File size: 10,544 Bytes
fcc23f9 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 |
DAMP (Digital Archive of Mobile Performances) Dataset Project Datasets extracted from the activities of Smule Users (Magic Piano and Sing! Karaoke). NOTE: Read the *[RESEARCH DATA AGREEMENT]* posted at the end of this README file. Summary: By downloading and using data from the DAMP dataset archives, you agree, among other things, that: * you will not redistribute the data, and * you will not use the data for commercial purposes. This document describes the sing300x30x2 dataset, assembled between November-Dec. 2017, from performances and SMule server events created/occuring between July and October, 2017. The sing300x30x2 dataset was formed by determining the most popular singers, female and male, of the 300 most popular arrangements (songs) in 30 countries. The most popular arrangements were determined by counting song starts or joins (duet/group) of recordings for each arrangement, within the country of interest. The term "arrangement" is used, because there might be multiple arrangements of the same song. The most popular performances and singers of those arrangements were determined by counting Listens (and/or Loves, if no complete Listens) for all performances of each arrangement. The sing_300x30x2 Data set: Found within the sing_300x30x2 folder, are 30 folders, one for each country. Here is the list of 30 countries and their codes: AE United Arab Emirates AR Argentina AU Australia BR Brazil CL Chile CN China DE Germany ES Spain FR France GB United Kingdom (Great Britain) HU Hungary IN India ID Indonesia IR Iran IQ Iraq IT Italy JP Japan KR South Korea MX Mexico MY Malaysia NO Norway PH Philippines PT Portugal RU Russia SA Saudi Arabia SG Singapore TH Thailand US United States of America VN Vietnam ZA South Africa Within each country folder, there are three directories, holding metadata for each of the 300 arrangements, lyric timings for each arrangement, and the 600 vocal performances (1 male, 1 female). Example: AE/AEArrangementMeta Metadata for each arrangement (song) AE/AELyrics Timed Lyrics (if available) for each arrangement AE/AEVocals Vocal tracks of top 2 performances, Male/Female Arrangement MetaData Files are named thusly: [arrangement_id].txt and contain information, where available, such as: Arrangement ID: 3771075_3771075 Arrangement title: Despacito Arrangement has lyrics: True Arrangement has PitchTrack: True Arrangement tags: 2010s, dance, flamenco, hiphop, latin, latinpop, pop, rap, reggaeton, spanish Arrangement artist: Luis Fonsi Care has been taken to preserve international UNICODE symbols were possible, for both Arrangement Meta Information and Lyrics Timing files: Arrangement ID: 386604149_188928 Arrangement title: على بالي Arrangement has lyrics: True Arrangement has PitchTrack: True Arrangement tags: arabicsongs, arab, sheren, arabic, egypt Arrangement artist: شيرين Lyrics Timing files are JSON files named thusly: [arrangement_id].json and contain either line-level or word-level timings for the lyrics (where available). Here is a short excerpt from one Lyrics Timing file: [ { "t": 14.419791666666667, "l": "You're givin' me a million reasons" }, { "t": 16.419791666666665, "l": "To let you go" }, ... ] Vocal performances are stored in each CCVocals directory, containing (roughly) 600 audio (m4a, OGG compressed) files named thusly: [arrangement_id]-[performance_id]-CC-GEND-[account_id].m4a where CC = 2 Letter Country Code and GEND = gender (M or F) Note that the delineators between segments in these filenames are dashes. Underscores occur in both arrangement and performance IDs, so dashes were selected to ensure unique automatic parsing. Here is a specific example vocal performance filename: 982160392_2914339-772360143_1567798916-AE-F-1048659059.m4a There is also a .csv file within each country folder, containing performance and user metadata for the performances/users from that country. At the top level sing_300x30x2 folder, there is a file, sing_300x30x2.csv which contains comma-separated user and performance level metadata for each performance, for all performances in the dataset. The items in each row, from left to right, are: arrangement_id unique song arrangement tag performance_id unique popular performance of this arrangement account_id of popular user singing on this performance country for this user locale for this user (includes country + language setting) latitude for this performance longitude for this performance city_id for this user gender for this user (NOTE: self reported, unverified) birth_year for this user (NOTE: self reported, unverified) creation_timestamp for this performance, UNIX integer time device_os for this performance headphones for this performance efx for this performance, audio efx selected/applied video for this performance ens_type SOLO, DUET, GROUP num_tracks for DUET/GROUP num_this_plyr for this performance, user might sing more than once listens for this performance loves for this performance rec_starts for this performance, how many times to user start record song_sing_restarts for this performance, user rewind/retry from within song song_review_restarts for this performance, same but from review screen at end total_user_perfs for this user, total number of performances of all songs app_inst_days user lifetime (actually for the sing app on this device type) followers for this user, total followers followees total other users followed by this user *************************************************************************************** *[RESEARCH DATA AGREEMENT]* SMULE RESEARCH DATA LICENSE AGREEMENT DIGITAL ARCHIVE OF MOBILE PERFORMANCES (DAMP) RESEARCHER (the downloader/user of the Smule Data) agrees as follows: 1. DEFINITIONS The following definitions apply herein: a) “Smule Data” means user data, user content, songs, lyrics or composition data uploaded, created or stored on the Platform, applications or website to be provided by Smule to RESEARCHER or otherwise accessed by RESEARCHER through the Platform, applications or website. b) “Platform” means archives of Smule Data hosted by Smule, or in other locations such as the Digital Archive of Mobile Performances (DAMP) portal hosted by the Stanford Center for Computer Research in Music and Acoustics (CCRMA) (https://ccrma.stanford.edu/damp/) c) “Use” means downloading, copying, or examining all or any portion of the Smule Data.. 2. SMULE LICENSE GRANT AND USE RESTRICTIONS a) Subject to the terms, conditions, and limitations herein, Smule grants to RESEARCHER a limited, nonexclusive, non-transferable, non-sublicensable, royalty-free license to (a) use the Smule Data in accordance with the terms and conditions of this Agreement, and (b) use the Smule Data solely for this Purpose. b) The license granted by Smule under this Agreement does not include the right to, and RESEARCHER will not: (i) use the Smule Data or Smule Software for any commercial purpose; (ii) report publicly or share with any third party any Smule Data or resulting financial information; or (iii) engage in or permit any infringing or unlawful activities involving the Smule Data; or (iv) engage with any third party, directly or indirectly, seeking to explore the Smule Data in an adversarial manner or threaten such exploration. c) As between RESEARCHER and Smule, RESEARCHER and Smule agree that Smule owns all proprietary rights, including patent, copyright, trade secret, trademark and other proprietary rights, in and to all Smule Software, Smule Data and Smule Confidential Information and any corrections, enhancements, or modifications, whether made by Smule, RESEARCHER, or any third party. d) RESEARCHER agrees that only Smule shall have the right to alter or otherwise modify the Smule Data. e) Except as authorized by this Agreement, RESEARCHER shall not sell, license, sublicense, publish, display, distribute, disclose or otherwise transfer, directly or indirectly, the Smule Data, any copy thereof, or any access thereto, in whole or in part, to any third party without Smule's prior written consent. f) RESEARCHER shall not sell, license, publish, display, distribute, disclose or otherwise make available the Smule Data or to any third party, nor use Smule Data except as explicitly authorized by this Agreement. This includes sharing any userIDs and/or passwords for accessing, viewing, or downloading the data, where applicable. 3. TECHNICAL SUPPORT RESEARCHER, at its option, may provide error reports, notes, and recommendations regarding the Smule Data to Smule. 4. GENERAL PROVISIONS This Agreement shall be governed by and construed in accordance with the laws of the State of California, excluding conflict of law rules and principles. The United Nations Convention on Contracts for the International Sale of Goods (1980) is specifically excluded and shall not apply. This Agreement is prepared, executed and will be interpreted in English only. Any dispute arising out of or connected with this Agreement shall be subject to the sole and exclusive jurisdiction of the California courts. Nothing contained in this Section shall preclude either Party from seeking or obtaining preliminary injunctive relief pending resolution of the dispute in issue. This Agreement does not create any agency or partnership relationship. Any failure to enforce any provision of this Agreement shall not constitute a waiver of any other provision. This Agreement shall be binding upon and inure to the benefit of the Parties hereto and their respective successors and permitted assigns. RESEARCHER shall not assign or otherwise transfer the Smule Software, the Smule Data, the Smule Confidential Information, or this Agreement to anyone, including any parent, subsidiary, affiliated entity or third party as part of the sale of any portion of its business or pursuant to a change of control, without Smule's prior written consent. The provisions of this Agreement shall be binding on, and inure to the benefit of, the parties, their successors and permitted assigns. If any provision is held to be illegal, invalid, or unenforceable, the remaining provisions shall remain in full force and effect. This Agreement contains the entire agreement between the parties with respect to the subject matter hereof. This Agreement may not be amended, nor any obligation waived, except by a writing signed by the parties hereto. |