Hugging Face's logo

Datasets:
nakas
/
stemflipper-dataset

Tasks:
Audio Classification
Modalities:
Tabular
Text
Time-series
Formats:
parquet
Size:
< 1K
Tags:
audio
synthesizer
audio-effects
parameter-estimation
Synthetic
Libraries:
Datasets
pandas
Polars
License:
Dataset card Data Studio Files
xet
 Community
1
Dataset Viewer
Auto-converted to Parquet Duplicate
kind
stringclasses
1 value
dry_id
stringclasses
8 values
seed
int32
0
3
sample_rate
int32
44.1k
44.1k
chain
listlengths
3
3
param_names
listlengths
25
25
params
listlengths
25
25
dry_peak
float32
0.26
1
audio
listlengths
44.1k
44.1k
effects
synth_b0_i0
0
44,100
[ "parametric_eq", "compressor", "distortion" ]
["parametric_eq.low_shelf_gain_db","parametric_eq.low_shelf_cutoff_freq","parametric_eq.low_shelf_q_(...TRUNCATED)
[3.2870805263519287,111.25947570800781,0.43277880549430847,-11.603336334228516,525.30810546875,2.773(...TRUNCATED)
0.99738
[0.0001654401421546936,0.0001674281375017017,0.00016960364882834256,0.00017215427942574024,0.0001741(...TRUNCATED)
effects
synth_b0_i0
1
44,100
[ "parametric_eq", "compressor", "distortion" ]
["parametric_eq.low_shelf_gain_db","parametric_eq.low_shelf_cutoff_freq","parametric_eq.low_shelf_q_(...TRUNCATED)
[0.283719003200531,240.58810424804688,0.5153276920318604,10.767586708068848,324.7325744628906,1.5006(...TRUNCATED)
1
[0.00040093151619657874,0.00040124033694155514,0.00040170352440327406,0.0004026814713142812,0.000401(...TRUNCATED)
effects
synth_b0_i0
2
44,100
[ "parametric_eq", "compressor", "distortion" ]
["parametric_eq.low_shelf_gain_db","parametric_eq.low_shelf_cutoff_freq","parametric_eq.low_shelf_q_(...TRUNCATED)
[-5.721308708190918,116.71331787109375,1.0513806343078613,-9.794017791748047,440.04022216796875,2.29(...TRUNCATED)
0.896526
[-0.00016467402747366577,-0.00016522558871656656,-0.0001653351355344057,-0.00016604014672338963,-0.0(...TRUNCATED)
effects
synth_b0_i0
3
44,100
[ "parametric_eq", "compressor", "distortion" ]
["parametric_eq.low_shelf_gain_db","parametric_eq.low_shelf_cutoff_freq","parametric_eq.low_shelf_q_(...TRUNCATED)
[-9.944419860839844,104.9939956665039,1.0410195589065552,1.9718888998031616,237.65145874023438,1.526(...TRUNCATED)
1
[-0.00003161192580591887,-0.000036306428228272125,-0.000039042075513862073,-0.000044175634684506804,(...TRUNCATED)
effects
synth_b0_i1
0
44,100
[ "parametric_eq", "compressor", "distortion" ]
["parametric_eq.low_shelf_gain_db","parametric_eq.low_shelf_cutoff_freq","parametric_eq.low_shelf_q_(...TRUNCATED)
[3.2870805263519287,111.25947570800781,0.43277880549430847,-11.603336334228516,525.30810546875,2.773(...TRUNCATED)
0.921895
[0.000019830256860586815,0.000024551109163439833,0.000039308826671913266,0.00006173622386995703,0.00(...TRUNCATED)
effects
synth_b0_i1
1
44,100
[ "parametric_eq", "compressor", "distortion" ]
["parametric_eq.low_shelf_gain_db","parametric_eq.low_shelf_cutoff_freq","parametric_eq.low_shelf_q_(...TRUNCATED)
[0.283719003200531,240.58810424804688,0.5153276920318604,10.767586708068848,324.7325744628906,1.5006(...TRUNCATED)
1
[0.00007931131403893232,0.00011466943396953866,0.00014884379925206304,0.00018744435510598123,0.00022(...TRUNCATED)
effects
synth_b0_i1
2
44,100
[ "parametric_eq", "compressor", "distortion" ]
["parametric_eq.low_shelf_gain_db","parametric_eq.low_shelf_cutoff_freq","parametric_eq.low_shelf_q_(...TRUNCATED)
[-5.721308708190918,116.71331787109375,1.0513806343078613,-9.794017791748047,440.04022216796875,2.29(...TRUNCATED)
0.626897
[5.584262908087112e-6,0.00001172225074697053,0.000016094587408588268,0.000021872130673727952,0.00002(...TRUNCATED)
effects
synth_b0_i1
3
44,100
[ "parametric_eq", "compressor", "distortion" ]
["parametric_eq.low_shelf_gain_db","parametric_eq.low_shelf_cutoff_freq","parametric_eq.low_shelf_q_(...TRUNCATED)
[-9.944419860839844,104.9939956665039,1.0410195589065552,1.9718888998031616,237.65145874023438,1.526(...TRUNCATED)
0.999984
[0.00005883426274522208,0.00008223962504416704,0.00008828516729408875,0.00009780942491488531,0.00010(...TRUNCATED)
effects
synth_b0_i2
0
44,100
[ "parametric_eq", "compressor", "distortion" ]
["parametric_eq.low_shelf_gain_db","parametric_eq.low_shelf_cutoff_freq","parametric_eq.low_shelf_q_(...TRUNCATED)
[3.2870805263519287,111.25947570800781,0.43277880549430847,-11.603336334228516,525.30810546875,2.773(...TRUNCATED)
0.999934
[-0.4950599670410156,-0.8308110237121582,-0.918022871017456,-0.9160065054893494,-0.8561067581176758,(...TRUNCATED)
effects
synth_b0_i2
1
44,100
[ "parametric_eq", "compressor", "distortion" ]
["parametric_eq.low_shelf_gain_db","parametric_eq.low_shelf_cutoff_freq","parametric_eq.low_shelf_q_(...TRUNCATED)
[0.283719003200531,240.58810424804688,0.5153276920318604,10.767586708068848,324.7325744628906,1.5006(...TRUNCATED)
1
[-0.8978560566902161,-0.8545442223548889,-0.9115028381347656,-0.9569015502929688,-0.9783453345298767(...TRUNCATED)
End of preview. Expand in Data Studio

StemFlipper — synth & effects parameter-estimation dataset (scaffold)

A synthetic (audio → parameters) dataset for the inverse problems StemFlipper targets: recover a synth patch from its sound, and recover an effect chain from wet audio. No public dataset pairs real audio with the synth-patch / effect-chain parameters that produced it — this fills that gap with deterministic synthetic generation. The moat is the generator + seeds, not stored audio: every example here regenerates bit-for-bit from its seed.

Configs

  • synth — torchsynth Voice renders (audio, params); params is the 78-dim normalized parameter vector (adsr_1.attack, vco_1.tuning, …).
  • effects — a known dasp-pytorch chain (parametric_eq, compressor, distortion) applied to clean synth voices → (wet audio, params); param_names labels each value.
from datasets import load_dataset
synth = load_dataset("nakas/stemflipper-dataset", "synth", split="train")
fx    = load_dataset("nakas/stemflipper-dataset", "effects", split="train")

Reproduce / extend from seeds

The published splits are a small demo. Regenerate or scale up deterministically:

from dataset.synth_gen import SynthGenConfig, iter_examples
list(iter_examples(SynthGenConfig(batch_indices=[0, 1, 2])))  # 3 batches

Generation spec (seeds):

{
  "sample_rate": 44100,
  "synth": {
    "generator": "dataset/synth_gen.py",
    "batch_size": 32,
    "batch_indices": [
      0,
      1
    ],
    "param_names_count": 78
  },
  "effects": {
    "generator": "dataset/effects_gen.py",
    "chain": [
      "parametric_eq",
      "compressor",
      "distortion"
    ],
    "seeds": [
      0,
      1,
      2,
      3
    ],
    "n_dry": 8,
    "param_names": [
      "parametric_eq.low_shelf_gain_db",
      "parametric_eq.low_shelf_cutoff_freq",
      "parametric_eq.low_shelf_q_factor",
      "parametric_eq.band0_gain_db",
      "parametric_eq.band0_cutoff_freq",
      "parametric_eq.band0_q_factor",
      "parametric_eq.band1_gain_db",
      "parametric_eq.band1_cutoff_freq",
      "parametric_eq.band1_q_factor",
      "parametric_eq.band2_gain_db",
      "parametric_eq.band2_cutoff_freq",
      "parametric_eq.band2_q_factor",
      "parametric_eq.band3_gain_db",
      "parametric_eq.band3_cutoff_freq",
      "parametric_eq.band3_q_factor",
      "parametric_eq.high_shelf_gain_db",
      "parametric_eq.high_shelf_cutoff_freq",
      "parametric_eq.high_shelf_q_factor",
      "compressor.threshold_db",
      "compressor.ratio",
      "compressor.attack_ms",
      "compressor.release_ms",
      "compressor.knee_db",
      "compressor.makeup_gain_db",
      "distortion.drive_db"
    ]
  }
}

Generated by dataset/build.py in the StemFlipper repo.
Downloads last month
34
Number of rows:
96
Total file size:
17 MB