SINES
Tools: Pitch and Beat Analysis Tool (Polyphonic & Batch Audio Signal Analysis with
Essentia.js)
Upload Audio(s)
1. Click on "Start" to reset all values on default.
2. Upload an mp3 or wav audio file of your choice (or multiple for Batch mode).
3. Select Audio Type:
4. optional: If you have uploaded multiple audio files to analyze them in a batch:
Additionally, export a table for each file containing all individual values over time as a .zip file
5. Click on "Analyse" to start the analyzing process.
6. optional: View data in the tables and plots (Single file mode).
7. Click on "Show JS Arrays" or "Export CSV",
to show the recorded data as JavaScript arrays or to export them as a CSV/Excel file.
Select one or more audio files (wav, mp3) and click "Analyse".
Global Features
Audio Feature
Wertebereich
Bedeutung
Duration [s]
≥ 0
Die tatsächliche Länge der Datei.
Effective Duration [s]
≥ 0
Die "effektive" Dauer unter Ausschluss von Stille.
Attack Time [s]
≥ 0
Die Zeit, die ein Klang benötigt, um seine maximale Amplitude zu erreichen.
Loudness [sone]
≥ 0
Psychoakustische Lautheit. Sone ist ein lineares Maß für die wahrgenommene Lautstärke.
Loudness Vickers [dB]
≤ 0
Lautheitsmessung basierend auf dem Vickers-Modell; oft genutzt für die Normalisierung von Audiomaterial.
Leq[dB]
≤ 0
Äquivalenter Dauerschallpegel; ein Maß für die durchschnittliche Energie über die Zeit.
LARM [dB]
≤ 0
Long-term Average RMS Measurement, spezifisches Maß für die Lautheit, das oft in der automatisierten Audio-Analyse verwendet wird.
Global RMS
0 bis 1
Der quadratische Mittelwert (Root Mean Square) der Amplitude über die gesamte Dateilänge; ein Maß für die durchschnittliche Lautstärke.
Replay Gain
-20 bis +20 [dB]
Der Korrekturwert in Dezibel, der nötig ist, um die Datei auf eine Standard-Referenzlautstärke zu bringen.
Dynamic Complexity
≤ 0
Gibt an, wie stark die Dynamik (Lautstärkeunterschiede) innerhalb des Stücks variiert.
Danceability
0 bis 3
Ein Maß dafür, wie geeignet ein Track zum Tanzen ist (basierend auf Rhythmusstabilität) [the higher, the more danceable].
Intensity
-1 bis +1
Beschreibt die aggressive Intensität des Stücks [-1 = relaxed, 0 = moderate, +1 = aggressive].
Key
Variabel
Die ermittelte Haupttonart des Musikstücks (z. B. C, G, F#).
Scale
Dur/Moll
Das Tongeschlecht der ermittelten Haupttonart.
Key [estimated Key]
Variabel
Die vom TonalExtractor geschätzte globale Tonart des Titels.
Scale [Scale of estimated Key]
Dur/Moll
Das Tongeschlecht der vom TonalExtractor geschätzten Tonart.
Key Strength
0 bis 1
Die Deutlichkeit bzw. Klarheit der Tonartbestimmung (~0 = atonal/unklar, >0.6 = sehr tonal ausgeprägt).
Chords Key
Variabel
Die Grundtonart des am häufigsten vorkommenden Akkords in der Akkordfolge.
Chords Scale
Dur/Moll
Das Tongeschlecht (Dur/Moll) des am häufigsten vorkommenden Akkords.
Chords Change Rate
0 bis 1
Die relative Häufigkeit der Akkordwechsel im Verhältnis zur Gesamtzahl der analysierten Frames.
Chords Number Rate
0 bis 1
Das Verhältnis der Anzahl unterschiedlicher Akkorde zur Gesamtzahl aller erkannten Akkorde.
Tuning Frequency
> 0
Die geschätzte Referenzfrequenz des Kammertons A in Hz (Standard: 440 Hz).
Ascending/Falling Pitch
≥ 0
Verhältnis der Tonhöhenenergie vor und nach dem Maximum (0-1: tendenziell ansteigend, >1: tendenziell abfallend).
Tempo
≥ 0
Das geschätzte Gesamttempo des Musikstücks in Beats per Minute (BPM), berechnet über Rhythmus-Deskriptoren.
Tempo (Percival)
≥ 0
Ein alternativer, robuster Schätzwert für das Tempo (in BPM), basierend auf dem Percival-Modell.
Frame Features
Audio Feature
Wertebereich
Bedeutung
RMS
0 bis 1
Durchschnittliche Amplitude des aktuellen Zeitfensters (Lautstärkeverlauf).
Pitch (Essentia)
≥ 0
Die Grundfrequenz (f0) in Hz, extrahiert über die standardmäßige Spektralanalyse von Essentia.
Pitch Salience
0 bis 1
Die Dominanz oder Klarheit der Tonhöhe im Spektrum (0 = Rauschen/perkussiv, 1 = rein tonaler Peak).
Pitch (Melodia)
≥ 0
Die Tonhöhe (f0 in Hz) der führenden Melodielinie, optimiert für die Extraktion dominanter Melodien in polyphonen Werken.
Pitch (Yin)
Variabel
Die Grundfrequenz in Hz, berechnet nach der präzisen YIN-Autokorrelationsmethode im Zeitbereich.
Pitch MIDI Notes
0-127
Die in ganzzahlige Standard-MIDI-Notenwerte (0-127) umgerechnete und quantisierte Tonhöhe.
Pitch Names
Variabel
Die musikalische Notenbezeichnung (z. B. A4, C#3), die der erkannten MIDI-Note entspricht.
Vibrato (Hz)
≥ 0
Die Modulationsfrequenz eines erkannten Vibratos (Anzahl der Modulationszyklen pro Sekunde).
Vibrato Depth (Cents)
≥ 0
Die Tiefe bzw. Ausprägung des Vibratos, gemessen in Cent (relative Abweichung von der Basis-Tonhöhe).
MultiPitch (Klapuri)
Array von Frequenzen (≥ 0)
Polyphone Tonhöhenerkennung nach dem Klapuri-Modell zur gleichzeitigen Erfassung mehrerer Grundtöne (Akkorde).
Global Features
Audio Feature
Value Range
Meaning
Duration [s]
≥ 0
The actual length of the file.
Effective Duration [s]
≥ 0
The "effective" duration excluding silence.
Attack Time [s]
≥ 0
The time required for a sound to reach its maximum amplitude.
Loudness [sone]
≥ 0
Psychoacoustic loudness. Sone is a linear measure of perceived loudness.
Loudness Vickers [dB]
≤ 0
Loudness measurement based on the Vickers model; often used for audio material normalization.
Leq[dB]
≤ 0
Equivalent continuous sound level; a measure of average energy over time.
LARM [dB]
≤ 0
Long-term Average RMS Measurement, a specific loudness metric often used in automated audio analysis.
Global RMS
0 to 1
The root mean square (RMS) of the amplitude over the entire file length; a measure of average loudness.
Replay Gain
-20 to +20 [dB]
The correction value in decibels needed to bring the file to a standard reference loudness.
Dynamic Complexity
≤ 0
Indicates how much the dynamics (loudness differences) vary within the piece.
Danceability
0 to 3
A measure of how suitable a track is for dancing (based on rhythm stability) [the higher, the more danceable].
Intensity
-1 to +1
Describes the aggressive intensity of the piece [-1 = relaxed, 0 = moderate, +1 = aggressive].
Key
Variable
The detected main key of the musical piece (e.g., C, G, F#).
Scale
Major/Minor
The mode (major/minor) of the detected main key.
Key [estimated Key]
Variable
The global key of the track estimated by the TonalExtractor.
Scale [Scale of estimated Key]
Major/Minor
The scale mode (major/minor) of the key estimated by the TonalExtractor.
Key Strength
0 to 1
The clarity or strength of the key determination (~0 = atonal/unclear, >0.6 = highly tonal).
Chords Key
Variable
The root key of the most frequently occurring chord in the chord progression.
Chords Scale
Major/Minor
The scale mode (major/minor) of the most frequently occurring chord.
Chords Change Rate
0 to 1
The relative frequency of chord changes in relation to the total number of analyzed frames.
Chords Number Rate
0 to 1
The ratio of the number of unique chords to the total number of all detected chords.
Tuning Frequency
> 0
The estimated reference frequency of concert pitch A in Hz (Standard: 440 Hz).
Ascending/Falling Pitch
≥ 0
Ratio of pitch energy after and before the maximum (0-1: tending to ascend, >1: tending to fall).
Tempo
≥ 0
The estimated overall tempo of the musical piece in beats per minute (BPM), calculated via rhythm descriptors.
Tempo (Percival)
≥ 0
An alternative, robust estimate for the tempo (in BPM), based on the Percival model.
Frame Features
Audio Feature
Value Range
Meaning
RMS
0 to 1
Average amplitude of the current time window (loudness curve).
Pitch (Essentia)
≥ 0
The fundamental frequency (f0) in Hz, extracted via Essentia's standard spectral analysis.
Pitch Salience
0 to 1
The dominance or clarity of the pitch in the spectrum (0 = noise/percussive, 1 = purely tonal peak).
Pitch (Melodia)
≥ 0
The pitch (f0 in Hz) of the leading melody line, optimized for extracting dominant melodies in polyphonic works.
Pitch (Yin)
Variable
The fundamental frequency in Hz, calculated using the precise YIN autocorrelation method in the time domain.
Pitch MIDI Notes
0-127
The pitch converted and quantized into standard integer MIDI note values (0-127).
Pitch Names
Variable
The musical note name (e.g., A4, C#3) corresponding to the detected MIDI note.
Vibrato (Hz)
≥ 0
The modulation frequency of a detected vibrato (number of modulation cycles per second).
Vibrato Depth (Cents)
≥ 0
The depth or extent of the vibrato, measured in cents (relative deviation from the base pitch).
MultiPitch (Klapuri)
Array of frequencies (≥ 0)
Polyphonic pitch detection according to the Klapuri model for the simultaneous detection of multiple fundamental tones (chords).
全局特征
音频特征
数值范围
含义
Duration [s]
≥ 0
文件的实际长度。
Effective Duration [s]
≥ 0
“有效”持续时间不包括沉默。
Attack Time [s]
≥ 0
声音达到最大振幅所需的时间。
Loudness [sone]
≥ 0
心理声学响度。Sone是感知响度的线性度量。
Loudness Vickers [dB]
≤ 0
基于维氏模型的响度测量；常用于音频素材标准化。
Leq[dB]
≤ 0
等效连续声级；一段时间内平均能量的度量。
LARM [dB]
≤ 0
长期平均 RMS 测量，一种常用于自动音频分析的特定响度指标。
Global RMS
0~1
整个文件长度上幅度的均方根 (RMS)；平均响度的度量。
Replay Gain
-20~+20 [dB]
使文件达到标准参考响度所需的修正值（以dB为单位）。
Dynamic Complexity
≤ 0
指示乐曲内动态（响度差异）的变化程度。
Danceability
0~3
衡量曲目是否适合跳舞的指标（基于节奏稳定性）[越高，越适合跳舞]。
Intensity
-1~+1
描述乐曲的侵略性强度 [-1 = 放松，0 = 中等，+1 = 侵略性]。
Key
可变
检测到的音乐作品的主调（例如 C、G、F#）。
Scale
主修/辅修
检测到的主键的模式（主要/次要）。
Key [estimated Key]
可变
由 TonalExtractor 估计的轨道的全局调。
Scale [Scale of estimated Key]
主修/辅修
TonalExtractor 估计的调的音阶模式（大调/小调）。
Key Strength
0~1
调性确定的清晰度或强度（~0 = 无调性/不清楚，>0.6 = 高调性）。
Chords Key
可变
和弦进行中最常出现的和弦的根音。
Chords Scale
主修/辅修
最常出现的和弦的音阶模式（大调/小调）。
Chords Change Rate
0~1
和弦变化的相对频率与分析帧的总数相关。
Chords Number Rate
0~1
唯一和弦的数量与所有检测到的和弦总数的比率。
Tuning Frequency
> 0
音乐会音高 A 的估计参考频率，以 Hz 为单位（标准：440 Hz）。
Ascending/Falling Pitch
≥ 0
最大值前后的俯仰能量比（0-1：趋于上升，>1：趋于下降）。
Tempo
≥ 0
通过节奏描述符计算得出的音乐作品的估计整体节奏（以每分钟节拍数 (BPM) 为单位）。
Tempo (Percival)
≥ 0
基于 Percival 模型的另一种稳健的节奏估计（以 BPM 为单位）。
帧特征
音频特征
数值范围
含义
RMS
0~1
当前时间窗口的平均幅度（响度曲线）。
Pitch (Essentia)
≥ 0
基频 (f0)（以 Hz 为单位），通过 Essentia 的标准频谱分析提取。
Pitch Salience
0~1
频谱中音高的主导度或清晰度（0 = 噪声/敲击声，1 = 纯音调峰值）。
Pitch (Melodia)
≥ 0
主旋律线的音高（f0，以 Hz 为单位），针对提取复调作品中的主旋律进行了优化。
Pitch (Yin)
可变
基频（以 Hz 为单位），在时域中使用精确的 YIN 自相关方法计算得出。
Pitch MIDI Notes
0~127
音高转换并量化为标准整数 MIDI 音符值 (0~127)。
Pitch Names
可变
与检测到的 MIDI 音符对应的音符名称（例如 A4、C#3）。
Vibrato (Hz)
≥ 0
检测到的颤音的调制频率（每秒的调制周期数）。
Vibrato Depth (Cents)
≥ 0
颤音的深度或范围，以音分为单位（与基本音高的相对偏差）。
MultiPitch (Klapuri)
频率数组 (≥ 0)
根据 Klapuri 模型进行复调音高检测，用于同时检测多个基音（和弦）。
Analyze Audio...
0%
Analysed File:
Feature
Feature
Duration [in s]
0
Loudness [sone]
0
Effective Duration [in s]
0
Loudness Vickers [in dB]
0
Attack Time [in s]
0
Leq [equivalent Sound Level in dB]
0
Dynamic Complexity
0
LARM [Long Term Loudness in dB]
0
Danceability [0 to 3, the higher, the more danceable]
Please note: Essentia and the underlying Essentia library are licensed under the Affero GPLv3 (AGPLv3).
This library is available under the AGPLv3 for non-commercial use. If you wish to use Essentia in a commercial product, you have to contact the Music Technology Group (MTG)
at Pompeu Fabra University (UPF) directly to negotiate commercial licensing terms. Translation of the Chinese version of the website: Li Hao