CCMusic database collects pop music, folk music and the sound materials of national musical instruments, and makes comprehensive annotation to form a multi-purpose music database for MIR researchers. This database is specially recorded by the conservatory of music, the recordists have high music literacy, professional recording environment and technology, high recording quality, no commodity copyright problems, and the recorded audio is free and public, which is convenient for large-scale promotion. This database professionally limits the recording environment, recording equipment, recording personnel and processes, so as to avoid variety noises interference and obtain high-quality audio materials. In addition, it is of great significance to the research of music information retrieval to record the melody part and accompaniment part independently. In the future, we will collect more music materials for recording and detailed annotation.
This database contains 10 sub-datasets, which are listed as follows:
For more detailed description of this database, please see the following papers:
This dataset contains 12 gamut audio files (.wav / .mp3 / .m4a format) and 1320 split single-tone audio files (.wav / .mp3 / .m4a format) of 7 types of pianos (Kawai upright piano, Kawai grand piano, YOUNG CHANG upright piano, HSINGHAI upright piano, Steinway grand piano in grand theatre, Steinway grand piano and Pearl River upright piano) in the piano-room of China Conservatory of Music, a total of 1332 files. In addition, there is a questionnaire on subjective evaluation of piano sound quality (.xls format), including the score of 29 people participating in the subjective evaluation of piano sound quality.
Take Kawai grand piano as an example, the list is as follows:
| Serial Number |
File Name |
Performance Content |
File Size |
Duration （min） |
File Format |
Demo |
|---|---|---|---|---|---|---|
| 1 | KAWAI-Grand.wav | KAWAI grand piano chromatic scale from C1 - C2 |
20.9 MB (21,965,182 bytes) |
01:23 | .wav（RIFF） | |
| 2 | 7100.wav | KAWAI grand piano C1 single-tone | 994 KB (1,017,928 bytes) |
00:05 | .wav（RIFF） | |
| 3 | 7101.wav | KAWAI grand piano #C1 single-tone | 1.15 MB (1,211,800 bytes) |
00:06 | .wav（RIFF） | |
| 4 | 7102.wav | KAWAI grand piano D1 single-tone | 1.13 MB (1,195,640 bytes) |
00:06 | .wav（RIFF） | |
| 5 | 7103.wav | KAWAI grand piano #D1 single-tone | 1.17 MB (1,227,960 bytes) |
00:06 | .wav（RIFF） | |
| 6 | 7104.wav | KAWAI grand piano E1 single-tone | 1.06 MB (1,114,864 bytes) |
00:06 | .wav（RIFF） | |
| 7 | 7105.wav | KAWAI grand piano F1 single-tone | 1.12 MB (1,179,488 bytes) |
00:06 | .wav（RIFF） | |
| 8 | 7106.wav | KAWAI grand piano #F1 single-tone | 1.20 MB (1,260,264 bytes) |
00:07 | .wav（RIFF） | |
| 9 | 7107.wav | KAWAI grand piano G1 single-tone | 1.00 MB (1,050,244 bytes) |
00:05 | .wav（RIFF） | |
| 10 | 7108.wav | KAWAI grand piano #G1 single-tone | 1.09 MB (1,147,180 bytes) |
00:06 | .wav（RIFF） | |
| 11 | 7109.wav | KAWAI grand piano A1 single-tone | 899 KB (920,996 bytes) |
00:05 | .wav（RIFF） | |
| 12 | 7110.wav | KAWAI grand piano #A1 single-tone | 1.06 MB (1,114,868 bytes) |
00:06 | .wav（RIFF） | |
| 13 | 7111.wav | KAWAI grand piano B1 single-tone | 946 KB (969,468 bytes) |
00:05 | .wav（RIFF） | |
| 14 | 7200.wav | KAWAI grand piano C2 single-tone | 946 KB (969,464 bytes) |
00:05 | .wav（RIFF） |
This dataset contains 6 Mandarin songs covered by 22 singers, with a total of 132 sections (.wav format). Each cover consists of a verse and a chorus. Four professional judges will evaluate and score from nine aspects: intonation, rhythm, range, timbre, pronunciation, vibrato, dynamic range, breath control and overall performance, with a full score of 10 points. The final score is recorded in the scoring results of the questionnaire.
Choose a singer from each of the six songs. The list is as follows:
| Serial Number |
File Name |
Singer | Singing Content |
File Size |
Duration （min） |
File Format |
Demo |
|---|---|---|---|---|---|---|---|
| 1 | Heyan_At least I have you_demo.wav | He Yan | Song 'At least I have you' | 16.8 MB (17,683,338 bytes) |
00:30 | .wav（RIFF） | |
| 2 | Li Jingqi_Dan Yuan Ren Chang Jiu_demo.wav | Li Jingqi | Song 'Dan Yuan Ren Chang Jiu' | 12.3 MB (12,960,102 bytes) |
00:22 | .wav（RIFF） | |
| 3 | Shi Chunxue_I Only Care about You_demo.wav | Shi Chunxue | Song 'I Only Care about You' | 7.47 MB (7,833,702 bytes) |
00:13 | .wav（RIFF） | |
| 4 | Wang Haolin_Without You_demo.wav | Wang Haolin | Song 'Without You' | 16.7 MB (17,568,102 bytes) |
00:30 | .wav（RIFF） | |
| 5 | Wang Zhongbo_The Moon Represents My Heart_demo.wav | Wang Zhongbo | Song 'The Moon Represents My Heart' | 16.7 MB (17,568,102 bytes) |
00:24 | .wav（RIFF） | |
| 6 | Yang Hongan_Tian Mi Mi_demo.wav | Yang Hongan | Song 'Tian Mi Mi' | 16.7 MB (17,568,102 bytes) |
00:19 | .wav（RIFF） |
This dataset contains 1280 single-tone audio files (.wav format) sung with chest voice or falsetto. Chest voice is labeled _chest, falsetto is labeled _falsetto. In addition, the labels, Mel spectrogram, MFCC, and spectral features of each audio segment are included in a total of 5120 .csv files.
Examples are as follows:
This dataset is used for the subjective timbre evaluation experiment of 37 national musical instruments, including 1 summary audio material (.wav format) used for subjective timbre evaluation experiment, and the scoring table (.xlsx format) of the subjective timbre evaluation experiment of 37 musical instruments on 16 timbre evaluation words by 14 participants. In addition, there are 10 spectrum analysis reports of 10 musical instruments (.docx format), and the instruments' audio comes from Chinese Traditional Instrument Sound Database（CTIS） 。
Partial documents are listed as follows:
This dataset contains at least 1700 audio recordings from different genres (.mp3 format, from Netease Cloud), and each audio is about 270 ~ 300 seconds long. The database is divided into 17 genres, and each genre corresponds to a label file. The annotation information is the genre classification label, which is used for the genre classification task. Main genre labels: classical (symphony, opera, solo, chamber), non-classical (pop, dance & house, indie, soul / R&B, rock).
Format of annotation information: file_name, duration, singer,
fst_level_label, sec_level_label,
thr_level_label
Take the Adult Alternative Rock genre (labeled 19) of Rock genre (labeled
11) of non-classical genre (labeled 2) as an example, part of the list
is as follows:
| Serial Number |
file_name | duration | singer | fst_level_label | sec_level_label | thr_level_label |
|---|---|---|---|---|---|---|
| 1 | A Fine Frenzy - Elements | 203s | A Fine Frenzy | 2 | 11 | 19 |
| 2 | Daniel Powter - Not Coming Back | 241s | Daniel Powter | 2 | 11 | 19 |
| 3 | Hit Crew Masters - A Place For My Head | 186s | Hit Crew Masters | 2 | 11 | 19 |
| 4 | R.E.M. - Everybody Hurts | 320s | R.E.M. | 2 | 11 | 19 |
| 5 | Black Strobe - Boogie in zero Gravity | 209s | Black Strobe | 2 | 11 | 19 |
| 6 | Hit Crew Masters - Futures | 237s | Hit Crew Masters | 2 | 11 | 19 |
This dataset contains two sub-databases: timbre database and range database.
1.The timbre dataset contains 775 recorded acapella of 9 singers, as well
as audio clips (.wav format).
2.The range dataset includes the up and down chromatic sclaes audio of several
vocals,
as well as the cut single-tone audio materials. In addition, there are several
audio waveform files.
The timbre dataset takes singer 2's singing as an example, and the list is as follows:
| Serial Number |
File Name |
Content | File Size |
Duration | File Format |
Demo |
|---|---|---|---|---|---|---|
| 1 | singer2.wav | Singer 2 singing acapella clip (6s) | 1.05 MB (1,109,742 bytes) | 6s | .wav（RIFF） | |
| 2 | singer2-1.wav | Singer 2 singing acapella clip after cutting and collage | 2.06 MB (2,162,706 bytes) | 24s | .wav（RIFF） | |
| 3 | singer2-1-1.wav | Singer 2 singing acapella clip was cut into 10 pieces-No.1 | 2.52 MB (2,646,042 bytes) | 29s | .wav（RIFF） | |
| 4 | singer2-1-2.wav | Singer 2 singing acapella clip was cut into 10 pieces-No.2 | 2.52 MB (2,646,042 bytes) | 29s | .wav（RIFF） | |
| 5 | singer2-1-3.wav | Singer 2 singing acapella clip was cut into 10 pieces-No.3 | 2.52 MB (2,646,042 bytes) | 29s | .wav（RIFF） | |
| 6 | singer2-1-4.wav | Singer 2 singing acapella clip was cut into 10 pieces-No.4 | 2.52 MB (2,646,042 bytes) | 29s | .wav（RIFF） | |
| 7 | singer2-1-5.wav | Singer 2 singing acapella clip was cut into 10 pieces-No.5 | 2.52 MB (2,646,042 bytes) | 29s | .wav（RIFF） | |
| 8 | singer2-1-6.wav | Singer 2 singing acapella clip was cut into 10 pieces-No.6 | 2.52 MB (2,646,042 bytes) | 29s | .wav（RIFF） | |
| 9 | singer2-1-7.wav | Singer 2 singing acapella clip was cut into 10 pieces-No.7 | 2.52 MB (2,646,042 bytes) | 29s | .wav（RIFF） | |
| 10 | singer2-1-8.wav | Singer 2 singing acapella clip was cut into 10 pieces-No.8 | 2.52 MB (2,646,042 bytes) | 29s | .wav（RIFF） | |
| 11 | singer2-1-9.wav | Singer 2 singing acapella clip was cut into 10 pieces-No.9 | 2.52 MB (2,646,042 bytes) | 29s | .wav（RIFF） | |
| 12 | singer2-1-10.wav | Singer 2 singing acapella clip was cut into 10 pieces-No.10 | 2.52 MB (2,646,042 bytes) | 29s | .wav（RIFF） |
The range dataset takes singer 19's singing as an example, and the list is as follows:
This dataset contains 300 pop songs (.mp3 format, from Netease Cloud), and the structure annotation file (.txt format) of each song. Song structure: intro, chorus, verse, pre chorus, post chorus, bridge, ending.
Take "Britney Spears - Toxic (Bloodshy & Avant's Intoxicated Remix)" and "Backstreet Boys - Darlin" as examples, the list of annotation information is as follows:
| Serial Number |
Start time（0.01s） |
End time（0.01s） |
Structure Annotation |
Demo | Serial Number |
Start time（0.01s） |
End time（0.01s） |
Structure Annotation |
Demo | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | 0000 | 4241 | "Intro" | Britney
Spears - Toxic |
1 | 0000 | 2486 | "Intro" | Backstreet
Boys - Darlin' |
||||||||
| 2 | 4241 | 6924 | "Verse A" | 2 | 2486 | 4054 | "Verse A" | ||||||||||
| 3 | 6924 | 8606 | "Pre-chorus A" | 3 | 4054 | 5628 | "Verse B" | ||||||||||
| 4 | 8606 | 11289 | "Chorus A" | 4 | 5628 | 8778 | "Chorus A" | ||||||||||
| 5 | 11289 | 12631 | "Re-intro A" | 5 | 8778 | 10350 | "Verse C" | ||||||||||
| 6 | 12631 | 13977 | "Verse B" | 6 | 10350 | 11920 | "Verse D" | ||||||||||
| 7 | 13977 | 15655 | "Pre-chorus B" | 7 | 11920 | 15072 | "Chorus B" | ||||||||||
| 8 | 15655 | 19681 | "Chorus B" | 8 | 15072 | 18607 | "Bridge" | ||||||||||
| 9 | 19681 | 24043 | "Re-intro B" | 9 | 18607 | 21763 | "Chorus C" | ||||||||||
| 10 | 24043 | 26730 | "Chorus C" | 10 | 21763 | 23334 | "Re-intro" | ||||||||||
| 11 | 26730 | 28072 | "Bridge A" | 11 | 23334 | 26861 | "Chorus D" | ||||||||||
| 12 | 28072 | 29417 | "Re-intro C" | 12 | 26861 | 30015 | "Chorus E" | ||||||||||
| 13 | 29417 | 33443 | "Chorus D" | 13 | 30015 | 32758 | "Chorus F" | ||||||||||
This dataset contains 1500erhu audio clips (.wav format), all of which are played by professional erhu players. According to the different performance techniques of erhu, they are divided into 11 categories (detache, diangong, harmonic, legato&slide&glissando, percussive, pizzicato, ricochet, staccato, tremolo, trill, vibrato). Each performance technique has a corresponding number of audio. Audio from: Chinese Traditional Instrument Sound Database（CTIS） .
Some audio lists are as follows:
| Serial Number |
File Name |
Performance Technique |
File Size |
File Format |
Demo |
|---|---|---|---|---|---|
| 1 | detache_01.wav | detache | 256 KB (262,372 bytes) | .wav | |
| 2 | diangong_01.wav | diangong | 114 KB (116,940 bytes) | .wav | |
| 3 | harmonic_natural_05.wav | harmonic-natural | 215 KB (220,874 bytes) | .wav | |
| 4 | harmonic_artificial_02.wav | harmonic-artificial | 153 KB (157,008 bytes) | .wav | |
| 5 | glissando_down_05.wav | glissando-glissando_down | 44.0 KB (45,064 bytes) | .wav | |
| 6 | glissando_up_03.wav | glissando-glissando_up | 39.5 KB (40,464 bytes) | .wav | |
| 7 | huihuayin_long_04.wav | slide-huihuayin_long | 178 KB (183,248 bytes) | .wav | |
| 8 | legato&slide_up_01.wav | legato&slide_up | 183 KB (188,206 bytes) | .wav | |
| 9 | slide_dianzhi_03.wav | slide-slide_dianzhi | 78.7 KB (80,626 bytes) | .wav | |
| 10 | dajigong_05.wav | percussive-dajigong | 188 KB (192,646 bytes) | .wav | |
| 11 | horse_03.wav | percussive-horse | 168 KB (172,920 bytes) | .wav | |
| 12 | pizzicato_07.wav | pizzicato | 25.1 KB (25,704 bytes) | .wav | |
| 13 | ricochet_11.wav | ricochet | 64.6 KB (66,246 bytes) | .wav | |
| 14 | staccato_07.wav | staccato | 31.0 KB (31,812 bytes) | .wav | |
| 15 | tremolo_03.wav | tremolo | 124 KB (127,082 bytes) | .wav | |
| 16 | trill_long_01.wav | trill-trill_long | 205 KB (210,490 bytes) | .wav | |
| 17 | vibrato_late_01.wav | vibrato | 236 KB (242,574 bytes) | .wav |
This dataset is specially used to distinguish Bel Canto and National singing. All audio are sung by professional singers.
Some audio lists are as follows:
| Serial Number |
File Name |
Gender | Singing Method |
File Size |
File Format |
Demo |
|---|---|---|---|---|---|---|
| 1 | Gan Niu Shan_female_National.wav | female | National Singing | 1.23 MB (1,291,292 bytes) | .wav | |
| 2 | Grassland Pastoral_male_National.wav | male | National Singing | 9.63 MB (10,105,288 bytes) | .wav | |
| 3 | Beautiful Countryside_female_Bel Canto.wav | female | Bel Canto | 10.7 MB (11,325,040 bytes) | .wav | |
| 4 | Huang He Song_male_Bel Canto.wav | male | Bel Canto | 3.27 MB (3,433,670 bytes) | .wav |
This dataset contains 2824 audio clips of guzheng playing techniques. Among them, 2328 pieces were collected from virtual sound banks, and 496 pieces were played and recorded by a professional guzheng performer. These clips cover almost all the tones in the range of guzheng and the most commonly used playing techniques in guzheng performance. According to the different playing techniques of guzheng, the clips are divided into 8 categories: Vibrato(chanyin), Upward Portamento(shanghuayin), Downward Portamento(xiahuayin), Returning Portamento(huihuayin), Glissando (guazou, huazhi), Tremolo(yaozhi), Harmonic(fanyin), Plucks(gou,da,mo,tuo…).
Some data lists are as follows: