comp.Tracks[0].Instrument = PatchNames.Acoustic_Grand; (comp.Tracks[0].GetMainSequence() as MelodySequence).ScaleVelocity(0.8f);
/* newComp.Tracks.Add(comp.Tracks[0]);
player.Play(newComp);*/
Console.ReadLine();
}
static void LoadSongTest()
{
Console.WriteLine("Playing test ff7tifa");
MusicPlayer player = new MusicPlayer();
Composition c = Composition.LoadFromMIDI("test/other/ff7tifa.mid");
player.Play(c);
}
static void InstrumentalTest()
{
Databank db = new Databank("lib");
var cat = db.Load("Classical");
InstrumentalGenerator gen = new InstrumentalGenerator(cat);
gen.Initialize();
var outMel = gen.GenerateInstrument(PatchNames.Acoustic_Grand, 40);
MusicPlayer player = new MusicPlayer();
player.Play(outMel);
Composition comp = new Composition();
Track t = new Track(PatchNames.Acoustic_Grand, 3);
t.AddSequence(outMel);
comp.Add(t);
comp.WriteToMidi("instrumental_acousticgrand.mid");
}
static void GeneticTest()
{
Databank db = new Databank("lib");
var cat = db.Load("Classical");
MetricSimilarity cosine = MetricSimilarity.GenerateMetricSimilarityMulti(cat.Compositions, new IMetric[] { new ChromaticToneDistance(), new MelodicInterval(), new ChromaticToneDuration(), new RhythmicBigram() });
GeneticGenerator gen = new GeneticGenerator(cosine, PatchNames.Orchestral_Strings, cat);
gen.OnPercentage += (sender, percentage, fitness) => { Console.WriteLine("{0}: {1}", percentage, fitness); };
gen.MaxGenerations = 1000;
var outMel = gen.Generate();
MusicPlayer player = new MusicPlayer();
player.Play(outMel);
Composition comp = new Composition();
Track t = new Track(gen.Instrument, 3);
t.AddSequence(outMel);
comp.Add(t);
comp.WriteToMidi("genetic_cosine_all.mid");
}
static void DemoTest2()
{
MusicPlayer player = new MusicPlayer();
Composition comp = new Composition();
Track track1 = new Track(PatchNames.Acoustic_Grand, 1);
Track track2 = new Track(PatchNames.Cello, 2);
Console.WriteLine("Hidden Markov Model, constant duration, twinkle");
var mel1 = GetMelodySequence(@"test\harry.mid");
var stoch = new StochasticGenerator(new MelodySequence[] { mel1 });
track1.AddSequence(stoch.Generate());
track2.AddSequence(stoch.Generate());
//comp.Add(track1);
comp.Add(track2);
player.Play(comp);
return;
}