github.com/r9y9/nnsvs

Test on a Neural source filter (NSF) model
NOTE: Training is still in progress (350k steps out of 400k)
NOTE: Haven't done any optimization yet

ust: www.nicovideo.jp/watch/sm39319737

Statistical parametric SVS (sampling rate: 48000)
Time-lag model: {
"_target_": "nnsvs.model.MDN",
"in_dim": 331,
"out_dim": 1,
"hidden_dim": 32,
"num_layers": 3,
"dropout": 0.5,
"num_gaussians": 1
}
Duration model: {
"_target_": "nnsvs.model.MDN",
"in_dim": 331,
"out_dim": 1,
"hidden_dim": 256,
"num_layers": 3,
"dropout": 0.5,
"num_gaussians": 1
}
Acoustic model: {
"_target_": "nnsvs.model.ResSkipF0FFConvLSTM",
"in_dim": 335,
"out_dim": 206,
"ff_hidden_dim": 1024,
"conv_hidden_dim": 512,
"lstm_hidden_dim": 256,
"dropout": 0.0,
"bidirectional": true,
"in_lf0_idx": 286,
"out_lf0_idx": 180,
"in_lf0_min": 5.162578582763672,
"in_lf0_max": 6.491110801696777,
"out_lf0_mean": 5.885598413281433,
"out_lf0_scale": 0.23067350503371858
}
Vocoder model: {
"generator_type": "HnSincNSF",
"generator_params": {
"aux_context_window": 2,
"cin_channels": 67,
"cnn_kernel_s": 5,
"cnn_num_in_block": 10,
"filter_block_num": 5,
"harmonic_num": 7,
"hidden_dim": 64,
"noise_std": 0.003,
"out_channels": 1,
"out_lf0_idx": 60,
"out_lf0_mean": 5.885598413281433,
"out_lf0_scale": 0.23067350503371858,
"out_vuv_idx": 61,
"sample_rate": 48000,
"sinc_order": 31,
"sine_amp": 0.1,
"upsample_net": "ConvInUpsampleNetwork",
"upsample_params": {
"upsample_scales": [
2,
3,
4,
10
],
"aux_channels": 67,
"aux_context_window": 2
},
"upsample_rate": 240
}
}

2022/04/06 Kiritan test SVS w/ NSF 350k steps Roshin Yuukai

Just some guy that loves music.
Sometimes makes some covers, but don't expect much

Avatar: 91zM0_bot

23 | Bi-Gender | British | Artist | Reposts a lot

CV for Kiiro, Kazuya, Miyuki, Kei

ᵖˡˢ ᵍᵒ ᵒⁿ ᵗʰᵉ ᵗʳᵃᶜᵏˢ ᵗᵃᵇ ᶦᵐ ᵇᵉᵍᵍᶦⁿᵍ ʸᵒᵘ

Creating UTAU covers, making mashups and reposting extremely loud music you probably won't like.
CV脆音リリア (Morone Lilia)

i will try to make the worst utau covers of all time

Not posting on SoundCloud anymore, YouTube only!
もうSoundCloudには投稿せず、YouTubeのみになります
応援ありがとうございました

25 yo gay ー he/him ー EN/IT/日本語OK ー f(x) & Britney fan ー producer, singer ー Yukio Acqua voicer
mic: RØDE NT-USB
software: FL Studio, VOCALOID4 (MEIKO V3), SynthV Pro (GENBU, Kasane Teto, Frimomen, Otomachi Una)
email: idoltrash.utau☆gmail.com (☆ → @)

previously 光る音君(Raiã).

email me for access to voice banks pre-release. 

email: uyopuyo.karunekun@outlook.com

VOICEBANKS available for pre-release:
龍花ユウタ - DeepVocal
DEMO: https://soundcloud.com/user-100576288/behind-the-door-deepvocal-remake
龍花ユウタU - UTAU CVVC 1.0
DEMO: https://on.soundcloud.com/aqmXw
光音ウラ - DeepVocal
DEMO: https://on.soundcloud.com/i19BZ
光音ウラ - VocalSharp
DEMO: https://on.soundcloud.com/c1Qqk
アマナ - DiffSinger
DEMO: https://on.soundcloud.com/zss9G

Released Voicebanks:
光音ウラCVVC1.0 https://www.youtube.com/watch?v=Kj1qyYf9YJI
光音ウラCV-VV (cute)https://www.youtube.com/watch?v=0S3XeaIhZrI
光音ウラ CVVC1.2 https://www.youtube.com/watch?v=WMd_TZhBrAA

xuu. amano naoki (天野なおき) cv.
synthesizer v / diff-svc / nnsvs+enunu / utau
generally all synthesizers

@xuutaite