ABCDEFGHIJKLMN
1
SR
DateDescriptionDatasetData
Epochs
Duration
Iterations
Train
Validation
Test
Converged?
Score
Comments
2
441002018 01 29jlin drumsjlin4655160.9220800.70.70.7cool6wow - recreates interesting syncopation
3
441002018 02 01sascha kuschel excerptkuschel4661282801.81.81.9great!6slurpy, wet sounds
4
2018 02 08jlin drumsjlin465546634800.50.60.6great6much more coherent rhythmic structures!
5
2018 02 16jlin drumsjlin465562855600.50.60.6great!6much better rhythm. seems to have 2-3 different tempo modes
6
2018 02 21ae "vi scose poise"scosepoise50916165601.21.21.2great!6gets the scose poise sound down... quieter bits sound just like the album!!
7
2018 02 21"can't sleep" vocalcantsleep_vocal50320207001.11.21.2great!6ethereal violin-like vocals right off the bat!
8
2018 03 07"what if" vocode longerwhatifvocode507882801.71.81.8great!6says "what if" really clearly, over and over, really good!
9
2018 03 14"what if" vocode longerwhatifvocode50721217351.51.61.6great!6more and more refined - gorgeous!
10
2018 03 20cattle auctioneersauctioneer510662101.21.21.2great!6this one is amazing
11
441002018 01 19
tap glass with spoon, layered
glassmix13091718.0478001.11.11.1great!5REALLY good. lots of interesting variations. rendering at higher resolution helped - double the data!
12
441002018 01 25jlin drumsjlin4655104.5138000.80.80.8great5more intricate rhythms
13
441002018 01 26jlin drumsjlin4655125.4193200.70.70.7great5seems to have plateaued...?
14
441002018 01 29uneasy melodyuneasy49781.482801.41.41.4great5really sick boards of canada vibe
15
441002018 01 29rashad becker "dances"dances590282.7291001.82.12.1great5dark noise, tonal patterns
16
441002018 01 30
flufdril_2 just drums @172bpm
flufdril_2a_drums74521289801.51.61.6good5rapid tapping, more drone/tone focusing
17
441002018 01 30jlin drumsjlin465524331200.60.60.6good5got sorta gappier, some cool small drum parts
18
441002018 02 01jlin drumsjlin465534469200.50.60.6great5beats have some structure now! seems slightly less arbitrary
19
441002018 02 04jlin drumsjlin465538524400.50.60.6great5rhythmic structure is very clear - larger chunks with syncopation
20
441002018 02 04rashad becker "dances v"dancesv45445310501.01.21.1great5really moody and interesting!
21
441002018 02 05jlin drumsjlin465541565800.50.60.6good5structure improves slowly / subtly
22
2018 02 07jlin drumsjlin465544607200.50.60.6good5more subtle improvements. seems to like one 3 beat rhythm
23
2018 02 08glassloop_left prunedglass_space155622227702.02.22.1great5noise decreases the further it trains.. but there is more silence and the breaking is further apart!
24
2018 02 09jlin drumsjlin465550690000.50.60.6great5good rhythm for stretches
25
2018 02 09vsnares "sundae"sundae41156386402.42.62.6great5develops more percussion, interesting rolls, still no larger structure (but original didnt have much)
26
2018 02 12jlin drumsjlin465554745200.50.60.6great5structure seems to last 2-3 seconds if it happens at all
27
2018 02 13jlin drumsjlin465558800400.50.60.6great5rendered 10s segments. good pattern, less rhythmic slippage. got a little weird static!
28
2018 02 16track 'fm9'fm954014144901.82.01.9great5mostly tone runs from the beginning of the track. funny bass/vocal tone in the background
29
2018 02 16hito "this is all i can say"thisisallicansay49571734851.71.92.0great5gets better at generating the phrase regularly, but still breaks up a bit
30
2018 02 19jlin drumsjlin465566910800.50.60.6great5still going.. wish it would generate triplets
31
2018 02 21autechre knownknown50814144901.51.61.6great5more of the same tones.. semi-evolving but not super varied
32
2018 02 21autechre seeonseeseeonsee51814144902.02.02.0great5very consistent.. generates legato changes. only a couple constant sticky tones.
33
2018 02 21"can't sleep" synthcantsleep_synth51116165602.32.42.4great5trumpet-like sounds - mostly slowly evolving tones
34
2018 02 21brigitte fontaine "l'ete l'ete"brigitte5042424248401.41.61.6great5generated a lot of vocal loops and horn hits
35
2018 02 21"can't sleep" source mixcantsleep_mix51224248402.22.32.3great5very windy.. ghostly vocal bits gradually appear
36
2018 02 21henry cow "teenbeat"teenbeat50324248402.22.42.4great5tense trumpet tones.. cymbals and rhythms evolve..
37
2018 02 21jlin drumsjlin465570966000.50.50.5great5pretty baseline consistent
38
2018 02 25ae "vi scose poise"scosepoise50920207001.11.21.2great5endless scose poise
39
2018 02 26ae "vi scose poise"scosepoise50924248401.11.21.2great5more scose poise variations. performs really well. it even gets better
40
2018 02 26sascha kuschel excerptkuschel46627186301.91.92.0great5making normal sounds again. didn't do anything different. fluke?
41
2018 02 26jlin drumsjlin4655801104000.50.50.5great5slight improvements but tbh this one is done
42
2018 02 28ae "vi scose poise"scosepoise50958600301.01.11.1great5half an hour of vi scose poise baby!!
43
2018 03 02ae "vi scose poise"scosepoise50964662400.91.11.1great5highly developed scose poise...
44
2018 03 03lyra bladerunner vocalslyraloop503882800.70.70.7great5really good singing! doesn't fixate so much.
45
2018 03 03lovelove looplovelove508882801.41.41.5great5produces "love love love" chants with odd spacing
46
2018 03 04lovelove looplovelove50817175951.31.31.3great5generates really well, doesn't skitter around
47
2018 03 06reich "clapping music"clapreich508662101.11.11.1great5really clear clapping happening at random
48
2018 03 06"music for pieces of wood"wood505662101.11.21.2great5random wood noises. sounds good
49
2018 03 06reich marimba musicmarimbareich511662101.41.51.5great5very clear marimba and organ tones (professionally recorded). becomes less dense toward the end
50
2018 03 07"music for pieces of wood"wood505772451.11.11.1great5the more it trains, the more silence it generates
51
2018 03 07reich "clapping music"clapreich50810103501.01.01.0great5clapping is super clear and begins to make more sense. fairly consistent rhythm
52
2018 03 07reich marimba musicmarimbareich51110103501.31.41.4great5fixated, but preserved timbre in a nice way. sustained organ sound
53
2018 03 10reich "clapping music"clapreich50814144900.90.90.9great5claps get farther apart
54
2018 03 10f.a.t.y. drumsjlin-faty51923238051.31.31.3great5very clear, trains like the other jlin
55
2018 03 11clapping fasterclapmulti51213134551.71.71.7great5produces claps, good room tone, random rhythms, cool combos of room saturation
56
2018 03 11stamping feetclapfeet50313134551.81.91.9great5foot tapping, with subtle wind noise
57
2018 03 11lip noiseslipnoise51413134553.43.53.5great5have to listen to this closer but it produces an extremely varying signal
58
2018 03 11"what if" vocode longerwhatifvocode50717175951.61.71.7great5would be interesting to see this go way farther
59
2018 03 12vsnares "sundae"sundae41173503702.12.62.6great5more snares chaos, still quite noisy
60
2018 03 13police sirenssiren 507662101.71.71.7great5like a child making siren noise
61
2018 03 14faty vocal2faty-vocal2506662102.22.22.2great5nice melody, really like this one
62
2018 03 14faty scrub2faty-scrub2501882800.70.80.8great5great melody yay
63
2018 03 14police sirenssiren 50710103501.51.61.6great5great siren stuff!
64
2018 03 14lip noiseslipnoise51417175953.43.53.5great5interesting dynamic - hard to tell if it's fixating - in a way, the sound resembles wind
65
2018 03 14f.a.t.y. drumsjlin-faty51935362251.21.31.3great5great drums, sounds amazing layered over the other parts
66
2018 03 20police sirenssiren 50714144901.51.61.6great5super sick siren noises as always
67
2018 03 20f.a.t.y. drumsjlin-faty51940414001.11.21.2great5loss stopped improving, more of the same
68
2018 03 24lip noiseslipnoise51416165603.33.43.4great5still good
69
2018 03 24autechre knownknown50817175951.51.61.6great5more weird evolving tones
70
2018 03 24police sirenssiren 50718186301.41.51.5great5still good
71
2018 03 24lovelove looplovelove50819196651.21.31.3great5interesting how "love" varies in length
72
2018 03 24"what if" vocode longerwhatifvocode50729300151.51.61.6great5gradually loses sentence structure as it overfits - whatifwwwwwww
73
2018 03 24sascha kuschel excerptkuschel46635362252.02.02.1great5loss stopped improving.. really interesting long noise sweeps taking place over time
74
2018 03 24f.a.t.y. drumsjlin-faty51952538201.11.21.2great5loss stopped improving, more of the same
75
2018 03 25tuuun - fmtrax 11tuuun12124202.52.62.6great5cool variations! must do more fm experiments.
76
2018 03 28evol fluf tracksevol2662100.91.00.9great5still good
77
2018 03 28evol fluf tracksevol1662101.31.31.3great5gets smoother with each iteration, you can hear it
78
2018 03 28billy splitterboosplitterboo1662101.71.71.7great5reproduces the sound with interesting variations. working well by epoch 6
79
2018 03 28evol fluf tracksevol2993150.90.90.9great5sounds interesting yes
80
2018 03 28evol fluf tracksevol112124201.11.21.2great5more smoothness, great
81
2018 03 28billy splitterboosplitterboo112124201.61.61.6great5works well, interesting variations
82
220502018 01 17
tap glass with spoon, layered
glassmix1309167.3221441.71.81.8good4actually sounds like the glass sample! low end sounds more like low-pitched noise
83
220502018 01 17all 3 singing "ah"ah_mix_long635162.883042.42.62.6good41st epoch already sounded good. 16th epoch SUPER clear, could hear everyone's voices.
84
220502018 01 20ensemble going "ah ah ah"ensemble796123.383002.52.62.6good4learned "ah ah ah" pretty quickly.. and then forgot it. toward the end it generated a lot of silence.
85
441002018 01 22ensemble compressedensemble238782.355202.92.92.9good4much better than previous ensemble training. compression helped, but didn't saturate.
86
441002018 01 24jlin drumsjlin465583.6110400.80.80.8good4continued processing.. wandering drum patterns, able to generate all the drum sounds
87
441002018 01 25
flufdril_2 just drums @172bpm
flufdril_2a_drums74542.155201.91.91.9good4very sporadic percussion!!
88
441002018 01 25ensemble hocketchapter373142.155202.42.32.3good4extremely whistly! comes into focus after an interation or two. still sounds like a pan flute.
89
441002018 01 26
flufdril_2 just drums @172bpm
flufdril_2a_drums74563.282201.71.71.7good4kicks are pretty big. harsh noise very quiet in background
90
441002018 01 26plastic cut drums @172bpmplasticcut_4600104.1103502.02.02.0good4
interesting variations of the source material, emphasis on bass notes. bird noises turned into whistling
91
441002018 01 28
flufdril_2 just drums @172bpm
flufdril_2a_drums74531.6179401.61.71.7good4slightly more active patterns
92
441002018 01 28jlin drumsjlin4655140.9193200.70.70.7good4not changing much but subtly different structure compared to older epochs
93
441002018 01 29shlucht "stolen pics"stolenpics41841.227602.12.12.1good4little pops and wind :)
94
441002018 01 29plastic cut drums @172bpmplasticcut_4600121.2155251.81.91.9good4longer patterns
95
441002018 01 29vibra tres @172bpmvibratres492140.7144900.70.80.8good4more arpeggiation..
96
441002018 01 30shlucht "stolen pics"stolenpics4181282801.91.91.9good4very cheeky tones !!
97
441002018 01 30plastic cut drums @172bpmplasticcut_460018186301.81.91.9good4more of the same
98
441002018 02 01uneasy melodyuneasy49728289801.01.11.1good4surprisingly generated sections of 3-4 notes
99
441002018 02 01plastic cut drums @172bpmplasticcut_460030310501.61.81.7good4likes weird bass notes
100
441002018 02 01
flufdril_2 just drums @172bpm
flufdril_2a_drums74533455401.41.51.5good4more of the same, still pretty spaced out
101
441002018 02 05glass foley (stretch)glassloop_stretch4601496603.23.23.2good4less whistling by the end - probably the best?
102
441002018 02 06re-rendered just quiet glassglass_space2544882801.71.81.8good4more whistly, but nice patterns
103
441002018 02 06glassloop_left prunedglass_space1556882802.32.32.3good4more defined "glass" sound, crystal tones. whistles the least.
104
441002018 02 07glassloop_left prunedglass_space155616165602.12.22.2good4still the best - some bursts of noise but it seems manageable
105
2018 02 08re-rendered just quiet glassglass_space254414144901.61.61.6good4very few crashes - mostly little pops and things
106
2018 02 08vsnares "sundae"sundae41136248402.52.62.6good4keeps developing... still noisy but maybe that's because the original recording is so loud
107
441002018 02 12rashad becker "dances v"dancesv45461420901.01.11.1good4no real development or new structure
108
2018 02 13hito "this is all i can say"thisisallicansay49536372601.81.92.0good4all shuffled up at the beginning, but almost generates a coherent sentence by the end
109
2018 02 15susumu "hagoromo" editsusumu44371069000.80.80.8good4nice drones, overfit by ep 10
110
2018 02 16(h)ywfd mixed acapellasywfdacc47615103503.03.13.1good4very windy - many voices mixed in background. occasional vowel overlap
111
2018 02 18"we found love" songwefoundlove-song50611113853.83.93.9good4random rhythms, a vague build. the organ stab is prominent toward the end
112
2018 02 18cata talkingcatamachine54814144901.31.41.4good4says "machine" a lot!!
113
2018 02 19autechre knownknown50810103501.61.71.7good4hypnotic autechre pads
114
2018 02 19autechre seeonseeseeonsee51810103502.12.22.1good4big ambient clouds of a couple forms
115
2018 02 19cee-lo "fuck you" acca-fuck55610103502.32.42.4good4very strained ceelo-like vocals
116
2018 02 19avril "what the hell" acca-hell50914144902.12.32.3good4consistent vocals, piercing noise blobs, could be cleaned up
117
2018 02 21"main artery of the mother"artery496882800.70.70.7good4really haunting noise. a bit staticky tho
118
2018 02 21harp of king davidharpdavid510882801.01.01.0good4deep singing.. the voice mixes with the harp
119
2018 02 25harp of king davidharpdavid51010103500.91.01.0good4seems to have fixated on a couple tones
120
2018 02 25susumu "hagoromo" editsusumu44371496600.70.80.8good4fixates, harp/violin tones
121
2018 02 25"can't sleep" synthcantsleep_synth51120207002.22.42.4good4fixates but has more variety
122
2018 02 25"can't sleep" source mixcantsleep_mix51228289802.12.32.3good4fixates, voice is most present thing from the original sound
123
2018 02 26"can't sleep" synthcantsleep_synth51124248402.22.32.3good4fixates, no change
124
2018 02 26"can't sleep" vocalcantsleep_vocal50328289801.01.21.2good4fixates, no change
125
2018 02 26"can't sleep" source mixcantsleep_mix51232331202.12.32.2good4fixates, no change
126
2018 02 26vsnares "sundae"sundae41170483002.42.62.6good4no real change, producing weird rhythms
127
2018 02 28f.a.t.y. drumsjlin-faty519882801.61.61.6good4sounds promising
128
2018 02 28f.a.t.y. voiceholly-faty509882801.92.02.0good4diverse loops and stuff, fwiw
129
2018 02 28psalm compressedpsalm_comp50512124202.12.22.2good4gets the guy's voice well, avoids sustained tones, but kind of random
130
2018 03 03lucierization of the officelucier511882800.90.90.9good4whistly, ghostly... interesting upper harmonics develop. mostly a big air cloud.
131
2018 03 03holly bladerunner chorushhbr510882801.11.11.1good4disordered vocal shapes, fixates but has cool vibrato / crescendo
132
2018 03 04f.a.t.y. voiceholly-faty50910103501.91.91.9good4really clear, doesn't seem to fixate exactly
133
2018 03 04f.a.t.y. drumsjlin-faty51920207001.31.41.4good4scattered percussion, makes all the notes
134
2018 03 06"chaotic alien wayne"chaotic508662101.61.61.6good4vocals emerging from wind. begins to fixate
135
2018 03 06"what if" vocode longerwhatifvocode507662101.81.81.8good4bizarre sounding vocoded voices!!
136
2018 03 12plastic cut drums @172bpmplasticcut_460032331201.61.81.7good4not much change, still unfocused, fixates on bass
137
2018 03 12
flufdril_2 just drums @172bpm
flufdril_2a_drums74536496801.41.51.5good4more aggressive drums..
138
2018 03 14rashad becker "dances"dances59063652051.72.12.1good4more of the dances stuff, still noisy
139
2018 03 14vsnares "sundae"sundae41177531302.42.62.6good4trained well but i find this kind of boring
140
2018 03 20throat fry noisethroat515662101.11.11.1good4sustained throat fry, some bursts of noise
141
2018 03 20dropping coinscoins509662101.71.71.7good4started generating coin drops after epoch 6
142
2018 03 20"what if" vocode longerwhatifvocode50725258751.51.61.6good4oddly less interesting than before. epoch 9 was best
143
2018 03 27iruuuuu "bufrd" trackirubufrd662101.01.01.0good4quiet occasional singing, some tone fixation
144
2018 03 28billy splitterboosplitterboo2662100.70.60.8good4the broken-up sound is now even more broken up
145
220502018 01 17
stacked "ahs" form major chord
bigahchord629152.877853.73.83.8ok3renders had areas with clear voices, but sometimes were overtaken by LOUD static
146
220502018 01 20holly granular stuffhollydream753123.383002.02.32.3ok3given this was a lot of granular/fft freeze stuff it produced similar material with no real transitions
147
220502018 01 20ensemble singing chordsomnir39260121.121002.12.12.1ok3very noisy, learned some phrasing. maybe the most promising once we get the dataset together
148
441002018 01 23clouds35bclouds264945.455200.70.60.6good3produced bubbily notes like the original. later training epochs popped less
149
441002018 01 23jlin drumsjlin465541.855201.00.90.9good3some drum noises, sporadic
150
441002018 01 23jlin master mixjlin367941.855201.41.41.4good3drum hit + vocal, more coherent sequences, bass sound
151
441002018 01 23
jlin mix, drums, vocal in sequence
jlin150041.341401.41.41.4ok3drum hit + vocal drone
152
441002018 01 23jlin holly vocal overlappedjlin246440.927602.32.32.3good3wandering noisy vocals
153
441002018 01 26ensemble hocketchapter373182.155201.91.91.9ok3more training didn't help! just made it more airy
154
441002018 01 27ensemble hocketchapter373152.369002.32.22.2ok3re-training but stopping sooner. got some maniacal flute stuff toward the end
155
441002018 01 28vibra tres @172bpmvibratres49220.7103500.70.80.8ok3not exciting development, seems to have converged, wish it would go farther..
156
441002018 01 28ensemble hocketchapter373160.582802.22.22.2ok3trained one more step to see if i can generalize why i like these epochs more
157
441002018 01 29rashad becker "dances v"dancesv45441.227601.41.41.4ok3a lot of silence, some pops, a big foghorn tone! wind
158
441002018 01 30vsnares "sundae"sundae4111282802.62.72.7ok3needs way more training but it's quite interesting
159
441002018 01 30uneasy melodyuneasy49716165601.21.21.2ok3not much development
160
441002018 02 01rashad becker "dances v"dancesv4544220801.11.21.2ok3got some stuff.. promising?
161
441002018 02 01vibra tres @172bpmvibratres49226269100.60.70.7ok3starts getting more active. still pretty indecisive, never got the structure
162
441002018 02 01shlucht "stolen pics"stolenpics41826165601.71.81.8good3cheeky tone is still prominent. harmonizes with itself once??
163
441002018 02 01vsnares "sundae"sundae41126179402.52.62.6ok3more chaotic / violent, able to reproduce a couple phrases at the end
164
441002018 02 01rashad becker "dances"dances59060621001.72.12.1ok3found a couple tones it likes, on the whole seems too complex to train on..
165
441002018 02 04sascha kuschel excerptkuschel46624165601.81.92.0ok3not too much development, still nice pops / splashy sounds
166
441002018 02 05glass foley (L/R/mix)glassloop4601496602.93.02.9ok3produced glass sounds, all on top of each other, some whistling
167
441002018 02 05glass foley (L)glassloop_left4601496603.03.13.1ok3whistly and vague
168
441002018 02 07space3 reversedglass_space_rev526882801.31.31.3ok3some good renders, no whistling, a lot of silence
169
2018 02 07rec1 compressedglass_rec1_comp446855201.61.61.6ok3best so far of the ones i recorded. a bit whistly, noisy
170
2018 02 07re-rendered just quiet glassglass_space254412124201.61.61.7ok3patterns ok but still whistly
171
2018 02 13susumu albumsusumu250419196651.01.11.0ok3sustained harp stuff. a little schizo given it trained on bits of 12 songs. some good tones at the end
172
2018 02 13lonne looplonneloop55919196651.11.21.2ok3very very bassy. ignored most of the song.
173
2018 02 15susumu "azukiiro no kaori"susumu352910103500.80.80.8ok3nice drones, got vocal, but also loud bursts of noise randomly
174
2018 02 15track 'fm9'fm954010103501.92.02.0ok3wobbly versions of the fm notes. promising technique for making weird fm sounds?
175
2018 02 16cata's 'clamstep'clamstep568882803.63.63.6ok3midrange whistle - like a howl - some shapes in the background resembling original sound
176
2018 02 18jules + cata simul talkingfreebsd52314144901.51.51.5good3weird talking, but how can we merge the voice?
177
2018 02 19waka "o lets do it"oletsdoit498551752.52.52.5ok3trained all weird. interesting combinations of the pager "bleep" sound and horns
178
2018 02 19cata talkingcatamachine54822227701.21.31.3ok3not much development
179
2018 02 19jules + cata simul talkingfreebsd52324248401.41.51.5ok3strained cycles
180
2018 02 21nsync "bye bye bye"byebyebye50816165603.03.13.1ok3some swipey nsync sounds, mostly caught in tones
181
2018 02 21waka "o lets do it"oletsdoit49822227702.02.12.1ok3mostly quiet static... almost sounds like data-loading / hard disk noise
182
2018 02 25"can't sleep" vocalcantsleep_vocal50324248401.11.21.2ok3fixates, vocal whorls
183
2018 02 26school chant 3school3483882801.21.21.2ok3boy vocalizations
184
2018 02 26school chant 1school1521882801.41.41.4ok3hoarse vocal noises
185
2018 02 26school chant 2school2511882801.41.51.5ok3early monster vocalizations, then breathy/fry voices
186
2018 03 04mix of bladerunner voxbrmix497882801.21.31.3ok3fixates, lyra voice dominates
187
2018 03 04psalm, 115 ms grainspsalm115509882801.51.51.5ok3fixates a bit more
188
2018 03 04psalm, 234 ms grainspsalm234509882801.51.51.5ok3fixates more
189
2018 03 04psalm, 468 ms grainspsalm468509882801.51.61.6ok3fixates on a couple quiet notes
190
2018 03 04psalm, 55 ms grainspsalm055509882801.51.61.6ok3skips around, likes producing a couple voices only
191
2018 03 04lucierization of the officelucier51115155250.70.80.8ok3fixates, occasionally with some upper harmonics
192
2018 03 07"chaotic alien wayne" shortchaotic2508993151.31.41.4ok3fixates, breathy singing
193
2018 03 07"what if" natural 30secwhatifshort504993151.31.41.4ok3babbles but says "what" from time to time
194
2018 03 07"what if" natural longerwhatifclean50912124201.31.31.3ok3babbles but says "what" in a couple different voices
195
2018 03 12rashad becker "dances"dances59062641701.72.12.1ok3super noisy as before
196
2018 03 14bubbles retrainedblblbl2503662101.31.31.3ok3same as first time - first epoch busy, second epoch more sparse, rest silent
197
2018 03 20faux eddie vedder singingvedder519662101.41.51.4ok3random wailing
198
2018 03 25huun huur tutuva12124201.61.71.7ok3many noise bursts
199
2018 03 28billy splitterboosplitterboo212124200.70.70.7ok3too much space between notes, inconsistent
200
441002018 01 22holly tts training sentenceshollytrainset73384.2110402.42.52.4ok2got some voice, but the more it trained, the more it turned into noise. epochs 7-8 the best
201
441002018 01 28cata's harsh noisetextile44640.975904.04.14.0ok2still big high pitched tones dominating. the harsh noise bits are more interesting tho!
202
441002018 01 30rashad becker "dances"dances59044455401.82.12.1ok2cool stuff but might need 2x-3x more training to denoise
203
441002018 02 01cata's harsh noisetextile44619131104.04.14.0ok2high pitched sound everywhere but the noise bursts more interesting
204
441002018 02 05glass foley (L compressed)glassloop_comp4601496603.13.33.2ok2very whistly
205
441002018 02 06space3 louder, roll off highsglass_space4573882801.31.31.3ok2less defined high end, whistles, not very glassy
206
2018 02 16"we found love" accwefoundlove4761496602.82.82.9ok2high pitched tone throughout. eventually stops rendering voice
207
2018 02 18liam talking (18 seconds)liam4511496600.90.90.9ok2mostly silence. some voice. seems like it picked up on the split-second of very quiet music mostly!
208
2018 02 25"main artery of the mother"artery49612124200.60.70.7ok2has a bit of static, no rhythm
209
2018 02 26harp of king davidharpdavid51014144900.90.90.9ok2fixates, on tones
210
2018 02 26"main artery of the mother"artery49616165600.50.60.6ok2staticky, irregular heartbeat
211
2018 02 28psalm cut up rearrangedpsalm_cut50812124201.71.81.8ok2certain epochs were very noisy. moves breathily between passages
212
2018 03 03really short countdowncountdown500551750.80.90.9ok2struggles to say "one"
213
2018 03 04really short countdowncountdown500772450.80.80.8ok2fixates, struggles to say "one"
214
2018 03 06"body percussion"bodyperc509662102.02.02.0ok2mostly silence. develops some noise around epoch 5 but it goes away
215
2018 03 07"short zumayyah"zumayyah501662101.91.91.9ok2fixated on "ah", different voice appears a few epochs in
216
2018 03 07"chaotic alien wayne"chaotic508882801.51.61.5ok2fixating on tones
217
2018 03 11cee-lo "fuck you" repeatedfuck50612124201.71.71.7ok2fixates on a choral note, mostly
218
2018 03 12jlin drumsjlin4655881214400.50.50.5ok2oddly scattered
219
2018 03 13bubblesblblbl503662101.31.41.4ok2noise in first epoch - silence thereafter
220
220502018 01 14slowed down glass soundshito glass16460084.02280002.02.72.8bad1turned into pitched wind sounds and bursts of noise
221
220502018 01 17
many glissandos mixed together
gliss_long_mix1164156.3181652.93.03.0bad1loud, harsh, distinct squeak at ?? hz - vague "ah" in the background. further renders all the same.
222
220502018 01 20holly glissando + ah concatgliss_and_ah432121.842002.22.32.3bad1the constant tones made it focus on certain notes. "hallucinated" a deep "ah"
223
441002018 01 21ensemble going "ah ah ah"ensemble19621218.2496800.90.90.9bad1generated mostly silence
224
441002018 01 22clouds35bclouds47282.355200.80.90.7bad1maybe the samples were too compressed? very cloudy noises, weird bass shapes
225
441002018 01 25jlin drums processedhhprocessedjlindrums52441.641404.84.84.8bad1turned into quickly-fading noise. might need a long time to come into focus
226
441002018 01 26400 kick drumskiq42810.26900.10.10.1bad1256 quantization steps. nothing rendered. too much silence!
227
441002018 01 26400 kick drumskiq42810.36900.20.10.2bad1512 quantization steps. ditto
228
441002018 01 27cata's harsh noisetextile44671.648304.24.14.1bad1sick harsh noise in places, however mostly resolves to a high pitched tone
229
441002018 01 27vibra tres @172bpmvibratres49282.782800.80.80.8bad1occasional sequences, but mostly focused on weird versions of the long tones
230
441002018 01 27dj shlucht pocket mixpocket49482.782802.42.42.4bad1again, a lot of harsh noise
231
441002018 01 27rashad becker "dances"dances59093.093151.92.12.2bad1mostly harsh noise, i still want to train it more
232
441002018 01 28rashad becker "dances"dances59082.7175951.82.12.1bad1still super harsh noise. more iterations gives sweet monster screaming / flute bits. train more!
233
441002018 01 28dj shlucht pocket mixpocket494121.4144902.32.42.4bad1harsh noise with some structure
234
441002018 01 29vsnares "sundae"sundae41181.255202.72.72.7bad1very chaotic but does seem to resemble something
235
441002018 02 06space2 prunedglass_space3588882801.31.31.3bad1really jumbled and noisy
236
441002018 02 06space4 timestretchglass_space5573882801.31.41.4bad1whistly, wet mud sound
237
2018 02 08rec1 compress + slowglass_rec1_slow484882800.91.01.0bad1sort of glassy but focused on repeating glass tones
238
2018 02 13susumu tracksusumu148819196650.90.91.1bad1bits of harps, a lot of loud noise though. why did this train poorly?
239
2018 02 19kesha "die young" acca-die52410103502.72.72.8bad1turns into a high pitched tone
240
2018 02 26sascha kuschel excerptkuschel46625172501.92.02.1bad1harsh noise! but this was training so well before?!
241
2018 02 28madli reading in estonianmadli508110352.52.22.2bad1T2 - still really noisy
242
2018 02 28school chant tunafishtunafish50012124201.31.41.4bad1sounds underwater and demonic, turns into wordlike tones
243
2018 02 28madli reading in estonianmadli50816165601.61.61.6bad1T1 - a few words toward the beginning - really scattered
244
2018 03 03inoj loveyoudown voxloveyoudown516882802.12.12.1bad1occasional mumbled words... mostly noise
245
2018 03 04madli reading in estonianmadli508220702.22.12.1bad1mostly noise
246
2018 03 07clapping in MGBclapping500551751.01.01.0bad1mostly "air sound" silence - the clapping turned into vinyl crackle
247
2018 03 07"body percussion"bodyperc509772451.91.91.9bad1basically all silence
248
2018 03 07"what if" vocode 30secwhatifvocodeshort510993151.31.41.4bad1fixated on really quiet stuff
249
2018 03 11clapping slowlyclapsingle51813134551.31.81.8bad1mostly silence and air, a couple noisy bits
250
2018 03 14jlin drumsjlin4655911255800.70.70.7bad1strangely this seems to have... forgotten everything...? or started retraining? very strange!
251
2018 03 20jingling keyskeys502662102.82.82.8bad1harsh noise with high pitched spectrum; silence
252
2018 03 27nyz rain wire recordingrainwire662101.51.51.5bad1some whistlers during epoch 2, but the rest of the epochs are noise
253
2018 03 28nyz rain wire recordingrainwire10103501.41.51.5bad1couple swishes during epoch 1 but then turns to noise
254
2018 03 29billy "steedish"steedish551753.23.23.2bad1converged at epoch 5, mostly noise/silence. one interesting moment - noise has reverb kinda
255
441002018 01 29rashad becker "dances v"dancesv45401.2xxxxxxxx
xxxx
xxxx
xxxx
didnt train
0used pretrained, unconverged "dances"
256
441002018 01 29rashad becker "dances v"dancesv45401.2xxxxxxxx
xxxx
xxxx
xxxx
didnt train
0used pretrained, converged "glassmix"
257
441002018 02 07rec1 reversedglass_rec1_rev471641400.90.90.9bad0silence, occasional glassy noise which sounds poorly trained
258
2018 02 07broken glass recordedglass_rec1476855200.80.80.8bad0noise - glassy in aspect - doesn't improve
259
2018 02 07rec1 repitch and overlapglass_rec1_mix471855202.42.42.4bad0noise - glassy in aspect - doesn't improve
260
2018 03 13faty medievalfaty-medieval502662100.70.70.7
261
2018 03 13faty crystalsfaty-crystals501662101.41.41.4
262
2018 03 13faty scrub1faty-scrub1506662101.71.71.7
263
2018 03 13faty vocal1faty-vocal1501662101.81.81.8