├── .env ├── LICENSE ├── README.md ├── data ├── IEMP_north_indian_raga │ └── Sample │ │ ├── Annotations │ │ ├── NIR_PrB_Jhinjhoti_2Gats_Annotation_Sample.csv │ │ ├── NIR_PrB_Jhinjhoti_2Gats_Onsets_Raw_Sarod_Sample.csv │ │ ├── NIR_PrB_Jhinjhoti_2Gats_Onsets_Raw_Tabla_Sample.csv │ │ ├── NIR_PrB_Jhinjhoti_2Gats_Onsets_Selected_Rupak_Sample.csv │ │ ├── NIR_PrB_jhinjhoti_2Gats_Metre_Rupak_Sample.csv │ │ ├── NIR_PrB_jhinjhoti_2Gats_OptFlow_CentralWide_Sarod_Sample.csv │ │ ├── NIR_PrB_jhinjhoti_2Gats_OptFlow_CentralWide_Tabla_Sample.csv │ │ ├── NIR_PrB_jhinjhoti_2Gats_OptFlow_StageR_Sarod_Sample.csv │ │ └── NIR_PrB_jhinjhoti_2Gats_OptFlow_StageR_Tabla_Sample.csv │ │ └── Media │ │ ├── NIR_PrB_Jhinjhoti_2Gats_Sarod_sample.mp3 │ │ ├── NIR_PrB_Jhinjhoti_2Gats_StereoMix_sample.mp3 │ │ └── NIR_PrB_Jhinjhoti_2Gats_Tabla_sample.mp3 └── compMusicDatasets │ └── turkishMakam │ ├── annotations.json │ └── readme.txt ├── docker-compose.yml └── notebooks ├── DownloadDataFromDunya_noToken.ipynb ├── downloadAllSARAGAContent.ipynb ├── external_utilities ├── converter.py ├── pitchdistribution.py ├── predominantmelodymakam.py └── readme.txt ├── formExpSubsets4ModeRecognition.ipynb ├── generateFileLists4Collections.ipynb ├── symbolicDataPro_symbTr.ipynb ├── tuningAnalysis_SetOfRecordings.ipynb ├── tuningAnalysis_SingleRecording.ipynb └── visualizeAnnotations.ipynb /.env: -------------------------------------------------------------------------------- 1 | JUPYTER_USER_ID=1000 2 | -------------------------------------------------------------------------------- /LICENSE: -------------------------------------------------------------------------------- 1 | GNU GENERAL PUBLIC LICENSE 2 | Version 3, 29 June 2007 3 | 4 | Copyright (C) 2007 Free Software Foundation, Inc. 5 | Everyone is permitted to copy and distribute verbatim copies 6 | of this license document, but changing it is not allowed. 7 | 8 | Preamble 9 | 10 | The GNU General Public License is a free, copyleft license for 11 | software and other kinds of works. 12 | 13 | The licenses for most software and other practical works are designed 14 | to take away your freedom to share and change the works. By contrast, 15 | the GNU General Public License is intended to guarantee your freedom to 16 | share and change all versions of a program--to make sure it remains free 17 | software for all its users. We, the Free Software Foundation, use the 18 | GNU General Public License for most of our software; it applies also to 19 | any other work released this way by its authors. You can apply it to 20 | your programs, too. 21 | 22 | When we speak of free software, we are referring to freedom, not 23 | price. Our General Public Licenses are designed to make sure that you 24 | have the freedom to distribute copies of free software (and charge for 25 | them if you wish), that you receive source code or can get it if you 26 | want it, that you can change the software or use pieces of it in new 27 | free programs, and that you know you can do these things. 28 | 29 | To protect your rights, we need to prevent others from denying you 30 | these rights or asking you to surrender the rights. Therefore, you have 31 | certain responsibilities if you distribute copies of the software, or if 32 | you modify it: responsibilities to respect the freedom of others. 33 | 34 | For example, if you distribute copies of such a program, whether 35 | gratis or for a fee, you must pass on to the recipients the same 36 | freedoms that you received. You must make sure that they, too, receive 37 | or can get the source code. And you must show them these terms so they 38 | know their rights. 39 | 40 | Developers that use the GNU GPL protect your rights with two steps: 41 | (1) assert copyright on the software, and (2) offer you this License 42 | giving you legal permission to copy, distribute and/or modify it. 43 | 44 | For the developers' and authors' protection, the GPL clearly explains 45 | that there is no warranty for this free software. For both users' and 46 | authors' sake, the GPL requires that modified versions be marked as 47 | changed, so that their problems will not be attributed erroneously to 48 | authors of previous versions. 49 | 50 | Some devices are designed to deny users access to install or run 51 | modified versions of the software inside them, although the manufacturer 52 | can do so. This is fundamentally incompatible with the aim of 53 | protecting users' freedom to change the software. The systematic 54 | pattern of such abuse occurs in the area of products for individuals to 55 | use, which is precisely where it is most unacceptable. Therefore, we 56 | have designed this version of the GPL to prohibit the practice for those 57 | products. If such problems arise substantially in other domains, we 58 | stand ready to extend this provision to those domains in future versions 59 | of the GPL, as needed to protect the freedom of users. 60 | 61 | Finally, every program is threatened constantly by software patents. 62 | States should not allow patents to restrict development and use of 63 | software on general-purpose computers, but in those that do, we wish to 64 | avoid the special danger that patents applied to a free program could 65 | make it effectively proprietary. To prevent this, the GPL assures that 66 | patents cannot be used to render the program non-free. 67 | 68 | The precise terms and conditions for copying, distribution and 69 | modification follow. 70 | 71 | TERMS AND CONDITIONS 72 | 73 | 0. Definitions. 74 | 75 | "This License" refers to version 3 of the GNU General Public License. 76 | 77 | "Copyright" also means copyright-like laws that apply to other kinds of 78 | works, such as semiconductor masks. 79 | 80 | "The Program" refers to any copyrightable work licensed under this 81 | License. Each licensee is addressed as "you". "Licensees" and 82 | "recipients" may be individuals or organizations. 83 | 84 | To "modify" a work means to copy from or adapt all or part of the work 85 | in a fashion requiring copyright permission, other than the making of an 86 | exact copy. The resulting work is called a "modified version" of the 87 | earlier work or a work "based on" the earlier work. 88 | 89 | A "covered work" means either the unmodified Program or a work based 90 | on the Program. 91 | 92 | To "propagate" a work means to do anything with it that, without 93 | permission, would make you directly or secondarily liable for 94 | infringement under applicable copyright law, except executing it on a 95 | computer or modifying a private copy. Propagation includes copying, 96 | distribution (with or without modification), making available to the 97 | public, and in some countries other activities as well. 98 | 99 | To "convey" a work means any kind of propagation that enables other 100 | parties to make or receive copies. Mere interaction with a user through 101 | a computer network, with no transfer of a copy, is not conveying. 102 | 103 | An interactive user interface displays "Appropriate Legal Notices" 104 | to the extent that it includes a convenient and prominently visible 105 | feature that (1) displays an appropriate copyright notice, and (2) 106 | tells the user that there is no warranty for the work (except to the 107 | extent that warranties are provided), that licensees may convey the 108 | work under this License, and how to view a copy of this License. If 109 | the interface presents a list of user commands or options, such as a 110 | menu, a prominent item in the list meets this criterion. 111 | 112 | 1. Source Code. 113 | 114 | The "source code" for a work means the preferred form of the work 115 | for making modifications to it. "Object code" means any non-source 116 | form of a work. 117 | 118 | A "Standard Interface" means an interface that either is an official 119 | standard defined by a recognized standards body, or, in the case of 120 | interfaces specified for a particular programming language, one that 121 | is widely used among developers working in that language. 122 | 123 | The "System Libraries" of an executable work include anything, other 124 | than the work as a whole, that (a) is included in the normal form of 125 | packaging a Major Component, but which is not part of that Major 126 | Component, and (b) serves only to enable use of the work with that 127 | Major Component, or to implement a Standard Interface for which an 128 | implementation is available to the public in source code form. A 129 | "Major Component", in this context, means a major essential component 130 | (kernel, window system, and so on) of the specific operating system 131 | (if any) on which the executable work runs, or a compiler used to 132 | produce the work, or an object code interpreter used to run it. 133 | 134 | The "Corresponding Source" for a work in object code form means all 135 | the source code needed to generate, install, and (for an executable 136 | work) run the object code and to modify the work, including scripts to 137 | control those activities. However, it does not include the work's 138 | System Libraries, or general-purpose tools or generally available free 139 | programs which are used unmodified in performing those activities but 140 | which are not part of the work. For example, Corresponding Source 141 | includes interface definition files associated with source files for 142 | the work, and the source code for shared libraries and dynamically 143 | linked subprograms that the work is specifically designed to require, 144 | such as by intimate data communication or control flow between those 145 | subprograms and other parts of the work. 146 | 147 | The Corresponding Source need not include anything that users 148 | can regenerate automatically from other parts of the Corresponding 149 | Source. 150 | 151 | The Corresponding Source for a work in source code form is that 152 | same work. 153 | 154 | 2. Basic Permissions. 155 | 156 | All rights granted under this License are granted for the term of 157 | copyright on the Program, and are irrevocable provided the stated 158 | conditions are met. This License explicitly affirms your unlimited 159 | permission to run the unmodified Program. The output from running a 160 | covered work is covered by this License only if the output, given its 161 | content, constitutes a covered work. This License acknowledges your 162 | rights of fair use or other equivalent, as provided by copyright law. 163 | 164 | You may make, run and propagate covered works that you do not 165 | convey, without conditions so long as your license otherwise remains 166 | in force. You may convey covered works to others for the sole purpose 167 | of having them make modifications exclusively for you, or provide you 168 | with facilities for running those works, provided that you comply with 169 | the terms of this License in conveying all material for which you do 170 | not control copyright. Those thus making or running the covered works 171 | for you must do so exclusively on your behalf, under your direction 172 | and control, on terms that prohibit them from making any copies of 173 | your copyrighted material outside their relationship with you. 174 | 175 | Conveying under any other circumstances is permitted solely under 176 | the conditions stated below. Sublicensing is not allowed; section 10 177 | makes it unnecessary. 178 | 179 | 3. Protecting Users' Legal Rights From Anti-Circumvention Law. 180 | 181 | No covered work shall be deemed part of an effective technological 182 | measure under any applicable law fulfilling obligations under article 183 | 11 of the WIPO copyright treaty adopted on 20 December 1996, or 184 | similar laws prohibiting or restricting circumvention of such 185 | measures. 186 | 187 | When you convey a covered work, you waive any legal power to forbid 188 | circumvention of technological measures to the extent such circumvention 189 | is effected by exercising rights under this License with respect to 190 | the covered work, and you disclaim any intention to limit operation or 191 | modification of the work as a means of enforcing, against the work's 192 | users, your or third parties' legal rights to forbid circumvention of 193 | technological measures. 194 | 195 | 4. Conveying Verbatim Copies. 196 | 197 | You may convey verbatim copies of the Program's source code as you 198 | receive it, in any medium, provided that you conspicuously and 199 | appropriately publish on each copy an appropriate copyright notice; 200 | keep intact all notices stating that this License and any 201 | non-permissive terms added in accord with section 7 apply to the code; 202 | keep intact all notices of the absence of any warranty; and give all 203 | recipients a copy of this License along with the Program. 204 | 205 | You may charge any price or no price for each copy that you convey, 206 | and you may offer support or warranty protection for a fee. 207 | 208 | 5. Conveying Modified Source Versions. 209 | 210 | You may convey a work based on the Program, or the modifications to 211 | produce it from the Program, in the form of source code under the 212 | terms of section 4, provided that you also meet all of these conditions: 213 | 214 | a) The work must carry prominent notices stating that you modified 215 | it, and giving a relevant date. 216 | 217 | b) The work must carry prominent notices stating that it is 218 | released under this License and any conditions added under section 219 | 7. This requirement modifies the requirement in section 4 to 220 | "keep intact all notices". 221 | 222 | c) You must license the entire work, as a whole, under this 223 | License to anyone who comes into possession of a copy. This 224 | License will therefore apply, along with any applicable section 7 225 | additional terms, to the whole of the work, and all its parts, 226 | regardless of how they are packaged. This License gives no 227 | permission to license the work in any other way, but it does not 228 | invalidate such permission if you have separately received it. 229 | 230 | d) If the work has interactive user interfaces, each must display 231 | Appropriate Legal Notices; however, if the Program has interactive 232 | interfaces that do not display Appropriate Legal Notices, your 233 | work need not make them do so. 234 | 235 | A compilation of a covered work with other separate and independent 236 | works, which are not by their nature extensions of the covered work, 237 | and which are not combined with it such as to form a larger program, 238 | in or on a volume of a storage or distribution medium, is called an 239 | "aggregate" if the compilation and its resulting copyright are not 240 | used to limit the access or legal rights of the compilation's users 241 | beyond what the individual works permit. Inclusion of a covered work 242 | in an aggregate does not cause this License to apply to the other 243 | parts of the aggregate. 244 | 245 | 6. Conveying Non-Source Forms. 246 | 247 | You may convey a covered work in object code form under the terms 248 | of sections 4 and 5, provided that you also convey the 249 | machine-readable Corresponding Source under the terms of this License, 250 | in one of these ways: 251 | 252 | a) Convey the object code in, or embodied in, a physical product 253 | (including a physical distribution medium), accompanied by the 254 | Corresponding Source fixed on a durable physical medium 255 | customarily used for software interchange. 256 | 257 | b) Convey the object code in, or embodied in, a physical product 258 | (including a physical distribution medium), accompanied by a 259 | written offer, valid for at least three years and valid for as 260 | long as you offer spare parts or customer support for that product 261 | model, to give anyone who possesses the object code either (1) a 262 | copy of the Corresponding Source for all the software in the 263 | product that is covered by this License, on a durable physical 264 | medium customarily used for software interchange, for a price no 265 | more than your reasonable cost of physically performing this 266 | conveying of source, or (2) access to copy the 267 | Corresponding Source from a network server at no charge. 268 | 269 | c) Convey individual copies of the object code with a copy of the 270 | written offer to provide the Corresponding Source. This 271 | alternative is allowed only occasionally and noncommercially, and 272 | only if you received the object code with such an offer, in accord 273 | with subsection 6b. 274 | 275 | d) Convey the object code by offering access from a designated 276 | place (gratis or for a charge), and offer equivalent access to the 277 | Corresponding Source in the same way through the same place at no 278 | further charge. You need not require recipients to copy the 279 | Corresponding Source along with the object code. If the place to 280 | copy the object code is a network server, the Corresponding Source 281 | may be on a different server (operated by you or a third party) 282 | that supports equivalent copying facilities, provided you maintain 283 | clear directions next to the object code saying where to find the 284 | Corresponding Source. Regardless of what server hosts the 285 | Corresponding Source, you remain obligated to ensure that it is 286 | available for as long as needed to satisfy these requirements. 287 | 288 | e) Convey the object code using peer-to-peer transmission, provided 289 | you inform other peers where the object code and Corresponding 290 | Source of the work are being offered to the general public at no 291 | charge under subsection 6d. 292 | 293 | A separable portion of the object code, whose source code is excluded 294 | from the Corresponding Source as a System Library, need not be 295 | included in conveying the object code work. 296 | 297 | A "User Product" is either (1) a "consumer product", which means any 298 | tangible personal property which is normally used for personal, family, 299 | or household purposes, or (2) anything designed or sold for incorporation 300 | into a dwelling. In determining whether a product is a consumer product, 301 | doubtful cases shall be resolved in favor of coverage. For a particular 302 | product received by a particular user, "normally used" refers to a 303 | typical or common use of that class of product, regardless of the status 304 | of the particular user or of the way in which the particular user 305 | actually uses, or expects or is expected to use, the product. A product 306 | is a consumer product regardless of whether the product has substantial 307 | commercial, industrial or non-consumer uses, unless such uses represent 308 | the only significant mode of use of the product. 309 | 310 | "Installation Information" for a User Product means any methods, 311 | procedures, authorization keys, or other information required to install 312 | and execute modified versions of a covered work in that User Product from 313 | a modified version of its Corresponding Source. The information must 314 | suffice to ensure that the continued functioning of the modified object 315 | code is in no case prevented or interfered with solely because 316 | modification has been made. 317 | 318 | If you convey an object code work under this section in, or with, or 319 | specifically for use in, a User Product, and the conveying occurs as 320 | part of a transaction in which the right of possession and use of the 321 | User Product is transferred to the recipient in perpetuity or for a 322 | fixed term (regardless of how the transaction is characterized), the 323 | Corresponding Source conveyed under this section must be accompanied 324 | by the Installation Information. But this requirement does not apply 325 | if neither you nor any third party retains the ability to install 326 | modified object code on the User Product (for example, the work has 327 | been installed in ROM). 328 | 329 | The requirement to provide Installation Information does not include a 330 | requirement to continue to provide support service, warranty, or updates 331 | for a work that has been modified or installed by the recipient, or for 332 | the User Product in which it has been modified or installed. Access to a 333 | network may be denied when the modification itself materially and 334 | adversely affects the operation of the network or violates the rules and 335 | protocols for communication across the network. 336 | 337 | Corresponding Source conveyed, and Installation Information provided, 338 | in accord with this section must be in a format that is publicly 339 | documented (and with an implementation available to the public in 340 | source code form), and must require no special password or key for 341 | unpacking, reading or copying. 342 | 343 | 7. Additional Terms. 344 | 345 | "Additional permissions" are terms that supplement the terms of this 346 | License by making exceptions from one or more of its conditions. 347 | Additional permissions that are applicable to the entire Program shall 348 | be treated as though they were included in this License, to the extent 349 | that they are valid under applicable law. If additional permissions 350 | apply only to part of the Program, that part may be used separately 351 | under those permissions, but the entire Program remains governed by 352 | this License without regard to the additional permissions. 353 | 354 | When you convey a copy of a covered work, you may at your option 355 | remove any additional permissions from that copy, or from any part of 356 | it. (Additional permissions may be written to require their own 357 | removal in certain cases when you modify the work.) You may place 358 | additional permissions on material, added by you to a covered work, 359 | for which you have or can give appropriate copyright permission. 360 | 361 | Notwithstanding any other provision of this License, for material you 362 | add to a covered work, you may (if authorized by the copyright holders of 363 | that material) supplement the terms of this License with terms: 364 | 365 | a) Disclaiming warranty or limiting liability differently from the 366 | terms of sections 15 and 16 of this License; or 367 | 368 | b) Requiring preservation of specified reasonable legal notices or 369 | author attributions in that material or in the Appropriate Legal 370 | Notices displayed by works containing it; or 371 | 372 | c) Prohibiting misrepresentation of the origin of that material, or 373 | requiring that modified versions of such material be marked in 374 | reasonable ways as different from the original version; or 375 | 376 | d) Limiting the use for publicity purposes of names of licensors or 377 | authors of the material; or 378 | 379 | e) Declining to grant rights under trademark law for use of some 380 | trade names, trademarks, or service marks; or 381 | 382 | f) Requiring indemnification of licensors and authors of that 383 | material by anyone who conveys the material (or modified versions of 384 | it) with contractual assumptions of liability to the recipient, for 385 | any liability that these contractual assumptions directly impose on 386 | those licensors and authors. 387 | 388 | All other non-permissive additional terms are considered "further 389 | restrictions" within the meaning of section 10. If the Program as you 390 | received it, or any part of it, contains a notice stating that it is 391 | governed by this License along with a term that is a further 392 | restriction, you may remove that term. If a license document contains 393 | a further restriction but permits relicensing or conveying under this 394 | License, you may add to a covered work material governed by the terms 395 | of that license document, provided that the further restriction does 396 | not survive such relicensing or conveying. 397 | 398 | If you add terms to a covered work in accord with this section, you 399 | must place, in the relevant source files, a statement of the 400 | additional terms that apply to those files, or a notice indicating 401 | where to find the applicable terms. 402 | 403 | Additional terms, permissive or non-permissive, may be stated in the 404 | form of a separately written license, or stated as exceptions; 405 | the above requirements apply either way. 406 | 407 | 8. Termination. 408 | 409 | You may not propagate or modify a covered work except as expressly 410 | provided under this License. Any attempt otherwise to propagate or 411 | modify it is void, and will automatically terminate your rights under 412 | this License (including any patent licenses granted under the third 413 | paragraph of section 11). 414 | 415 | However, if you cease all violation of this License, then your 416 | license from a particular copyright holder is reinstated (a) 417 | provisionally, unless and until the copyright holder explicitly and 418 | finally terminates your license, and (b) permanently, if the copyright 419 | holder fails to notify you of the violation by some reasonable means 420 | prior to 60 days after the cessation. 421 | 422 | Moreover, your license from a particular copyright holder is 423 | reinstated permanently if the copyright holder notifies you of the 424 | violation by some reasonable means, this is the first time you have 425 | received notice of violation of this License (for any work) from that 426 | copyright holder, and you cure the violation prior to 30 days after 427 | your receipt of the notice. 428 | 429 | Termination of your rights under this section does not terminate the 430 | licenses of parties who have received copies or rights from you under 431 | this License. If your rights have been terminated and not permanently 432 | reinstated, you do not qualify to receive new licenses for the same 433 | material under section 10. 434 | 435 | 9. Acceptance Not Required for Having Copies. 436 | 437 | You are not required to accept this License in order to receive or 438 | run a copy of the Program. Ancillary propagation of a covered work 439 | occurring solely as a consequence of using peer-to-peer transmission 440 | to receive a copy likewise does not require acceptance. However, 441 | nothing other than this License grants you permission to propagate or 442 | modify any covered work. These actions infringe copyright if you do 443 | not accept this License. Therefore, by modifying or propagating a 444 | covered work, you indicate your acceptance of this License to do so. 445 | 446 | 10. Automatic Licensing of Downstream Recipients. 447 | 448 | Each time you convey a covered work, the recipient automatically 449 | receives a license from the original licensors, to run, modify and 450 | propagate that work, subject to this License. You are not responsible 451 | for enforcing compliance by third parties with this License. 452 | 453 | An "entity transaction" is a transaction transferring control of an 454 | organization, or substantially all assets of one, or subdividing an 455 | organization, or merging organizations. If propagation of a covered 456 | work results from an entity transaction, each party to that 457 | transaction who receives a copy of the work also receives whatever 458 | licenses to the work the party's predecessor in interest had or could 459 | give under the previous paragraph, plus a right to possession of the 460 | Corresponding Source of the work from the predecessor in interest, if 461 | the predecessor has it or can get it with reasonable efforts. 462 | 463 | You may not impose any further restrictions on the exercise of the 464 | rights granted or affirmed under this License. For example, you may 465 | not impose a license fee, royalty, or other charge for exercise of 466 | rights granted under this License, and you may not initiate litigation 467 | (including a cross-claim or counterclaim in a lawsuit) alleging that 468 | any patent claim is infringed by making, using, selling, offering for 469 | sale, or importing the Program or any portion of it. 470 | 471 | 11. Patents. 472 | 473 | A "contributor" is a copyright holder who authorizes use under this 474 | License of the Program or a work on which the Program is based. The 475 | work thus licensed is called the contributor's "contributor version". 476 | 477 | A contributor's "essential patent claims" are all patent claims 478 | owned or controlled by the contributor, whether already acquired or 479 | hereafter acquired, that would be infringed by some manner, permitted 480 | by this License, of making, using, or selling its contributor version, 481 | but do not include claims that would be infringed only as a 482 | consequence of further modification of the contributor version. For 483 | purposes of this definition, "control" includes the right to grant 484 | patent sublicenses in a manner consistent with the requirements of 485 | this License. 486 | 487 | Each contributor grants you a non-exclusive, worldwide, royalty-free 488 | patent license under the contributor's essential patent claims, to 489 | make, use, sell, offer for sale, import and otherwise run, modify and 490 | propagate the contents of its contributor version. 491 | 492 | In the following three paragraphs, a "patent license" is any express 493 | agreement or commitment, however denominated, not to enforce a patent 494 | (such as an express permission to practice a patent or covenant not to 495 | sue for patent infringement). To "grant" such a patent license to a 496 | party means to make such an agreement or commitment not to enforce a 497 | patent against the party. 498 | 499 | If you convey a covered work, knowingly relying on a patent license, 500 | and the Corresponding Source of the work is not available for anyone 501 | to copy, free of charge and under the terms of this License, through a 502 | publicly available network server or other readily accessible means, 503 | then you must either (1) cause the Corresponding Source to be so 504 | available, or (2) arrange to deprive yourself of the benefit of the 505 | patent license for this particular work, or (3) arrange, in a manner 506 | consistent with the requirements of this License, to extend the patent 507 | license to downstream recipients. "Knowingly relying" means you have 508 | actual knowledge that, but for the patent license, your conveying the 509 | covered work in a country, or your recipient's use of the covered work 510 | in a country, would infringe one or more identifiable patents in that 511 | country that you have reason to believe are valid. 512 | 513 | If, pursuant to or in connection with a single transaction or 514 | arrangement, you convey, or propagate by procuring conveyance of, a 515 | covered work, and grant a patent license to some of the parties 516 | receiving the covered work authorizing them to use, propagate, modify 517 | or convey a specific copy of the covered work, then the patent license 518 | you grant is automatically extended to all recipients of the covered 519 | work and works based on it. 520 | 521 | A patent license is "discriminatory" if it does not include within 522 | the scope of its coverage, prohibits the exercise of, or is 523 | conditioned on the non-exercise of one or more of the rights that are 524 | specifically granted under this License. You may not convey a covered 525 | work if you are a party to an arrangement with a third party that is 526 | in the business of distributing software, under which you make payment 527 | to the third party based on the extent of your activity of conveying 528 | the work, and under which the third party grants, to any of the 529 | parties who would receive the covered work from you, a discriminatory 530 | patent license (a) in connection with copies of the covered work 531 | conveyed by you (or copies made from those copies), or (b) primarily 532 | for and in connection with specific products or compilations that 533 | contain the covered work, unless you entered into that arrangement, 534 | or that patent license was granted, prior to 28 March 2007. 535 | 536 | Nothing in this License shall be construed as excluding or limiting 537 | any implied license or other defenses to infringement that may 538 | otherwise be available to you under applicable patent law. 539 | 540 | 12. No Surrender of Others' Freedom. 541 | 542 | If conditions are imposed on you (whether by court order, agreement or 543 | otherwise) that contradict the conditions of this License, they do not 544 | excuse you from the conditions of this License. If you cannot convey a 545 | covered work so as to satisfy simultaneously your obligations under this 546 | License and any other pertinent obligations, then as a consequence you may 547 | not convey it at all. For example, if you agree to terms that obligate you 548 | to collect a royalty for further conveying from those to whom you convey 549 | the Program, the only way you could satisfy both those terms and this 550 | License would be to refrain entirely from conveying the Program. 551 | 552 | 13. Use with the GNU Affero General Public License. 553 | 554 | Notwithstanding any other provision of this License, you have 555 | permission to link or combine any covered work with a work licensed 556 | under version 3 of the GNU Affero General Public License into a single 557 | combined work, and to convey the resulting work. The terms of this 558 | License will continue to apply to the part which is the covered work, 559 | but the special requirements of the GNU Affero General Public License, 560 | section 13, concerning interaction through a network will apply to the 561 | combination as such. 562 | 563 | 14. Revised Versions of this License. 564 | 565 | The Free Software Foundation may publish revised and/or new versions of 566 | the GNU General Public License from time to time. Such new versions will 567 | be similar in spirit to the present version, but may differ in detail to 568 | address new problems or concerns. 569 | 570 | Each version is given a distinguishing version number. If the 571 | Program specifies that a certain numbered version of the GNU General 572 | Public License "or any later version" applies to it, you have the 573 | option of following the terms and conditions either of that numbered 574 | version or of any later version published by the Free Software 575 | Foundation. If the Program does not specify a version number of the 576 | GNU General Public License, you may choose any version ever published 577 | by the Free Software Foundation. 578 | 579 | If the Program specifies that a proxy can decide which future 580 | versions of the GNU General Public License can be used, that proxy's 581 | public statement of acceptance of a version permanently authorizes you 582 | to choose that version for the Program. 583 | 584 | Later license versions may give you additional or different 585 | permissions. However, no additional obligations are imposed on any 586 | author or copyright holder as a result of your choosing to follow a 587 | later version. 588 | 589 | 15. Disclaimer of Warranty. 590 | 591 | THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY 592 | APPLICABLE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT 593 | HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY 594 | OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, 595 | THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR 596 | PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM 597 | IS WITH YOU. SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF 598 | ALL NECESSARY SERVICING, REPAIR OR CORRECTION. 599 | 600 | 16. Limitation of Liability. 601 | 602 | IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING 603 | WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MODIFIES AND/OR CONVEYS 604 | THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY 605 | GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE 606 | USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF 607 | DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD 608 | PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER PROGRAMS), 609 | EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF 610 | SUCH DAMAGES. 611 | 612 | 17. Interpretation of Sections 15 and 16. 613 | 614 | If the disclaimer of warranty and limitation of liability provided 615 | above cannot be given local legal effect according to their terms, 616 | reviewing courts shall apply local law that most closely approximates 617 | an absolute waiver of all civil liability in connection with the 618 | Program, unless a warranty or assumption of liability accompanies a 619 | copy of the Program in return for a fee. 620 | 621 | END OF TERMS AND CONDITIONS 622 | 623 | How to Apply These Terms to Your New Programs 624 | 625 | If you develop a new program, and you want it to be of the greatest 626 | possible use to the public, the best way to achieve this is to make it 627 | free software which everyone can redistribute and change under these terms. 628 | 629 | To do so, attach the following notices to the program. It is safest 630 | to attach them to the start of each source file to most effectively 631 | state the exclusion of warranty; and each file should have at least 632 | the "copyright" line and a pointer to where the full notice is found. 633 | 634 | 635 | Copyright (C) 636 | 637 | This program is free software: you can redistribute it and/or modify 638 | it under the terms of the GNU General Public License as published by 639 | the Free Software Foundation, either version 3 of the License, or 640 | (at your option) any later version. 641 | 642 | This program is distributed in the hope that it will be useful, 643 | but WITHOUT ANY WARRANTY; without even the implied warranty of 644 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the 645 | GNU General Public License for more details. 646 | 647 | You should have received a copy of the GNU General Public License 648 | along with this program. If not, see . 649 | 650 | Also add information on how to contact you by electronic and paper mail. 651 | 652 | If the program does terminal interaction, make it output a short 653 | notice like this when it starts in an interactive mode: 654 | 655 | Copyright (C) 656 | This program comes with ABSOLUTELY NO WARRANTY; for details type `show w'. 657 | This is free software, and you are welcome to redistribute it 658 | under certain conditions; type `show c' for details. 659 | 660 | The hypothetical commands `show w' and `show c' should show the appropriate 661 | parts of the General Public License. Of course, your program's commands 662 | might be different; for a GUI interface, you would use an "about box". 663 | 664 | You should also get your employer (if you work as a programmer) or school, 665 | if any, to sign a "copyright disclaimer" for the program, if necessary. 666 | For more information on this, and how to apply and follow the GNU GPL, see 667 | . 668 | 669 | The GNU General Public License does not permit incorporating your program 670 | into proprietary programs. If your program is a subroutine library, you 671 | may consider it more useful to permit linking proprietary applications with 672 | the library. If this is what you want to do, use the GNU Lesser General 673 | Public License instead of this License. But first, please read 674 | . 675 | -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | # Ismir2018TutorialNotebooks 2 | 3 | Jupyter notebooks for Ismir-2018 tutorial titled "Computational approaches for analysis of non-Western music traditions" by Serra, Clayton and Bozkurt. 4 | https://www.upf.edu/web/mtg/non-western-music-tutorial 5 | 6 | To run a Jupyter server, we use docker. 7 | 8 | ## Install docker 9 | 10 | ### Windows 11 | https://docs.docker.com/docker-for-windows/install/ 12 | 13 | ### Mac 14 | https://docs.docker.com/docker-for-mac/install/ 15 | 16 | ### Ubuntu 17 | https://docs.docker.com/engine/installation/linux/docker-ce/ubuntu/#install-docker-ce 18 | 19 | ## Run 20 | In a terminal/console window, change to this directory 21 | 22 | On MacOS or Windows, run: 23 | 24 | docker-compose up 25 | 26 | On Linux, run the following (this command ensures that any files you create are owned by your own user): 27 | 28 | JUPYTER_USER_ID=$(id -u) docker-compose up 29 | 30 | Then accesss http://localhost:8888 with your browser and when asked for a 31 | password use the token created automatically and printed in the last line 32 | 33 | Then, you can access the notebooks from the browser and run them. 34 | 35 | Most of the notebooks use our [API](https://github.com/MTG/dunya/blob/master/API_README.md) to access Dunya-server for downloading data. They then process this data and visualize outputs. Accessing Dunya-server requires use of a user specific token. Please refer to https://dunya.compmusic.upf.edu/developers/ for creating a user and getting a token. 36 | 37 | ## List of notebooks (ordered as presented in the tutorial): 38 | * DownloadDataFromDunya_noToken.ipynb: demonstrates use of Dunya API to access CompMusic corpora 39 | * downloadAllSARAGAContent.ipynb: notebook for downloading annotations, features, metadata and audio for the open access collections: Saraga-Hindustani, Saraga-Carnatic 40 | * visualizeAnnotations.ipynb: visualizing manual annotations of Saraga dataset and performing some rhythm analysis tasks 41 | * tuningAnalysis_SingleRecording.ipynb: case study for tuning analysis of a single recording in makam Huseyni 42 | * tuningAnalysis_SetOfRecordings.ipynb: tuning analysis for a set of recordings from the same mode 43 | * generateFileLists4Collections.ipynb: creating files-lists with mode information to help forming experiment subsets for mode recognition tasks 44 | * formExpSubsets4ModeRecognition.ipynb: forms the subsets by grouping recordings with respect to mode or rhythm mode while also checking available files (ex: tonic annotation) for the recording 45 | * symbolicDataPro_symbTr.ipynb: accessing pieces in a specific form and makam from the Turkish Makam Music Symbolic Data Collection (SymbTr), reading a specific a section of the form and ploting the melodic curves 46 | 47 | The presentation files are available on: https://www.upf.edu/web/mtg/non-western-music-tutorial. 48 | 49 | 50 | 51 | -------------------------------------------------------------------------------- /data/IEMP_north_indian_raga/Sample/Annotations/NIR_PrB_Jhinjhoti_2Gats_Annotation_Sample.csv: -------------------------------------------------------------------------------- 1 | START-END,,6.999,77,1610.367,Start-End 2 | FORM,,6.999,77,796.5,Rupak tal 3 | SAROD,,6.9,72.6,65.7,Sthayi 4 | TABLA,,18.1,72.5,54.4,Tabla solo 5 | INTERACTION,,72.352,76.029,3.677,Mutual nod and smile 6 | -------------------------------------------------------------------------------- /data/IEMP_north_indian_raga/Sample/Annotations/NIR_PrB_Jhinjhoti_2Gats_Onsets_Raw_Sarod_Sample.csv: -------------------------------------------------------------------------------- 1 | Time,Peak 2 | 6.654465,0.064164 3 | 7.510898,0.115438 4 | 8.046478,0.061773 5 | 9.037787,0.155225 6 | 10.687933,0.083104 7 | 15.030866,0.113743 8 | 17.032264,0.068177 9 | 20.448615,0.083414 10 | 21.009995,0.244811 11 | 21.986443,0.110318 12 | 24.84104,0.111825 13 | 27.236914,0.470011 14 | 27.558275,0.093422 15 | 27.705234,0.33227 16 | 28.648847,0.089509 17 | 30.016176,0.115782 18 | 30.422514,0.062319 19 | 30.863805,0.177085 20 | 31.37858,0.069081 21 | 31.49741,0.066053 22 | 34.019319,0.18107 23 | 34.912877,0.077511 24 | 36.238417,0.114916 25 | 38.571219,0.060659 26 | 40.219383,0.146915 27 | 41.03454,0.079014 28 | 42.287525,0.172645 29 | 42.736012,0.109845 30 | 43.161772,0.063547 31 | 43.623489,0.108309 32 | 43.749181,0.082967 33 | 44.400379,0.104089 34 | 45.678716,0.365 35 | 45.937994,0.076233 36 | 46.076039,0.25712 37 | 46.955099,0.075965 38 | 48.266099,0.108493 39 | 48.729801,0.090887 40 | 49.547411,0.069434 41 | 52.108928,0.163917 42 | 52.894892,0.074692 43 | 53.300287,0.07172 44 | 53.625471,0.168525 45 | 54.010913,0.114236 46 | 54.117022,0.132827 47 | 54.571554,0.0959 48 | 55.36104,0.136978 49 | 55.457438,0.07631 50 | 56.212532,0.161258 51 | 56.82965,0.121586 52 | 56.955105,0.116735 53 | 57.227531,0.088119 54 | 57.761522,0.285095 55 | 58.556401,0.101941 56 | 59.701892,0.124227 57 | 59.970009,0.099414 58 | 60.360423,0.083499 59 | 60.876558,0.093443 60 | 60.97764,0.067068 61 | 62.531726,0.086926 62 | 63.33696,0.262952 63 | 64.174685,0.140367 64 | 64.52747,0.077891 65 | 65.093084,0.184715 66 | 65.431313,0.141663 67 | 65.741668,0.077519 68 | 65.896684,0.161935 69 | 66.689185,0.162338 70 | 66.786297,0.093178 71 | 67.640932,0.063377 72 | 68.355252,0.204815 73 | 68.72738,0.063966 74 | 68.857779,0.158393 75 | 69.231089,0.325124 76 | 69.32269,0.099038 77 | 70.111151,0.16398 78 | 70.391156,0.180908 79 | 71.231204,0.327126 80 | 71.593535,0.105385 81 | 71.685266,0.123634 82 | 72.600555,0.07637 83 | 76.329646,0.065026 84 | -------------------------------------------------------------------------------- /data/IEMP_north_indian_raga/Sample/Annotations/NIR_PrB_Jhinjhoti_2Gats_Onsets_Raw_Tabla_Sample.csv: -------------------------------------------------------------------------------- 1 | Time,Peak 2 | 21.959794,0.084094 3 | 22.908233,0.038139 4 | 23.355005,0.088653 5 | 24.114575,0.057181 6 | 24.81613,0.061803 7 | 26.751293,0.07851 8 | 27.229643,0.041012 9 | 27.457619,0.041754 10 | 27.691986,0.076369 11 | 28.152662,0.039431 12 | 28.637231,0.115679 13 | 29.102688,0.089389 14 | 29.974645,0.076757 15 | 30.85341,0.035824 16 | 31.316971,0.067409 17 | 32.710212,0.100748 18 | 33.620579,0.061796 19 | 34.062892,0.06832 20 | 34.460394,0.046534 21 | 34.888367,0.077473 22 | 36.219345,0.07196 23 | 36.936069,0.035125 24 | 37.588469,0.080386 25 | 38.300573,0.06797 26 | 38.709158,0.053206 27 | 39.354697,0.063191 28 | 39.765959,0.058915 29 | 40.180953,0.038992 30 | 40.423358,0.050605 31 | 40.61617,0.068416 32 | 41.405211,0.076447 33 | 41.865829,0.0602 34 | 42.094793,0.038837 35 | 42.3019,0.091482 36 | 43.1194,0.077679 37 | 43.539144,0.060925 38 | 44.199414,0.040596 39 | 44.40293,0.038755 40 | 45.012533,0.03693 41 | 45.223481,0.080116 42 | 47.141171,0.056121 43 | 47.359176,0.106347 44 | 49.541484,0.070571 45 | 50.228191,0.0573 46 | 50.676596,0.037113 47 | 50.898974,0.051277 48 | 51.295024,0.084898 49 | 51.513546,0.060273 50 | 51.700791,0.086391 51 | 52.094668,0.074543 52 | 52.308787,0.049702 53 | 52.906825,0.152287 54 | 52.996307,0.046802 55 | 53.090315,0.08912 56 | 53.216022,0.342599 57 | 53.428057,0.059016 58 | 53.549847,0.380126 59 | 53.718837,0.056505 60 | 53.9538,0.256632 61 | 54.064697,0.047533 62 | 54.147695,0.096468 63 | 54.395435,0.073437 64 | 54.52821,0.075821 65 | 54.745651,0.082709 66 | 54.912791,0.127923 67 | 55.143868,0.0406 68 | 55.309802,0.136722 69 | 55.543315,0.057936 70 | 55.732367,0.107966 71 | 56.145863,0.352672 72 | 56.358945,0.043495 73 | 56.562835,0.055883 74 | 56.763665,0.303011 75 | 56.8691,0.066603 76 | 56.981913,0.067333 77 | 57.166302,0.217612 78 | 57.369544,0.100364 79 | 57.578617,0.058594 80 | 57.726792,0.082189 81 | 57.99297,0.066879 82 | 58.120654,0.15048 83 | 58.352029,0.038273 84 | 58.525131,0.128279 85 | 58.742578,0.067335 86 | 59.315689,0.117749 87 | 59.402332,0.068528 88 | 59.486433,0.10642 89 | 59.611612,0.334044 90 | 59.735698,0.067174 91 | 59.923728,0.340654 92 | 60.125607,0.0369 93 | 60.294368,0.278231 94 | 60.499449,0.120244 95 | 60.914031,0.049279 96 | 61.290952,0.171562 97 | 61.922906,0.159617 98 | 62.305083,0.048439 99 | 62.533442,0.284754 100 | 63.179283,0.158155 101 | 63.275363,0.035913 102 | 63.596391,0.055692 103 | 63.788985,0.273743 104 | 64.446069,0.262541 105 | 64.570697,0.042821 106 | 64.839571,0.08277 107 | 65.039085,0.452637 108 | 65.13332,0.131245 109 | 65.661053,0.274943 110 | 66.091659,0.05157 111 | 66.28659,0.25442 112 | 66.712496,0.036529 113 | 66.911177,0.239887 114 | 67.004179,0.043323 115 | 67.324502,0.084627 116 | 67.541006,0.323882 117 | 67.709988,0.050882 118 | 68.199782,0.298536 119 | 68.327151,0.03978 120 | 68.614725,0.059713 121 | 68.815725,0.453055 122 | 68.908733,0.152921 123 | 69.278074,0.044446 124 | 69.465454,0.308829 125 | 69.866246,0.04344 126 | 70.054743,0.460426 127 | 70.15992,0.12422 128 | 70.674853,0.167636 129 | 71.071962,0.041474 130 | 71.272987,0.249707 131 | 71.913068,0.10551 132 | 72.546201,0.125968 133 | 74.328735,0.062759 134 | 74.543106,0.112919 135 | 74.747615,0.041002 136 | 74.941632,0.039804 137 | 75.109931,0.128187 138 | 75.381678,0.044252 139 | 75.557953,0.079107 140 | 75.766474,0.053467 141 | 75.927495,0.066768 142 | 76.733434,0.09996 143 | -------------------------------------------------------------------------------- /data/IEMP_north_indian_raga/Sample/Annotations/NIR_PrB_Jhinjhoti_2Gats_Onsets_Selected_Rupak_Sample.csv: -------------------------------------------------------------------------------- 1 | Session,Inst Name,Tala,Label,Matra,Half beat,Half,Misc 1,Misc 2,Cadence,Tabla solo,Inst,Tabla,Inst Density,Tabla Density,Inst Peak,Tabla Peak,Inst Player,Tabla Player ,Chunk 2 | PrB_Rup,Sarod,Rupak,01:01,1,On,2,,,,TS,10.687933,,1,,0.083104,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 3 | PrB_Rup,Sarod,Rupak,1:1&,1,&,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 4 | PrB_Rup,Sarod,Rupak,01:02,2,On,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 5 | PrB_Rup,Sarod,Rupak,1:2&,2,&,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 6 | PrB_Rup,Sarod,Rupak,01:03,3,On,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 7 | PrB_Rup,Sarod,Rupak,1:3&,3,&,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 8 | PrB_Rup,Sarod,Rupak,01:04,4,On,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 9 | PrB_Rup,Sarod,Rupak,1:4&,4,&,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 10 | PrB_Rup,Sarod,Rupak,01:05,5,On,2,,,,TS,15.030866,,0.5,,0.113743,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 11 | PrB_Rup,Sarod,Rupak,1:5&,5,&,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 12 | PrB_Rup,Sarod,Rupak,01:06,6,On,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 13 | PrB_Rup,Sarod,Rupak,1:6&,6,&,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 14 | PrB_Rup,Sarod,Rupak,01:07,7,On,2,,,,TS,17.032264,,0.5,,0.068177,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 15 | PrB_Rup,Sarod,Rupak,1:7&,7,&,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 16 | PrB_Rup,Sarod,Rupak,02:01,1,On,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 17 | PrB_Rup,Sarod,Rupak,2:1&,1,&,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 18 | PrB_Rup,Sarod,Rupak,02:02,2,On,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 19 | PrB_Rup,Sarod,Rupak,2:2&,2,&,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 20 | PrB_Rup,Sarod,Rupak,02:03,3,On,2,,,,TS,,20.032782,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 21 | PrB_Rup,Sarod,Rupak,2:3&,3,&,2,,,,TS,20.448615,,0.5,,0.083414,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 22 | PrB_Rup,Sarod,Rupak,02:04,4,On,2,,,,TS,21.009995,,1,,0.244811,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 23 | PrB_Rup,Sarod,Rupak,2:4&,4,&,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 24 | PrB_Rup,Sarod,Rupak,02:05,5,On,2,,,,TS,21.986443,21.959794,1.5,,0.110318,0.084094,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 25 | PrB_Rup,Sarod,Rupak,2:5&,5,&,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 26 | PrB_Rup,Sarod,Rupak,02:06,6,On,2,,,,TS,,22.908233,,,,0.038139,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 27 | PrB_Rup,Sarod,Rupak,2:6&,6,&,2,,,,TS,,23.355005,,,,0.088653,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 28 | PrB_Rup,Sarod,Rupak,02:07,7,On,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 29 | PrB_Rup,Sarod,Rupak,2:7&,7,&,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 30 | PrB_Rup,Sarod,Rupak,03:01,1,On,2,,,,TS,24.84104,24.81613,0.5,2,0.111825,0.061803,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 31 | PrB_Rup,Sarod,Rupak,3:1&,1,&,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 32 | PrB_Rup,Sarod,Rupak,03:02,2,On,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 33 | PrB_Rup,Sarod,Rupak,3:2&,2,&,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 34 | PrB_Rup,Sarod,Rupak,03:03,3,On,2,,,,TS,,26.751293,,1,,0.07851,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 35 | PrB_Rup,Sarod,Rupak,3:3&,3,&,2,,,,TS,27.236914,27.229643,0.5,1,0.470011,0.041012,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 36 | PrB_Rup,Sarod,Rupak,03:04,4,On,2,,,,TS,27.705234,27.691986,1.5,2,0.33227,0.076369,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 37 | PrB_Rup,Sarod,Rupak,3:4&,4,&,2,,,,TS,,28.152662,,2.5,,0.039431,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 38 | PrB_Rup,Sarod,Rupak,03:05,5,On,2,,,,TS,28.648847,28.637231,2,3,0.089509,0.115679,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 39 | PrB_Rup,Sarod,Rupak,3:5&,5,&,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 40 | PrB_Rup,Sarod,Rupak,03:06,6,On,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 41 | PrB_Rup,Sarod,Rupak,3:6&,6,&,2,,,,TS,30.016176,29.974645,1,2,0.115782,0.076757,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 42 | PrB_Rup,Sarod,Rupak,03:07,7,On,2,,,,TS,30.422514,,1.5,,0.062319,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 43 | PrB_Rup,Sarod,Rupak,3:7&,7,&,2,,,,TS,30.863805,30.85341,1.5,1.5,0.177085,0.035824,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 44 | PrB_Rup,Sarod,Rupak,04:01,1,On,2,,,,TS,31.37858,31.316971,2,1.5,0.069081,0.067409,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 45 | PrB_Rup,Sarod,Rupak,4:1&,1,&,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 46 | PrB_Rup,Sarod,Rupak,04:02,2,On,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 47 | PrB_Rup,Sarod,Rupak,4:2&,2,&,2,,,,TS,,32.710212,,1.5,,0.100748,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 48 | PrB_Rup,Sarod,Rupak,04:03,3,On,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 49 | PrB_Rup,Sarod,Rupak,4:3&,3,&,2,,,,TS,,33.620579,,1,,0.061796,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 50 | PrB_Rup,Sarod,Rupak,04:04,4,On,2,,,,TS,34.019319,34.062892,0.5,1.5,0.18107,0.06832,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 51 | PrB_Rup,Sarod,Rupak,4:4&,4,&,2,,,,TS,,34.460394,,2,,0.046534,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 52 | PrB_Rup,Sarod,Rupak,04:05,5,On,2,,,,TS,34.912877,34.888367,1,2,0.077511,0.077473,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 53 | PrB_Rup,Sarod,Rupak,4:5&,5,&,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 54 | PrB_Rup,Sarod,Rupak,04:06,6,On,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 55 | PrB_Rup,Sarod,Rupak,4:6&,6,&,2,,,,TS,36.238417,36.219345,1,1.5,0.114916,0.07196,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 56 | PrB_Rup,Sarod,Rupak,04:07,7,On,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 57 | PrB_Rup,Sarod,Rupak,4:7&,7,&,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 58 | PrB_Rup,Sarod,Rupak,05:01,1,On,2,,,,TS,,37.588469,,1.5,,0.080386,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 59 | PrB_Rup,Sarod,Rupak,5:1&,1,&,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 60 | PrB_Rup,Sarod,Rupak,05:02,2,On,2,,,,TS,38.571219,,0.5,,0.060659,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 61 | PrB_Rup,Sarod,Rupak,5:2&,2,&,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 62 | PrB_Rup,Sarod,Rupak,05:03,3,On,2,,,,TS,,39.354697,,2,,0.063191,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 63 | PrB_Rup,Sarod,Rupak,5:3&,3,&,2,,,,TS,,39.765959,,2,,0.058915,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 64 | PrB_Rup,Sarod,Rupak,05:04,4,On,2,,,,TS,40.219383,40.180953,1,2.5,0.146915,0.038992,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 65 | PrB_Rup,Sarod,Rupak,5:4&,4,&,2,,,,TS,,40.61617,,3,,0.068416,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 66 | PrB_Rup,Sarod,Rupak,05:05,5,On,2,,,,TS,41.03454,,1,,0.079014,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 67 | PrB_Rup,Sarod,Rupak,5:5&,5,&,2,,,,TS,,41.405211,,2.5,,0.076447,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 68 | PrB_Rup,Sarod,Rupak,05:06,6,On,2,,,,TS,,41.865829,,2.5,,0.0602,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 69 | PrB_Rup,Sarod,Rupak,5:6&,6,&,2,,,,TS,42.287525,42.3019,1,3,0.172645,0.091482,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 70 | PrB_Rup,Sarod,Rupak,05:07,7,On,2,,,,TS,42.736012,,1.5,,0.109845,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 71 | PrB_Rup,Sarod,Rupak,5:7&,7,&,2,,,,TS,43.161772,43.1194,1.5,2.5,0.063547,0.077679,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 72 | PrB_Rup,Sarod,Rupak,06:01,1,On,2,,,,TS,43.623489,43.539144,2,2.5,0.108309,0.060925,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 73 | PrB_Rup,Sarod,Rupak,6:1&,1,&,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 74 | PrB_Rup,Sarod,Rupak,06:02,2,On,2,,,,TS,44.400379,44.40293,2.5,2,0.104089,0.038755,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 75 | PrB_Rup,Sarod,Rupak,6:2&,2,&,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 76 | PrB_Rup,Sarod,Rupak,06:03,3,On,2,,,,TS,,45.223481,,2.5,,0.080116,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 77 | PrB_Rup,Sarod,Rupak,6:3&,3,&,2,,,,TS,45.678716,,1.5,,0.365,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 78 | PrB_Rup,Sarod,Rupak,06:04,4,On,2,,,,TS,46.076039,,2,,0.25712,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 79 | PrB_Rup,Sarod,Rupak,6:4&,4,&,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 80 | PrB_Rup,Sarod,Rupak,06:05,5,On,2,,,,TS,46.955099,,2,,0.075965,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 81 | PrB_Rup,Sarod,Rupak,6:5&,5,&,2,,,,TS,,47.359176,,1,,0.106347,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 82 | PrB_Rup,Sarod,Rupak,06:06,6,On,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 83 | PrB_Rup,Sarod,Rupak,6:6&,6,&,2,,,,TS,48.266099,,1,,0.108493,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 84 | PrB_Rup,Sarod,Rupak,06:07,7,On,2,,,,TS,48.729801,,1.5,,0.090887,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 85 | PrB_Rup,Sarod,Rupak,6:7&,7,&,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 86 | PrB_Rup,Sarod,Rupak,07:01,1,On,2,,,,TS,49.547411,49.541484,1.5,0.5,0.069434,0.070571,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 87 | PrB_Rup,Sarod,Rupak,7:1&,1,&,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 88 | PrB_Rup,Sarod,Rupak,07:02,2,On,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 89 | PrB_Rup,Sarod,Rupak,7:2&,2,&,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 90 | PrB_Rup,Sarod,Rupak,07:03,3,On,2,,,,TS,,51.295024,,2.5,,0.084898,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 91 | PrB_Rup,Sarod,Rupak,7:3&,3,&,2,,,,TS,,51.700791,,3,,0.086391,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 92 | PrB_Rup,Sarod,Rupak,07:04,4,On,2,,,,TS,52.108928,52.094668,0.5,3.5,0.163917,0.074543,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 93 | PrB_Rup,Sarod,Rupak,7:4&,4,&,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 94 | PrB_Rup,Sarod,Rupak,07:05,5,On,2,,,,TS,52.894892,52.906825,1,3,0.074692,0.152287,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 95 | PrB_Rup,Sarod,Rupak,7:5&,5,&,2,,,,TS,53.300287,,1.5,,0.07172,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 96 | PrB_Rup,Sarod,Rupak,07:06,6,On,2,,,,TS,,53.718837,,4.5,,0.056505,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 97 | PrB_Rup,Sarod,Rupak,7:6&,6,&,2,,,,TS,54.117022,54.064697,2.5,5.5,0.132827,0.047533,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 98 | PrB_Rup,Sarod,Rupak,07:07,7,On,2,,,,TS,54.571554,54.52821,3,6,0.0959,0.075821,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 99 | PrB_Rup,Sarod,Rupak,7:7&,7,&,2,,,,TS,,54.912791,,6.5,,0.127923,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 100 | PrB_Rup,Sarod,Rupak,08:01,1,On,2,,,,TS,55.36104,55.309802,2.5,6,0.136978,0.136722,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 101 | PrB_Rup,Sarod,Rupak,8:1&,1,&,2,,,,TS,,55.732367,,5.5,,0.107966,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 102 | PrB_Rup,Sarod,Rupak,08:02,2,On,2,,,,TS,56.212532,56.145863,2,5,0.161258,0.352672,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 103 | PrB_Rup,Sarod,Rupak,8:2&,2,&,2,,,,TS,,56.562835,,4.5,,0.055883,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 104 | PrB_Rup,Sarod,Rupak,08:03,3,On,2,,,,TS,56.955105,56.981913,2.5,5,0.116735,0.067333,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 105 | PrB_Rup,Sarod,Rupak,8:3&,3,&,2,,,,TS,,57.369544,,5,,0.100364,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 106 | PrB_Rup,Sarod,Rupak,08:04,4,On,2,,,,TS,57.761522,57.726792,2.5,5.5,0.285095,0.082189,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 107 | PrB_Rup,Sarod,Rupak,8:4&,4,&,2,,,,TS,,58.120654,,6,,0.15048,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 108 | PrB_Rup,Sarod,Rupak,08:05,5,On,2,,,,TS,58.556401,58.525131,2.5,6,0.101941,0.128279,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 109 | PrB_Rup,Sarod,Rupak,8:5&,5,&,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 110 | PrB_Rup,Sarod,Rupak,08:06,6,On,2,,,,TS,,59.315689,,4.5,,0.117749,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 111 | PrB_Rup,Sarod,Rupak,8:6&,6,&,2,,,,TS,59.701892,59.735698,1.5,5,0.124227,0.067174,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 112 | PrB_Rup,Sarod,Rupak,08:07,7,On,2,,,,TS,,60.125607,,5,,0.0369,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 113 | PrB_Rup,Sarod,Rupak,8:7&,7,&,2,,,,TS,,60.499449,,5.5,,0.120244,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 114 | PrB_Rup,Sarod,Rupak,09:01,1,On,2,,,,TS,60.876558,60.914031,2,5,0.093443,0.049279,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 115 | PrB_Rup,Sarod,Rupak,9:1&,1,&,2,,,,TS,,61.290952,,5.5,,0.171562,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 116 | PrB_Rup,Sarod,Rupak,09:02,2,On,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 117 | PrB_Rup,Sarod,Rupak,9:2&,2,&,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 118 | PrB_Rup,Sarod,Rupak,09:03,3,On,2,,,,TS,62.531726,62.533442,1.5,2.5,0.086926,0.284754,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 119 | PrB_Rup,Sarod,Rupak,9:3&,3,&,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 120 | PrB_Rup,Sarod,Rupak,09:04,4,On,2,,,,TS,63.33696,,1,,0.262952,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 121 | PrB_Rup,Sarod,Rupak,9:4&,4,&,2,,,,TS,,63.788985,,3.5,,0.273743,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 122 | PrB_Rup,Sarod,Rupak,09:05,5,On,2,,,,TS,64.174685,,1.5,,0.140367,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 123 | PrB_Rup,Sarod,Rupak,9:5&,5,&,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 124 | PrB_Rup,Sarod,Rupak,09:06,6,On,2,,,,TS,,65.039085,,4,,0.452637,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 125 | PrB_Rup,Sarod,Rupak,9:6&,6,&,2,,,,TS,65.431313,,2,,0.141663,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 126 | PrB_Rup,Sarod,Rupak,09:07,7,On,2,,,,TS,65.896684,,3,,0.161935,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 127 | PrB_Rup,Sarod,Rupak,9:7&,7,&,2,,,,TS,,66.28659,,4,,0.25442,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 128 | PrB_Rup,Sarod,Rupak,10:01,1,On,2,,,,TS,66.689185,,2.5,,0.162338,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 129 | PrB_Rup,Sarod,Rupak,10:1&,1,&,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 130 | PrB_Rup,Sarod,Rupak,10:02,2,On,2,,,,TS,,67.541006,,4,,0.323882,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 131 | PrB_Rup,Sarod,Rupak,10:2&,2,&,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 132 | PrB_Rup,Sarod,Rupak,10:03,3,On,2,,,,TS,68.355252,,2,,0.204815,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 133 | PrB_Rup,Sarod,Rupak,10:3&,3,&,2,,,,TS,68.857779,68.815725,2,4.5,0.158393,0.453055,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 134 | PrB_Rup,Sarod,Rupak,10:04,4,On,2,,,,TS,69.231089,,2.5,,0.325124,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 135 | PrB_Rup,Sarod,Rupak,10:4&,4,&,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 136 | PrB_Rup,Sarod,Rupak,10:05,5,On,2,,,,TS,,70.054743,,4.5,,0.460426,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 137 | PrB_Rup,Sarod,Rupak,10:5&,5,&,2,,,,TS,70.391156,,3,,0.180908,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 138 | PrB_Rup,Sarod,Rupak,10:06,6,On,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 139 | PrB_Rup,Sarod,Rupak,10:6&,6,&,2,,,,TS,71.231204,71.272987,2,4,0.327126,0.249707,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 140 | PrB_Rup,Sarod,Rupak,10:07,7,On,2,,,,TS,71.685266,,2.5,,0.123634,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 141 | PrB_Rup,Sarod,Rupak,10:7&,7,&,2,,,,TS,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 142 | PrB_Rup,Sarod,Rupak,11:01,1,On,2,,,C,N,72.600555,72.546201,2,2.5,0.07637,0.125968,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 143 | PrB_Rup,Sarod,Rupak,11:1&,1,&,2,,,,N,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 144 | PrB_Rup,Sarod,Rupak,11:02,2,On,2,,,,N,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 145 | PrB_Rup,Sarod,Rupak,11:2&,2,&,2,,,,N,,,,,,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 146 | PrB_Rup,Sarod,Rupak,11:03,3,On,2,,,,N,,74.328735,,1,,0.062759,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 147 | PrB_Rup,Sarod,Rupak,11:3&,3,&,2,,,,N,,74.747615,,1.5,,0.041002,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 148 | PrB_Rup,Sarod,Rupak,11:04,4,On,2,,,,N,,75.109931,,2.5,,0.128187,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 149 | PrB_Rup,Sarod,Rupak,11:4&,4,&,2,,,,N,,75.557953,,3.5,,0.079107,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 150 | PrB_Rup,Sarod,Rupak,11:05,5,On,2,,,,N,,75.927495,,4.5,,0.066768,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 151 | PrB_Rup,Sarod,Rupak,11:5&,5,&,2,,,,N,76.329646,,0.5,,0.065026,,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 152 | PrB_Rup,Sarod,Rupak,11:06,6,On,2,,,,N,,76.733434,,4,,0.09996,Prattyush Banerjee,Gauri Shankar Karmarkar,1A 153 | -------------------------------------------------------------------------------- /data/IEMP_north_indian_raga/Sample/Annotations/NIR_PrB_jhinjhoti_2Gats_Metre_Rupak_Sample.csv: -------------------------------------------------------------------------------- 1 | Cycle,Time,Notes 2 | 1,10.704, 3 | 2,18.02, 4 | 3,24.80666667, 5 | 4,31.30666667, 6 | 5,37.58666667, 7 | 6,43.568, 8 | 7,49.54133333, 9 | 8,55.308, 10 | 9,60.94114583, 11 | 10,66.708, 12 | 11,72.552, 13 | -------------------------------------------------------------------------------- /data/IEMP_north_indian_raga/Sample/Media/NIR_PrB_Jhinjhoti_2Gats_Sarod_sample.mp3: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/MTG/Ismir2018TutorialNotebooks/8011c1c999164eb36d9b03e76fba35cf6f0a1147/data/IEMP_north_indian_raga/Sample/Media/NIR_PrB_Jhinjhoti_2Gats_Sarod_sample.mp3 -------------------------------------------------------------------------------- /data/IEMP_north_indian_raga/Sample/Media/NIR_PrB_Jhinjhoti_2Gats_StereoMix_sample.mp3: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/MTG/Ismir2018TutorialNotebooks/8011c1c999164eb36d9b03e76fba35cf6f0a1147/data/IEMP_north_indian_raga/Sample/Media/NIR_PrB_Jhinjhoti_2Gats_StereoMix_sample.mp3 -------------------------------------------------------------------------------- /data/IEMP_north_indian_raga/Sample/Media/NIR_PrB_Jhinjhoti_2Gats_Tabla_sample.mp3: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/MTG/Ismir2018TutorialNotebooks/8011c1c999164eb36d9b03e76fba35cf6f0a1147/data/IEMP_north_indian_raga/Sample/Media/NIR_PrB_Jhinjhoti_2Gats_Tabla_sample.mp3 -------------------------------------------------------------------------------- /data/compMusicDatasets/turkishMakam/readme.txt: -------------------------------------------------------------------------------- 1 | The annotations file in this folder comes from the OTMM Makam Recognition Dataset at this repository: 2 | https://github.com/MTG/otmm_makam_recognition_dataset 3 | 4 | OTMM Makam Recognition Dataset 5 | 6 | This repository hosts the dataset designed to test makam recognition methodologies on Ottoman-Turkish makam music. It is composed of 50 recording from each of the 20 most common makams in CompMusic Project's Dunya Ottoman-Turkish Makam Music collection. Currently the dataset is the largest makam recognition dataset. 7 | 8 | Please cite the publication below, if you use this dataset in your work: 9 | Karakurt, A., Şentürk S., & Serra X. (2016). MORTY: A Toolbox for Mode Recognition and Tonic Identification. 3rd International Digital Libraries for Musicology Workshop. New York, USA 10 | 11 | For more information, please refer to the hithub repository read-me. -------------------------------------------------------------------------------- /docker-compose.yml: -------------------------------------------------------------------------------- 1 | version: "2" 2 | services: 3 | mir-tool: 4 | image: mtgupf/mir-toolbox 5 | ports: 6 | - "8888:8888" 7 | volumes: 8 | - .:/notebooks 9 | environment: 10 | - JUPYTER_USER_ID=$JUPYTER_USER_ID 11 | -------------------------------------------------------------------------------- /notebooks/DownloadDataFromDunya_noToken.ipynb: -------------------------------------------------------------------------------- 1 | { 2 | "cells": [ 3 | { 4 | "cell_type": "markdown", 5 | "metadata": {}, 6 | "source": [ 7 | "## Using Dunya API\n", 8 | "\n", 9 | "This notebook demonstrates downloading data using \n", 10 | "the CompMusic Python library: https://github.com/MTG/pycompmusic which includes a client library to access Dunya. `pycompmusic` is already installed in the docker image, and is ready to use. \n", 11 | "\n", 12 | "To be able to download sounds from Dunya, you need to have a user and obtain an API authentication key (token). Please create a user: https://dunya.compmusic.upf.edu/developers/ \n", 13 | "In order to get your API token you have to log in to dunya and then go to your profile where you will find your token. \n", 14 | "\n", 15 | "This example demonstrates:\n", 16 | " * downloading a single file using a recording's MusicBrainz ID\n", 17 | " * downloading files of a CompMusic dataset (https://github.com/MTG/otmm_makam_recognition_dataset)\n", 18 | " \n", 19 | "The [MusicBrainz ID](https://musicbrainz.org/doc/MusicBrainz_Identifier) for a recording is the UUID at the end of a URL for a MusicBrainz page. For example, the recording https://musicbrainz.org/recording/e666ec52-b752-492d-9423-24e1c7bffbc7 has the MusicBrainz ID `e666ec52-b752-492d-9423-24e1c7bffbc7`" 20 | ] 21 | }, 22 | { 23 | "cell_type": "code", 24 | "execution_count": null, 25 | "metadata": {}, 26 | "outputs": [], 27 | "source": [ 28 | "# Set your token here from https://dunya.compmusic.upf.edu/user/profile/\n", 29 | "token = '...yourAPITokenGoesHere...'" 30 | ] 31 | }, 32 | { 33 | "cell_type": "code", 34 | "execution_count": null, 35 | "metadata": {}, 36 | "outputs": [], 37 | "source": [ 38 | "import collections\n", 39 | "import json\n", 40 | "import os\n", 41 | "from compmusic import dunya\n", 42 | "\n", 43 | "dunya.set_token(token)" 44 | ] 45 | }, 46 | { 47 | "cell_type": "markdown", 48 | "metadata": {}, 49 | "source": [ 50 | "#### Downloading a single/specific file: \n", 51 | "https://musicbrainz.org/recording/e666ec52-b752-492d-9423-24e1c7bffbc7\n" 52 | ] 53 | }, 54 | { 55 | "cell_type": "code", 56 | "execution_count": null, 57 | "metadata": {}, 58 | "outputs": [], 59 | "source": [ 60 | "musicbrainzid = 'e666ec52-b752-492d-9423-24e1c7bffbc7'\n", 61 | "data_dir = '../data/compMusicDatasets/turkishMakam/'\n", 62 | "_ = dunya.makam.download_mp3(musicbrainzid, data_dir)" 63 | ] 64 | }, 65 | { 66 | "cell_type": "markdown", 67 | "metadata": {}, 68 | "source": [ 69 | "#### Downloading a set of files\n", 70 | "Example: Audio from the following dataset https://github.com/MTG/otmm_makam_recognition_dataset\n", 71 | "\n", 72 | "The OTMM Makam Recognition Dataset comes with a JSON file listing a number of recordings which exist in Dunya, along with some additional metadata. This file has been copied to this repository.\n", 73 | "\n", 74 | "We are going to download two audio files from each Makam." 75 | ] 76 | }, 77 | { 78 | "cell_type": "code", 79 | "execution_count": null, 80 | "metadata": {}, 81 | "outputs": [], 82 | "source": [ 83 | "# Reading the dataset description file which contains a list of references to audio\n", 84 | "with open(os.path.join(data_dir, 'annotations.json')) as fp:\n", 85 | " collectionFiles = json.load(fp)\n", 86 | "\n", 87 | "# Collecting the list of makams in this dataset\n", 88 | "makams = collections.defaultdict(list)\n", 89 | "for file in collectionFiles:\n", 90 | " makam = file['makam']\n", 91 | " makams[makam].append(file)\n", 92 | "\n", 93 | "# Create sub-directories for makams and download a few files for each makam\n", 94 | "num_files_per_makam = 2\n", 95 | "\n", 96 | "print('Downloading files for {} makams'.format(len(makams)))\n", 97 | "for makam, files in makams.items():\n", 98 | " print(' {}'.format(makam))\n", 99 | " makam_dir = os.path.join(data_dir, makam)\n", 100 | " os.makedirs(makam_dir, exist_ok=True)\n", 101 | " \n", 102 | " for file in files[:num_files_per_makam]:\n", 103 | " musicbrainzid = file['mbid'].split('http://musicbrainz.org/recording/')[-1]\n", 104 | " dunya.makam.download_mp3(musicbrainzid, makam_dir)\n", 105 | "\n", 106 | "print('Sub-folders and files created in {}'.format(data_dir))" 107 | ] 108 | } 109 | ], 110 | "metadata": { 111 | "kernelspec": { 112 | "display_name": "Python 3", 113 | "language": "python", 114 | "name": "python3" 115 | }, 116 | "language_info": { 117 | "codemirror_mode": { 118 | "name": "ipython", 119 | "version": 3 120 | }, 121 | "file_extension": ".py", 122 | "mimetype": "text/x-python", 123 | "name": "python", 124 | "nbconvert_exporter": "python", 125 | "pygments_lexer": "ipython3", 126 | "version": "3.5.2" 127 | } 128 | }, 129 | "nbformat": 4, 130 | "nbformat_minor": 1 131 | } 132 | -------------------------------------------------------------------------------- /notebooks/downloadAllSARAGAContent.ipynb: -------------------------------------------------------------------------------- 1 | { 2 | "cells": [ 3 | { 4 | "cell_type": "markdown", 5 | "metadata": {}, 6 | "source": [ 7 | "## Accessing Saraga database\n", 8 | "\n", 9 | "This notebook demonstrates the use of the Dunya api for downloading Saraga dataset files, which includes Audio (mp3s under a Creative Commons license), metadata, automatically extracted features, and manual annotation files .\n", 10 | "\n", 11 | "The full Saraga dataset is also available for direct download on Zenodo: https://doi.org/10.5281/zenodo.1256126\n", 12 | "\n", 13 | "The Saraga dataset is composed of two collections:\n", 14 | "- [Hindustani collection](https://musicbrainz.org/collection/6adc54c6-6605-4e57-8230-b85f1de5be2b)\n", 15 | "- [Carnatic collection](https://musicbrainz.org/collection/a163c8f2-b75f-4655-86be-1504ea2944c2) \n", 16 | "\n", 17 | "This notebook creates two subfolders and saves all data in these folders. Each annotation is saved in a separate text file. \n", 18 | "\n", 19 | "To be able to download sounds from Dunya, you need to have a user and obtain an API authentication key (token). Please create a user: https://dunya.compmusic.upf.edu/developers/ \n", 20 | "In order to get your API token you have to log in to dunya and then go to your profile where you will find your token. \n", 21 | "\n", 22 | "For example visualisations of the annotations of this data, refer to the 'visualizeAnnotations' notebook \n", 23 | "\n", 24 | "Authors: Sankalp Gulati, Baris Bozkurt" 25 | ] 26 | }, 27 | { 28 | "cell_type": "code", 29 | "execution_count": null, 30 | "metadata": {}, 31 | "outputs": [], 32 | "source": [ 33 | "# Set your token here from https://dunya.compmusic.upf.edu/user/profile/\n", 34 | "token = '...yourAPITokenGoesHere...'" 35 | ] 36 | }, 37 | { 38 | "cell_type": "code", 39 | "execution_count": null, 40 | "metadata": {}, 41 | "outputs": [], 42 | "source": [ 43 | "import codecs\n", 44 | "import json, os, sys\n", 45 | "import pickle\n", 46 | "import csv\n", 47 | "import time\n", 48 | "import datetime\n", 49 | "import collections\n", 50 | "\n", 51 | "import numpy as np\n", 52 | "\n", 53 | "import compmusic\n", 54 | "\n", 55 | "from compmusic import dunya as dn\n", 56 | "from compmusic.dunya import hindustani as hi\n", 57 | "from compmusic.dunya import carnatic as ca\n", 58 | "from compmusic.dunya import docserver as ds\n", 59 | "from compmusic import musicbrainz\n", 60 | "\n", 61 | "dn.set_token(token)" 62 | ] 63 | }, 64 | { 65 | "cell_type": "code", 66 | "execution_count": null, 67 | "metadata": {}, 68 | "outputs": [], 69 | "source": [ 70 | "#Features list\n", 71 | "features_dunya_all = [{'type': 'pitch', 'subtype': 'pitch', 'extension': '.pitch', 'version': 'noguessunv'},\n", 72 | " {'type': 'ctonic', 'subtype': 'tonic', 'extension': '.tonic', 'version': '0.3'},\n", 73 | " {'type': 'sama-manual', 'subtype': None, 'extension': '.sama', 'version': None},\n", 74 | " {'type': 'sections-manual', 'subtype': None, 'extension': '.sections', 'version': None},\n", 75 | " {'type': 'tempo-manual', 'subtype': None, 'extension': '.tempo', 'version': None},\n", 76 | " {'type': 'pitch-vocal', 'subtype': None, 'extension': '.mpitch', 'version': None},\n", 77 | " {'type': 'mphrases-manual', 'subtype': None, 'extension': '.mphrases', 'version': None},\n", 78 | " {'type': 'sections-manual-p', 'subtype': None, 'extension': '.sections_p', 'version': None},\n", 79 | " {'type': 'bpm-manual', 'subtype': None, 'extension': '.bpm', 'version': None}\n", 80 | " ]" 81 | ] 82 | }, 83 | { 84 | "cell_type": "markdown", 85 | "metadata": {}, 86 | "source": [ 87 | "### Functions for accessing files, computing statistics and writing/saving files" 88 | ] 89 | }, 90 | { 91 | "cell_type": "code", 92 | "execution_count": null, 93 | "metadata": {}, 94 | "outputs": [], 95 | "source": [ 96 | "def getStatsDunyaCorpus():\n", 97 | " \"\"\"\n", 98 | " Compute and save statistics for the Hindustani and Carnatic collections.\n", 99 | " \n", 100 | " Outputs:\n", 101 | " A Pickle to 'stats_{collection}_cc.pkl' of the MusicBrainz IDs that appear in the collection\n", 102 | " A text file to 'stats_{collection}_cc.txt' showing summary counts of items that appear in the collection\n", 103 | " \"\"\"\n", 104 | "\n", 105 | " carnatic_stats = get_stats_carnatic(DUNYA_COLLECTIONS['carnatic'])\n", 106 | " output_file = 'stats_carnatic_cc.pkl'\n", 107 | " output_file_pretty = 'stats_carnatic_cc.txt'\n", 108 | " save_stats(carnatic_stats, output_file, output_file_pretty)\n", 109 | " \n", 110 | " hindustani_stats = get_stats_hindustani(DUNYA_COLLECTIONS['hindustani'])\n", 111 | " output_file = 'stats_hindustani_cc.pkl'\n", 112 | " output_file_pretty = 'stats_hindustani_cc.txt'\n", 113 | " save_stats(hindustani_stats, output_file, output_file_pretty)\n", 114 | "\n", 115 | "\n", 116 | "def get_stats_hindustani(dunya_collections=None):\n", 117 | " \"\"\"Get information about hindustani recordings and return a summary of attributes.\n", 118 | " For the following attributes:\n", 119 | " release\n", 120 | " works\n", 121 | " raags\n", 122 | " taals\n", 123 | " forms\n", 124 | " layas\n", 125 | " album artists\n", 126 | " artists (musicians)\n", 127 | " generate a list of identifiers for these attributes (mbid or uuid [raag, taal, laya] or name [form])\n", 128 | " present in the collection\n", 129 | " \n", 130 | " Args:\n", 131 | " dunya_collections: a list of MusicBrainz/Dunya Collection IDs to restrict the Dunya API to\n", 132 | " \"\"\"\n", 133 | " \n", 134 | " hi.set_collections(dunya_collections)\n", 135 | " recordings = hi.get_recordings()\n", 136 | "\n", 137 | " stats = collections.defaultdict(list)\n", 138 | " for r in recordings:\n", 139 | " mbid = r['mbid']\n", 140 | "\n", 141 | " try:\n", 142 | " rec_info = hi.get_recording(mbid)\n", 143 | "\n", 144 | " stats['release'].append([r['mbid'] for r in rec_info.get('release', [])])\n", 145 | " stats['works'].append([w['mbid'] for w in rec_info.get('works', [])])\n", 146 | " stats['raags'].append([r['uuid'] for r in rec_info.get('raags', [])])\n", 147 | " stats['taals'].append([t['uuid'] for t in rec_info.get('taals', [])])\n", 148 | " stats['forms'].append([f['name'] for f in rec_info.get('forms', [])])\n", 149 | " stats['layas'].append([l['uuid'] for l in rec_info.get('layas', [])])\n", 150 | " stats['album_artists'].append([a['mbid'] for a in rec_info.get('album_artists', [])])\n", 151 | " stats['artists'].append([a['artist']['mbid'] for a in rec_info.get('artists', [])])\n", 152 | " stats['length'].append(rec_info.get('length'))\n", 153 | " except:\n", 154 | " failure+=1\n", 155 | " print(\"%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%\")\n", 156 | " print(\"Failed to fetch info for recording %s\" % mbid) \n", 157 | " print(\"%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%\")\n", 158 | " \n", 159 | " # Filter empty lists from the stats\n", 160 | " for k, vals in stats.items():\n", 161 | " stats[k] = [v for v in vals if v]\n", 162 | " \n", 163 | " return stats\n", 164 | "\n", 165 | "def get_stats_carnatic(dunya_collections=None):\n", 166 | " \"\"\"Get information about carnatic recordings and return a summary of attributes.\n", 167 | " For the following attributes:\n", 168 | " concert\n", 169 | " work\n", 170 | " raaga\n", 171 | " taala\n", 172 | " form\n", 173 | " album artists\n", 174 | " artists (musicians)\n", 175 | " generate a list of identifiers for these attributes (mbid or uuid [raaga, taala] or name [form])\n", 176 | " present in the collection\n", 177 | " \n", 178 | " Args:\n", 179 | " dunya_collections: a list of MusicBrainz/Dunya Collection IDs to restrict the Dunya API to\n", 180 | " \"\"\"\n", 181 | " ca.set_collections(dunya_collections)\n", 182 | " recordings = ca.get_recordings()\n", 183 | " \n", 184 | " stats = collections.defaultdict(list)\n", 185 | " for r in recordings:\n", 186 | " mbid = r['mbid']\n", 187 | "\n", 188 | " try:\n", 189 | " rec_info = ca.get_recording(mbid)\n", 190 | "\n", 191 | " stats['concert'].append([c['mbid'] for c in rec_info.get('concert', [])])\n", 192 | " stats['work'].append([w['mbid'] for w in rec_info.get('work', [])])\n", 193 | " stats['raaga'].append([r['uuid'] for r in rec_info.get('raaga', [])])\n", 194 | " stats['taala'].append([t['uuid'] for t in rec_info.get('taala', [])])\n", 195 | " stats['form'].append([f['name'] for f in rec_info.get('form', [])])\n", 196 | " stats['album_artists'].append([a['mbid'] for a in rec_info.get('album_artists', [])])\n", 197 | " stats['artists'].append([a['artist']['mbid'] for a in rec_info.get('artists', [])])\n", 198 | " stats['length'].append(rec_info.get('length'))\n", 199 | " except dn.HTTPError:\n", 200 | " print(\"%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%\")\n", 201 | " print(\"Failed to fetch info for recording %s\" % mbid) \n", 202 | " print(\"%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%\")\n", 203 | " \n", 204 | " # Filter empty lists from the stats\n", 205 | " for k, vals in stats.items():\n", 206 | " stats[k] = [v for v in vals if v]\n", 207 | " \n", 208 | " return stats\n", 209 | "\n", 210 | " \n", 211 | "def save_stats(stats, stats_file, summary_file):\n", 212 | " \"\"\"Write statistics to file\n", 213 | " Args:\n", 214 | " stats (dict): the statistics to write\n", 215 | " stats_file (str): file path to write the statistics summary\n", 216 | " summary_file (str): file path to write a readable statistics summary\n", 217 | " \"\"\"\n", 218 | " \n", 219 | " output_stats = {}\n", 220 | " for k, v in stats.items():\n", 221 | " if k == 'length':\n", 222 | " output_stats[k] = {'total_length': np.sum(v), 'total_recs': len(v)}\n", 223 | " else:\n", 224 | " output_stats[k] = {'total_unique': len(np.unique(sum(v, []))), 'unique_elems': np.unique(sum(v, [])).tolist(), 'total_rels': len(sum(v, [])), 'total_recs': len(v)}\n", 225 | " pickle.dump(output_stats, codecs.open(stats_file, 'wb'))\n", 226 | " \n", 227 | " \n", 228 | " with codecs.open(summary_file, 'w') as fp:\n", 229 | " for key1, val in output_stats.items():\n", 230 | " fp.write('------------ %s ------------\\n'%str(key1))\n", 231 | " if key1 == 'length':\n", 232 | " for key2, val2 in val.items():\n", 233 | " fp.write('%s\\t%f\\n'%(str(key2), float(val2)/(1000.0*3600.0)))\n", 234 | " else:\n", 235 | " for key2, val2 in val.items():\n", 236 | " if key2 == 'unique_elems':\n", 237 | " fp.write('%s\\t%d\\n'%(str(key2), len(val2)))\n", 238 | " else:\n", 239 | " fp.write('%s\\t%d\\n'%(str(key2), val2))\n", 240 | " fp.write('\\n')\n", 241 | "\n", 242 | "def saveSections(content, output_file):\n", 243 | " \"\"\"\n", 244 | " This function saves the content(section annotations) into a file in a structured manner\n", 245 | " Annotations are already stored nicely but due to differences in the delimiters of Hindustani and Carnatic\n", 246 | " we needed this function\n", 247 | " \n", 248 | " Args:\n", 249 | " content (str): data read from dunya api\n", 250 | " output_file (str): file path for output file\n", 251 | " Outputs:\n", 252 | " Saves statistics to a text file\n", 253 | " \"\"\"\n", 254 | " \n", 255 | " # detecting delimiter automatically\n", 256 | " snf = csv.Sniffer()\n", 257 | " delimiter = snf.sniff(content).delimiter\n", 258 | " rows = [k.split(delimiter) for k in content.split('\\n') if k != '']\n", 259 | " csv.writer(output_file, rows, delimiter = '\\t')\n", 260 | "\n", 261 | "\n", 262 | "def download_files_for_collection(collection_name, collection_ids, features, numFiles=5):\n", 263 | " \"\"\"Download all files of a collection\n", 264 | " Args:\n", 265 | " collection (dict): dictionary containig name and id of the collection\n", 266 | " features (list of dicts): feature types\n", 267 | " numFiles (int): the maximum number of files to download\n", 268 | " Returns:\n", 269 | " A dictionary counting how many files for each feature was unable to be downloaded\n", 270 | " Outputs:\n", 271 | " Saves mp3 and annotation files of the collection\n", 272 | " \"\"\"\n", 273 | " dataDir = collection_name\n", 274 | " os.makedirs(dataDir, exist_ok=True)\n", 275 | "\n", 276 | " if collection_name == 'hindustani':\n", 277 | " tradition = hi\n", 278 | " elif collection_name == 'carnatic':\n", 279 | " tradition = ca\n", 280 | " \n", 281 | " tradition.set_collections(collection_ids)\n", 282 | " recs = tradition.get_recordings()\n", 283 | " \n", 284 | " numFiles = min(numFiles, len(recs))\n", 285 | " \n", 286 | " print('Number of files in collection {}: {}'.format(collection_name, len(recs)))\n", 287 | " print('...will download {} files'.format(numFiles))\n", 288 | " \n", 289 | " # Creating data structure for keeping list of missing files\n", 290 | " missingData = collections.Counter()\n", 291 | " for feature in features:\n", 292 | " missingData[feature['type']] = 0\n", 293 | " \n", 294 | " # Downloading data\n", 295 | " for i, recording in enumerate(recs[:numFiles], 1):\n", 296 | " mbid = recording['mbid']\n", 297 | " print('{}/{}: {}'.format(i, len(recs), mbid))\n", 298 | " mp3_filename = tradition.download_mp3(mbid, dataDir)\n", 299 | " json_file = mp3_filename.replace('.mp3', '.json')\n", 300 | " with open(os.path.join(dataDir, json_file), 'w') as outfile:\n", 301 | " json.dump(tradition.get_recording(mbid), outfile)\n", 302 | " \n", 303 | " print(mp3_filename)\n", 304 | "\n", 305 | " for feature in features:\n", 306 | " try:\n", 307 | " content = ds.file_for_document(mbid, feature['type'], feature['subtype'], version=feature['version'])\n", 308 | " \n", 309 | " out_file = os.path.join(dataDir, mp3_filename.replace('.mp3','.{}.txt'.format(feature['type'])))\n", 310 | " if feature['type'] == 'pitch':\n", 311 | " content = json.loads(content.decode())\n", 312 | " content = np.array(content)\n", 313 | " np.savetxt(out_file, content, fmt='%.7f', delimiter='\\t')\n", 314 | " #elif feature['type'] == 'sections-manual' or feature['type'] == 'sections-manual-p':\n", 315 | " # saveSections(content.decode(), out_file)\n", 316 | " else:\n", 317 | " with open(out_file, 'w') as fp:\n", 318 | " fp.write(content.decode())\n", 319 | " except dn.HTTPError:\n", 320 | " #print('Does not have ',feature['type'],' content for :',mbid)\n", 321 | " missingData[feature['type']] += 1\n", 322 | " \n", 323 | " print('Collection download finished.')\n", 324 | " print('----------------------------------------------------------')\n", 325 | " return dict(missingData)" 326 | ] 327 | }, 328 | { 329 | "cell_type": "markdown", 330 | "metadata": {}, 331 | "source": [ 332 | "### Setting collections to be downloaded\n", 333 | "Collections are specified with a name and musicBrainz id. [All CompMusic collections are listed here](https://musicbrainz.org/user/compmusic/collections)\n", 334 | "\n", 335 | "ID refers to the last part of a musicBrainz link for the collection such as\n", 336 | "https://musicbrainz.org/collection/a163c8f2-b75f-4655-86be-1504ea2944c2 for the Carnatic collection" 337 | ] 338 | }, 339 | { 340 | "cell_type": "code", 341 | "execution_count": null, 342 | "metadata": {}, 343 | "outputs": [], 344 | "source": [ 345 | "DUNYA_COLLECTIONS = {'hindustani': ['6adc54c6-6605-4e57-8230-b85f1de5be2b'],\n", 346 | " 'carnatic': ['a163c8f2-b75f-4655-86be-1504ea2944c2']\n", 347 | " }" 348 | ] 349 | }, 350 | { 351 | "cell_type": "code", 352 | "execution_count": null, 353 | "metadata": {}, 354 | "outputs": [], 355 | "source": [ 356 | "# Calling functions to produce statistics and download data\n", 357 | "\n", 358 | "missingDatas = {}\n", 359 | "NUM_FILES = 5 #set to 200 if you like to download all data (CAUTION: 8Gb)\n", 360 | "\n", 361 | "print('Starting process: {}'.format(datetime.datetime.now()))\n", 362 | "print('Collecting statistics of carnatic collection')\n", 363 | "carnatic_stats = get_stats_carnatic(DUNYA_COLLECTIONS['carnatic'])\n", 364 | "output_file = 'stats_carnatic_cc.pkl'\n", 365 | "output_file_pretty = 'stats_carnatic_cc.txt'\n", 366 | "save_stats(carnatic_stats, output_file, output_file_pretty)\n", 367 | "\n", 368 | "print('Downloading files ... ')\n", 369 | "missingData = download_files_for_collection('carnatic', DUNYA_COLLECTIONS['carnatic'], features_dunya_all, NUM_FILES)\n", 370 | "missingDatas['carnatic'] = missingData\n", 371 | "\n", 372 | "print('...Done')\n", 373 | "print('Collecting statistics of hindustani collection')\n", 374 | "\n", 375 | "hindustani_stats = get_stats_hindustani(DUNYA_COLLECTIONS['hindustani'])\n", 376 | "output_file = 'stats_hindustani_cc.pkl'\n", 377 | "output_file_pretty = 'stats_hindustani_cc.txt'\n", 378 | "save_stats(hindustani_stats, output_file, output_file_pretty)\n", 379 | "\n", 380 | "print('Downloading files ... ')\n", 381 | "missingData = download_files_for_collection('hindustani', DUNYA_COLLECTIONS['hindustani'], features_dunya_all, NUM_FILES)\n", 382 | "missingDatas['hindustani'] = missingData\n", 383 | "print('...Done')\n", 384 | "\n", 385 | "pickle.dump(missingDatas, codecs.open('missingData.pkl', 'wb'))\n", 386 | "print('Missing data list stored in missingData.pkl')\n", 387 | "\n", 388 | "print('Finished! {}'.format(datetime.datetime.now()))" 389 | ] 390 | } 391 | ], 392 | "metadata": { 393 | "kernelspec": { 394 | "display_name": "Python 3", 395 | "language": "python", 396 | "name": "python3" 397 | }, 398 | "language_info": { 399 | "codemirror_mode": { 400 | "name": "ipython", 401 | "version": 3 402 | }, 403 | "file_extension": ".py", 404 | "mimetype": "text/x-python", 405 | "name": "python", 406 | "nbconvert_exporter": "python", 407 | "pygments_lexer": "ipython3", 408 | "version": "3.5.2" 409 | } 410 | }, 411 | "nbformat": 4, 412 | "nbformat_minor": 2 413 | } 414 | -------------------------------------------------------------------------------- /notebooks/external_utilities/converter.py: -------------------------------------------------------------------------------- 1 | from __future__ import division 2 | import numpy as np 3 | _NUM_CENTS_IN_OCTAVE = 1200.0 4 | 5 | 6 | class Converter(object): 7 | @staticmethod 8 | def hz_to_cent(hz_track, ref_freq, min_freq=20.0): 9 | """-------------------------------------------------------------------- 10 | Converts an array of Hertz values into cents. 11 | ----------------------------------------------------------------------- 12 | hz_track : The 1-D array of Hertz values 13 | ref_freq : Reference frequency for cent conversion 14 | min_freq : The minimum frequency allowed (exclusive) 15 | --------------------------------------------------------------------""" 16 | # The 0 Hz values are removed, not only because they are meaningless, 17 | # but also logarithm of 0 is problematic. 18 | assert min_freq >= 0.0, 'min_freq cannot be less than 0' 19 | 20 | hz_track = np.array(hz_track).astype(float) 21 | 22 | # change values less than the min_freq to nan 23 | hz_track[hz_track <= min_freq] = np.nan 24 | 25 | return np.log2(hz_track / ref_freq) * _NUM_CENTS_IN_OCTAVE 26 | 27 | @staticmethod 28 | def cent_to_hz(cent_track, ref_freq): 29 | """-------------------------------------------------------------------- 30 | Converts an array of cent values into Hertz. 31 | ----------------------------------------------------------------------- 32 | cent_track : The 1-D array of cent values 33 | ref_freq : Reference frequency for cent conversion 34 | --------------------------------------------------------------------""" 35 | cent_track = np.array(cent_track).astype(float) 36 | 37 | return 2 ** (cent_track / _NUM_CENTS_IN_OCTAVE) * ref_freq 38 | -------------------------------------------------------------------------------- /notebooks/external_utilities/pitchdistribution.py: -------------------------------------------------------------------------------- 1 | # -*- coding: utf-8 -*- 2 | from __future__ import division 3 | import essentia 4 | import essentia.standard as std 5 | import numpy as np 6 | import json 7 | import copy 8 | import scipy.stats 9 | import scipy.integrate 10 | import matplotlib.pyplot as plt 11 | from external_utilities.converter import Converter 12 | import numbers 13 | import pickle 14 | import logging 15 | logging.basicConfig(level=logging.INFO) 16 | logger = logging.getLogger(__name__) 17 | 18 | 19 | class PitchDistribution(object): 20 | def __init__(self, pd_bins, pd_vals, kernel_width=7.5, ref_freq=440.0): 21 | """------------------------------------------------------------------- 22 | The main data structure that wraps all the relevant information about a 23 | pitch distribution. 24 | ---------------------------------------------------------------------- 25 | pd_bins : Bins of the pitch distribution. It is a 1-D list of 26 | equally spaced monotonically increasing frequency 27 | values. 28 | pd_vals : Values of the pitch distribution 29 | kernel_width : The standard deviation of the Gaussian kernel. See 30 | generate_pd() of ModeFunctions for more detail. 31 | ref_freq : Reference frequency that is used while generating the 32 | distribution. If the tonic of a recording is annotated, 33 | this is variable that stores it. 34 | --------------------------------------------------------------------""" 35 | assert len(pd_bins) == len(pd_vals), 'Lengths of bins and vals are ' \ 36 | 'different.' 37 | 38 | self.bins = np.array(pd_bins) # force numpy array 39 | self.vals = np.array(pd_vals) # force numpy array 40 | self.kernel_width = kernel_width 41 | if ref_freq is None: 42 | self.ref_freq = None 43 | else: 44 | self.ref_freq = np.array(ref_freq) # force numpy array 45 | 46 | @property 47 | def step_size(self): 48 | # get step size in cents 49 | if self.has_hz_bin(): 50 | temp_ss = Converter.hz_to_cent(self.bins[1], self.bins[0]) 51 | else: # has_cent_bin 52 | temp_ss = self.bins[1] - self.bins[0] 53 | 54 | # TEMPORARY FIX: round step_size to one decimal point 55 | return temp_ss if temp_ss == (round(temp_ss * 10) / 10) \ 56 | else round(temp_ss * 10) / 10 57 | 58 | @property 59 | def bin_unit(self): 60 | err_str = 'Invalid reference. ref_freq should be either None ' \ 61 | '(bin unit is Hz) or a number greater than 0.' 62 | if self.ref_freq is None: 63 | return 'Hz' 64 | elif isinstance(self.ref_freq, (numbers.Number, np.ndarray)): 65 | assert self.ref_freq > 0, err_str 66 | return 'cent' 67 | else: 68 | return ValueError(err_str) 69 | 70 | @staticmethod 71 | def from_cent_pitch(cent_track, ref_freq=440.0, kernel_width=7.5, 72 | step_size=7.5, norm_type='sum'): 73 | """-------------------------------------------------------------------- 74 | Given the pitch track in the unit of cents, generates the Pitch 75 | Distribution of it. the pitch track from a text file. 0th column is the 76 | time-stamps and 77 | 1st column is the corresponding frequency values. 78 | ----------------------------------------------------------------------- 79 | cent_track: 1-D array of frequency values in cents. 80 | ref_freq: Reference frequency used while converting Hz values to 81 | cents. 82 | This number isn't used in the computations, but is to 83 | be recorded in the PitchDistribution object. 84 | kernel_width: The standard deviation of the gaussian kernel, used in 85 | Kernel Density Estimation. If 0, a histogram is given 86 | step_size: The step size of the Pitch Distribution bins. 87 | --------------------------------------------------------------------""" 88 | assert step_size > 0, 'The step size should have a positive value' 89 | 90 | # Some extra interval is added to the beginning and end since the 91 | # superposed Gaussian for kernel_width would introduce some tails in 92 | # the ends. These vanish after 3 sigmas(=kernel_width). 93 | 94 | # The limits are also quantized to be a multiple of chosen step-size 95 | # kernel_width = standard deviation of the gaussian kernel 96 | # parse the cent_track 97 | try: 98 | cent_track = np.loadtxt(cent_track) 99 | except ValueError: 100 | logger.debug('cent_track is already a numpy array') 101 | 102 | if cent_track.ndim > 1: # pitch is given as [time, pitch, (conf)] 103 | cent_track = cent_track[:, 1] 104 | 105 | # filter out NaN, and infinity 106 | cent_track = cent_track[~np.isnan(cent_track)] 107 | cent_track = cent_track[~np.isinf(cent_track)] 108 | 109 | # Finds the endpoints of the histogram edges. Histogram bins will be 110 | # generated as the midpoints of these edges. 111 | min_edge = min(cent_track) - (step_size / 2.0) 112 | max_edge = max(cent_track) + (step_size / 2.0) 113 | pd_edges = np.concatenate( 114 | [np.arange(-step_size / 2.0, min_edge, -step_size)[::-1], 115 | np.arange(step_size / 2.0, max_edge, step_size)]) 116 | 117 | # An exceptional case is when min_bin and max_bin are both positive 118 | # In this case, pd_edges would be in the range of [step_size/2, max_ 119 | # bin]. If so, a -step_size is inserted to the head, to make sure 0 120 | # would be in pd_bins. The same procedure is repeated for the case 121 | # when both are negative. Then, step_size is inserted to the tail. 122 | pd_edges = pd_edges if -step_size / 2.0 in pd_edges else np.insert( 123 | pd_edges, 0, -step_size / 2.0) 124 | pd_edges = pd_edges if step_size / 2.0 in pd_edges else np.append( 125 | pd_edges, step_size / 2.0) 126 | 127 | # Generates the histogram and bins (i.e. the midpoints of edges) 128 | pd_vals, pd_edges = np.histogram(cent_track, bins=pd_edges, 129 | density=False) 130 | pd_bins = np.convolve(pd_edges, [0.5, 0.5])[1:-1] # the bin centers 131 | 132 | # initialize the distribution 133 | pd = PitchDistribution(pd_bins, pd_vals, kernel_width=0, 134 | ref_freq=ref_freq) 135 | pd.smoothen(kernel_width=kernel_width) 136 | 137 | # normalize 138 | pd.normalize(norm_type=norm_type) 139 | 140 | return pd 141 | 142 | def smoothen(self, kernel_width=7.5): 143 | if kernel_width > 0: 144 | # smooth the histogram 145 | normal_dist = scipy.stats.norm(loc=0, scale=kernel_width) 146 | xn = np.concatenate( 147 | [np.arange(0, - 5 * kernel_width, -self.step_size)[::-1], 148 | np.arange(self.step_size, 5 * kernel_width, self.step_size)]) 149 | sampled_norm = normal_dist.pdf(xn) 150 | if len(sampled_norm) <= 1: 151 | raise ValueError( 152 | "the smoothing factor is too small compared to the step " 153 | "size, such that the convolution kernel returns a single " 154 | "point Gaussian. Either increase the value to at least " 155 | "(step size/3) or assign kernel width to 0, for no " 156 | "smoothing.") 157 | # convolution generates tails 158 | extra_num_bins = np.floor(len(sampled_norm) / 2) 159 | 160 | self.bins = np.concatenate( 161 | (np.arange(self.bins[0] - extra_num_bins * self.step_size, 162 | self.bins[0], self.step_size), self.bins, 163 | np.arange(self.bins[-1] + self.step_size, self.bins[-1] + 164 | extra_num_bins * self.step_size + self.step_size, 165 | self.step_size))) 166 | self.vals = np.convolve(self.vals, sampled_norm) 167 | assert len(self.bins) == len(self.vals), 'Lengths of bins and ' \ 168 | 'vals are different.' 169 | self.kernel_width = (kernel_width if self.kernel_width == 0 else 170 | self.kernel_width * kernel_width) 171 | 172 | @staticmethod 173 | def from_hz_pitch(hz_track, ref_freq=440.0, kernel_width=7.5, 174 | step_size=7.5, norm_type='sum'): 175 | try: 176 | hz_track = np.loadtxt(hz_track) 177 | except ValueError: 178 | logger.debug('hz_track is already a numpy array') 179 | 180 | if hz_track.ndim > 1: # pitch is given as [time, pitch, (conf)] array 181 | hz_track = hz_track[:, 1] 182 | 183 | # filter out the NaN, -infinity and +infinity and values < 20 184 | hz_track = hz_track[~np.isnan(hz_track)] 185 | hz_track = hz_track[~np.isinf(hz_track)] 186 | hz_track = hz_track[hz_track >= 20.0] 187 | cent_track = Converter.hz_to_cent(hz_track, ref_freq, min_freq=20.0) 188 | 189 | return PitchDistribution.from_cent_pitch( 190 | cent_track, ref_freq=ref_freq, kernel_width=kernel_width, 191 | step_size=step_size, norm_type=norm_type) 192 | 193 | def __eq__(self, other): 194 | eq_bool = True 195 | self_dict = self.__dict__ 196 | other_dict = other.__dict__ 197 | 198 | # numpy array need to be compared with np.allclose 199 | eq_bool = eq_bool and np.allclose(self_dict.pop("vals", None), 200 | other_dict.pop("vals", None)) 201 | eq_bool = eq_bool and np.allclose(self_dict.pop("bins", None), 202 | other_dict.pop("bins", None)) 203 | 204 | return eq_bool and self_dict == other_dict 205 | 206 | def is_pcd(self): 207 | """-------------------------------------------------------------------- 208 | The boolean flag of whether the instance is PCD or not. 209 | --------------------------------------------------------------------""" 210 | if self.has_cent_bin(): # cent bins; compare directly 211 | return np.isclose(max(self.bins) - min(self.bins), 212 | 1200 - self.step_size) 213 | else: 214 | dummy_d = copy.deepcopy(self) 215 | 216 | dummy_d.hz_to_cent(dummy_d.bins[0]) 217 | 218 | return np.isclose(max(dummy_d.bins) - min(dummy_d.bins), 219 | 1200 - dummy_d.step_size) 220 | 221 | def is_pdf(self): 222 | return np.isclose(np.sum(self.vals), 1) 223 | 224 | def distrib_type(self): 225 | return 'pcd' if self.is_pcd() else 'pd' 226 | 227 | def has_hz_bin(self): 228 | return self.bin_unit in ['hz', 'Hz', 'Hertz', 'hertz'] 229 | 230 | def has_cent_bin(self): 231 | return self.bin_unit in ['cent', 'Cent', 'cents', 'Cents'] 232 | 233 | def normalize(self, norm_type='sum'): 234 | if norm_type is None: # nothing, keep the occurrences (histogram) 235 | normval = 1 236 | elif norm_type == 'area': # area under the curve using simpsons rule 237 | normval = scipy.integrate.simps(self.vals, dx=self.step_size) 238 | elif norm_type == 'sum': # sum normalization 239 | normval = np.sum(self.vals) 240 | elif norm_type == 'max': # max number becomes 1 241 | normval = max(self.vals) 242 | else: 243 | raise ValueError("norm_type can be None, 'area', 'sum' or 'max'") 244 | 245 | self.vals = self.vals / normval 246 | 247 | def detect_peaks(self, min_peak_ratio=0.15): 248 | """-------------------------------------------------------------------- 249 | Finds the peak indices of the distribution. These are treated as tonic 250 | candidates in higher order functions. 251 | min_peak_ratio: The minimum ratio between the max peak value and the 252 | value of a detected peak 253 | --------------------------------------------------------------------""" 254 | assert 1 >= min_peak_ratio >= 0, \ 255 | 'min_peak_ratio should be between 0 (keep all peaks) and ' \ 256 | '1 (keep only the highest peak)' 257 | 258 | # Peak detection is handled by Essentia 259 | detector = std.PeakDetection() 260 | peak_bins, peak_vals = detector(essentia.array(self.vals)) 261 | 262 | # Essentia normalizes the positions to 1, they are converted here 263 | # to actual index values to be used in bins. 264 | peak_inds = np.array([int(round(bn * (len(self.bins) - 1))) 265 | for bn in peak_bins]) 266 | 267 | # if the object is pcd and there is a peak at zeroth index, 268 | # there will be another in the last index. Since a pcd is circular 269 | # remove the lower value 270 | if self.is_pcd() and peak_inds[0] == 0: 271 | if peak_vals[0] >= peak_vals[-1]: 272 | peak_inds = peak_inds[:-1] 273 | peak_vals = peak_vals[:-1] 274 | else: 275 | peak_inds = peak_inds[1:] 276 | peak_vals = peak_vals[1:] 277 | 278 | # remove peaks lower than the min_peak_ratio 279 | peak_bool = peak_vals / max(peak_vals) >= min_peak_ratio 280 | 281 | return peak_inds[peak_bool], peak_vals[peak_bool] 282 | 283 | def to_pcd(self): 284 | """-------------------------------------------------------------------- 285 | Given the pitch distribution of a recording, generates its pitch class 286 | distribution, by octave wrapping. 287 | ----------------------------------------------------------------------- 288 | pD: PitchDistribution object. Its attributes include everything we need 289 | --------------------------------------------------------------------""" 290 | assert not self.is_pcd(), 'The object is already a PCD' 291 | 292 | has_hz_bin = self.has_hz_bin() # remember the bin unit for later 293 | if self.has_hz_bin(): 294 | self.hz_to_cent(self.bins[0]) 295 | 296 | # Initializations 297 | pcd_bins = np.arange(0, 1200, self.step_size) 298 | pcd_vals = np.zeros(len(pcd_bins)) 299 | 300 | # Octave wrapping 301 | for bb, vv in zip(self.bins, self.vals): 302 | 303 | idx = int(round((bb % 1200) / self.step_size)) 304 | idx = idx if idx != 160 else 0 305 | pcd_vals[idx] += vv 306 | 307 | self.bins = pcd_bins 308 | self.vals = pcd_vals 309 | 310 | assert len(pcd_bins) == len(pcd_vals), 'Lengths of bins and vals ' \ 311 | 'are different.' 312 | 313 | # convert the unit back to what is was 314 | if has_hz_bin: 315 | self.cent_to_hz() 316 | 317 | def hz_to_cent(self, ref_freq): 318 | if self.has_hz_bin(): 319 | self.bins = Converter.hz_to_cent(self.bins, ref_freq) 320 | self.ref_freq = ref_freq 321 | 322 | # make sure all the bins stay between 0 - 1200 for PCDs 323 | if self.is_pcd(): 324 | self.bins = np.mod(self.bins, 1200) 325 | 326 | idx = np.argsort(self.bins) 327 | self.bins = self.bins[idx] 328 | self.vals = self.vals[idx] 329 | else: 330 | raise ValueError('The bin unit should be "hz".') 331 | 332 | def cent_to_hz(self): 333 | if self.has_cent_bin(): 334 | self.bins = Converter.cent_to_hz(self.bins, self.ref_freq) 335 | self.ref_freq = None 336 | else: 337 | raise ValueError('The bin unit should be "cent".') 338 | 339 | def shift(self, shift_idx): 340 | """-------------------------------------------------------------------- 341 | Shifts the distribution by the given number of samples 342 | ----------------------------------------------------------------------- 343 | shift_idx : The number of samples that the distribution is to be 344 | shifted 345 | --------------------------------------------------------------------""" 346 | # Shift only if the index is non-zero and the distribution is in 347 | # cent units 348 | if shift_idx and self.has_cent_bin(): 349 | # update reference frequency 350 | self.ref_freq = Converter.cent_to_hz( 351 | self.bins[shift_idx] - self.bins[0], 352 | ref_freq=self.ref_freq) 353 | 354 | # If distribution is a PCD, we do a circular shift 355 | if self.is_pcd(): 356 | self.vals = np.concatenate((self.vals[shift_idx:], 357 | self.vals[:shift_idx])) 358 | else: # If distribution is a PD, shift the bins. 359 | self.bins -= self.step_size * shift_idx 360 | 361 | def merge(self, distrib): 362 | """ 363 | Merges the distribution with another distribution 364 | :param distrib: input distribution (PD or PCD) 365 | """ 366 | assert self.bin_unit == distrib.bin_unit, \ 367 | 'The bin units of the compared distributions should match.' 368 | assert self.distrib_type() == distrib.distrib_type(), \ 369 | 'The features should be of the same type' 370 | assert self.step_size == distrib.step_size, \ 371 | 'The step_sizes should be the same' 372 | assert self.is_pdf() == distrib.is_pdf(), \ 373 | 'The normalization should be the same' 374 | 375 | # find the max and min bins 376 | min_bin = np.min([np.min(self.bins), np.min(distrib.bins[0])]) 377 | max_bin = np.max([np.max(self.bins[-1]), np.max(distrib.bins[-1])]) 378 | 379 | # initialize the bins and vals 380 | bins = np.arange(min_bin, max_bin + self.step_size / 2.0, 381 | self.step_size) 382 | assert 0 in bins, 'Zero should be in the bins' 383 | vals = np.zeros(len(bins)) 384 | 385 | # add the vals in the distributions to the corresponding bins 386 | for dd in (self, distrib): 387 | bin_bool = np.logical_and(bins >= np.min(dd.bins), 388 | bins <= np.max(dd.bins)) 389 | vals[bin_bool] += dd.vals 390 | 391 | # update self 392 | is_pdf = self.is_pdf() # record if pdf 393 | self.bins = bins 394 | self.vals = vals 395 | if is_pdf: 396 | self.normalize() 397 | 398 | def plot(self): 399 | plt.plot(self.bins, self.vals) 400 | self.label_figure() 401 | 402 | def bar(self): 403 | bars = plt.bar(self.bins, self.vals, width=self.step_size, 404 | align='center') 405 | self.label_figure() 406 | 407 | return bars 408 | 409 | def label_figure(self): 410 | if self.is_pcd(): 411 | plt.title('Pitch class distribution') 412 | ref_freq_str = 'Hz x 2^n' 413 | else: 414 | plt.title('Pitch distribution') 415 | ref_freq_str = 'Hz' 416 | if self.has_hz_bin(): 417 | plt.xlabel('Frequency (Hz)') 418 | else: 419 | plt.xlabel('Normalized Frequency (cents), ref = {0}{1}'.format( 420 | str(self.ref_freq), ref_freq_str)) 421 | plt.ylabel('Occurence') 422 | 423 | @staticmethod 424 | def from_pickle(input_str): 425 | try: # file given 426 | return pickle.load(open(input_str, 'rb')) 427 | except IOError: # string given 428 | return pickle.loads(input_str, 'rb') 429 | 430 | def to_pickle(self, file_name=None): 431 | if file_name is None: 432 | return pickle.dumps(self) 433 | else: 434 | pickle.dump(self, open(file_name, 'wb')) 435 | 436 | @staticmethod 437 | def from_json(file_name): 438 | """-------------------------------------------------------------------- 439 | Loads a PitchDistribution object from JSON file. 440 | ----------------------------------------------------------------------- 441 | file_name : The filename of the JSON file 442 | -------------------------------------------------------------------- 443 | """ 444 | try: 445 | distrib = json.load(open(file_name, 'r')) 446 | except IOError: # json string 447 | distrib = json.loads(file_name) 448 | 449 | distrib = distrib if isinstance(distrib, dict) else distrib[0] 450 | 451 | return PitchDistribution.from_dict(distrib) 452 | 453 | def to_json(self, file_name=None): 454 | """-------------------------------------------------------------------- 455 | Saves the PitchDistribution object to a JSON file. 456 | ----------------------------------------------------------------------- 457 | file_name : The file path of the JSON file to be created. 458 | --------------------------------------------------------------------""" 459 | dist_json = self.to_dict() 460 | 461 | if file_name is None: 462 | return json.dumps(dist_json, indent=4) 463 | else: 464 | json.dump(dist_json, open(file_name, 'w'), indent=4) 465 | 466 | @staticmethod 467 | def from_dict(distrib_dict): 468 | return PitchDistribution(distrib_dict['bins'], distrib_dict['vals'], 469 | kernel_width=distrib_dict['kernel_width'], 470 | ref_freq=distrib_dict['ref_freq']) 471 | 472 | def to_dict(self): 473 | pdict = self.__dict__ 474 | for key in pdict.keys(): 475 | try: 476 | # convert to list from np array 477 | pdict[key] = pdict[key].tolist() 478 | except AttributeError: 479 | pass 480 | 481 | return pdict 482 | -------------------------------------------------------------------------------- /notebooks/external_utilities/predominantmelodymakam.py: -------------------------------------------------------------------------------- 1 | # -*- coding: utf-8 -*- 2 | # Copyright 2014 - 2017 Music Technology Group - Universitat Pompeu Fabra 3 | # 4 | # This program is distributed in the hope that it will be useful, 5 | # but WITHOUT ANY ARRANTY; without even the implied warranty of 6 | # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Affero 7 | # General Public License v3.0 for more details. 8 | # 9 | # You should have received a copy of the GNU Affero General Public License v3.0 10 | # along with this program. If not, see http://www.gnu.org/licenses/ 11 | # 12 | # If you are using this extractor please cite the following paper: 13 | # 14 | # Atlı, H. S., Uyar, B., Şentürk, S., Bozkurt, B., and Serra, X. (2014). Audio 15 | # feature extraction for exploring Turkish makam music. In Proceedings of 3rd 16 | # International Conference on Audio Technologies for Music and Media, Ankara, 17 | # Turkey. 18 | 19 | from essentia import Pool 20 | from essentia import array as e_array 21 | import essentia.standard as estd 22 | from math import ceil 23 | import numpy as np 24 | import warnings 25 | 26 | def xrange(k,m): 27 | return range(k,m) 28 | 29 | 30 | class PredominantMelodyMakam(object): 31 | def __init__(self, hop_size=128, frame_size=2048, bin_resolution=1.0, 32 | min_frequency=55, max_frequency=1760, magnitude_threshold=0, 33 | peak_distribution_threshold=1.4, filter_pitch=True, 34 | confidence_threshold=36, min_chunk_size=50): 35 | 36 | self.hop_size = hop_size # default hopSize of PredominantMelody 37 | self.frame_size = frame_size # default frameSize of PredominantMelody 38 | self.bin_resolution = bin_resolution # ~1/3 Hc; recommended for makams 39 | self.min_frequency = min_frequency # default: minimum of 40 | # PitchSalienceFunction 41 | self.max_frequency = max_frequency # default: maximum of 42 | # PitchSalienceFunction 43 | self.magnitude_threshold = magnitude_threshold # default of 44 | # SpectralPeaks; 0 dB? 45 | self.peak_distribution_threshold = peak_distribution_threshold 46 | # default in PitchContours is 0.9; we need higher in makams 1.4 47 | self.filter_pitch = filter_pitch # call PitchFilter 48 | self.confidence_threshold = confidence_threshold # default 49 | # confidenceThreshold for pitchFilter 50 | self.min_chunk_size = min_chunk_size # number of minimum allowed 51 | # samples of a chunk in PitchFilter; ~145 ms with 52 | # 128 sample hopSize & 44100 Fs 53 | 54 | self.sample_rate = 44100 55 | 56 | def get_settings(self): 57 | from essentia import __version__ as essentia_version 58 | citation = u"Atlı, H. S., Uyar, B., Şentürk, S., Bozkurt, B., " \ 59 | u"and Serra, X. (2014). Audio feature extraction for " \ 60 | u"exploring Turkish makam music. In Proceedings of 3rd " \ 61 | u"International Conference on Audio Technologies for " \ 62 | u"Music and Media, Ankara, Turkey." 63 | 64 | return {'hopSize': self.hop_size, 'frameSize': self.frame_size, 65 | 'pitchUnit': 'Hz', 'binResolution': self.bin_resolution, 66 | 'minFrequency': self.min_frequency, 67 | 'maxFrequency': self.max_frequency, 68 | 'magnitudeThreshold': self.magnitude_threshold, 69 | 'peakDistributionThreshold': self.peak_distribution_threshold, 70 | 'filterPitch': self.filter_pitch, 71 | 'confidenceThreshold': self.confidence_threshold, 72 | 'sampleRate': self.sample_rate, 73 | 'minChunkSize': self.min_chunk_size, 74 | 'essentiaVersion': essentia_version, 75 | 'citation': citation} 76 | 77 | def run(self, fname): 78 | # load audio and eqLoudness 79 | # Note: MonoLoader resamples the audio signal to 44100 Hz by default 80 | audio = estd.MonoLoader(filename=fname)() 81 | audio = estd.EqualLoudness()(audio) 82 | 83 | contours_bins, contours_start_times, contour_saliences, duration = \ 84 | self._extract_pitch_contours(audio) 85 | 86 | # run the simplified contour selection 87 | [pitch, pitch_salience] = self.select_contours( 88 | contours_bins, contour_saliences, contours_start_times, duration) 89 | 90 | # cent to Hz conversion 91 | pitch = [0. if p == 0 92 | else 55. * 2. ** (self.bin_resolution * p / 1200.) 93 | for p in pitch] 94 | pitch = e_array(pitch) 95 | pitch_salience = e_array(pitch_salience) 96 | 97 | # pitch filter 98 | if self.filter_pitch: 99 | pitch, pitch_salience = self._post_filter_pitch( 100 | pitch, pitch_salience) 101 | 102 | # generate time stamps 103 | time_stamps = self._gen_time_stamps(0, len(pitch)) 104 | 105 | # [time pitch salience] matrix 106 | out = np.transpose( 107 | np.vstack((time_stamps, pitch.tolist(), pitch_salience.tolist()))) 108 | out = out.tolist() 109 | 110 | # settings 111 | settings = self.get_settings() 112 | settings.update({'source': fname}) 113 | 114 | return {'pitch': out, 'settings': settings} 115 | 116 | def extract(self, fname): 117 | """ 118 | Alias of self.run 119 | :param fname: filename 120 | :return: dictionary with 'pitch' and 'settings' keys 121 | """ 122 | return self.run(fname) 123 | 124 | def _extract_pitch_contours(self, audio): 125 | # Hann window with x4 zero padding 126 | run_windowing = estd.Windowing(zeroPadding=3 * self.frame_size) 127 | run_spectrum = estd.Spectrum(size=self.frame_size * 4) 128 | run_spectral_peaks = estd.SpectralPeaks( 129 | minFrequency=self.min_frequency, maxFrequency=self.max_frequency, 130 | magnitudeThreshold=self.magnitude_threshold, 131 | sampleRate=self.sample_rate, orderBy='magnitude') 132 | 133 | # convert unit to cents, PitchSalienceFunction takes 55 Hz as the 134 | # default reference 135 | run_pitch_salience_function = estd.PitchSalienceFunction( 136 | binResolution=self.bin_resolution) 137 | run_pitch_salience_function_peaks = estd.PitchSalienceFunctionPeaks( 138 | binResolution=self.bin_resolution, minFrequency=self.min_frequency, 139 | maxFrequency=self.max_frequency) 140 | run_pitch_contours = estd.PitchContours( 141 | hopSize=self.hop_size, binResolution=self.bin_resolution, 142 | peakDistributionThreshold=self.peak_distribution_threshold) 143 | 144 | # compute frame by frame 145 | pool = Pool() 146 | for frame in estd.FrameGenerator(audio, frameSize=self.frame_size, 147 | hopSize=self.hop_size): 148 | frame = run_windowing(frame) 149 | spectrum = run_spectrum(frame) 150 | peak_frequencies, peak_magnitudes = run_spectral_peaks(spectrum) 151 | salience = run_pitch_salience_function(peak_frequencies, 152 | peak_magnitudes) 153 | salience_peaks_bins, salience_peaks_contour_saliences = \ 154 | run_pitch_salience_function_peaks(salience) 155 | if not np.size(salience_peaks_bins): 156 | salience_peaks_bins = np.array([0]) 157 | if not np.size(salience_peaks_contour_saliences): 158 | salience_peaks_contour_saliences = np.array([0]) 159 | 160 | pool.add('allframes_salience_peaks_bins', salience_peaks_bins) 161 | pool.add('allframes_salience_peaks_contourSaliences', 162 | salience_peaks_contour_saliences) 163 | 164 | # post-processing: contour tracking 165 | contours_bins, contour_saliences, contours_start_times, duration = \ 166 | run_pitch_contours( 167 | [f.tolist() for f in pool['allframes_salience_peaks_bins']], 168 | [f.tolist() for f in pool['allframes_salience_peaks_contourSaliences']]) 169 | return contours_bins, contours_start_times, contour_saliences, duration 170 | 171 | def _post_filter_pitch(self, pitch, pitch_salience): 172 | try: 173 | run_pitch_filter = estd.PitchFilter( 174 | confidenceThreshold=self.confidence_threshold, 175 | minChunkSize=self.min_chunk_size) 176 | pitch = run_pitch_filter(pitch, pitch_salience) 177 | 178 | except AttributeError: # fall back to python implementation 179 | from pitchfilter.pitchfilter import PitchFilter 180 | run_pitch_filter = PitchFilter() 181 | 182 | # generate time stamps 183 | time_stamps = self._gen_time_stamps(0, len(pitch)) 184 | 185 | temp_pitch = np.vstack(( 186 | time_stamps, pitch, pitch_salience)).transpose() 187 | 188 | temp_pitch = run_pitch_filter.run(temp_pitch) 189 | 190 | pitch = temp_pitch[:, 1] 191 | pitch_salience = temp_pitch[:, 2] 192 | 193 | return pitch, pitch_salience 194 | 195 | def _gen_time_stamps(self, start_samp, end_samp): 196 | time_stamps = [s * self.hop_size / float( 197 | self.sample_rate) for s in xrange(start_samp, end_samp)] 198 | return time_stamps 199 | 200 | def select_contours(self, pitch_contours, contour_saliences, start_times, 201 | duration): 202 | sample_rate = self.sample_rate 203 | 204 | hop_size = self.hop_size 205 | 206 | # number in samples in the audio 207 | num_samples = int(ceil((duration * sample_rate) / hop_size)) 208 | 209 | # Start points of the contours in samples 210 | start_samples = [ 211 | int(round(start_times[i] * sample_rate / float(hop_size))) 212 | for i in xrange(0, len(start_times))] 213 | 214 | pitch_contours_no_overlap = [] 215 | start_samples_no_overlap = [] 216 | contour_saliences_no_overlap = [] 217 | lens_no_overlap = [] 218 | try: 219 | # the pitch contours is a list of numpy arrays, parse them starting 220 | # with the longest contour 221 | while pitch_contours: # terminate when all the contours are 222 | # checked 223 | # print len(pitchContours) 224 | 225 | # get the lengths of the pitchContours 226 | lens = [len(k) for k in pitch_contours] 227 | 228 | # find the longest pitch contour 229 | long_idx = lens.index(max(lens)) 230 | 231 | # pop the lists related to the longest pitchContour and append 232 | # it to the new list 233 | pitch_contours_no_overlap.append(pitch_contours.pop(long_idx)) 234 | contour_saliences_no_overlap.append( 235 | contour_saliences.pop(long_idx)) 236 | start_samples_no_overlap.append(start_samples.pop(long_idx)) 237 | lens_no_overlap.append(lens.pop(long_idx)) 238 | 239 | # accumulate the filled samples 240 | acc_idx = range(start_samples_no_overlap[-1], 241 | start_samples_no_overlap[-1] + 242 | lens_no_overlap[-1]) 243 | 244 | # remove overlaps 245 | [start_samples, pitch_contours, contour_saliences] = \ 246 | self._remove_overlaps(start_samples, pitch_contours, 247 | contour_saliences, lens, acc_idx) 248 | except ValueError: 249 | # if the audio input is very short such that Essentia returns a 250 | # single contour as a numpy array (of length 1) of numpy array 251 | # (of length 1). In this case the while loop fails directly 252 | # as it tries to check all the truth value of an all pitch values, 253 | # instead of checking whether the list is empty or not. 254 | # Here we handle the error in a Pythonic way by simply breaking the 255 | # loop and assigning the inputs to outputs since a single contour 256 | # means nothing to filter 257 | pitch_contours_no_overlap = pitch_contours 258 | contour_saliences_no_overlap = contour_saliences 259 | start_samples_no_overlap = start_samples 260 | 261 | pitch, salience = self._join_contours(pitch_contours_no_overlap, 262 | contour_saliences_no_overlap, 263 | start_samples_no_overlap, 264 | num_samples) 265 | 266 | return pitch, salience 267 | 268 | @staticmethod 269 | def _join_contours(pitch_contours_no_overlap, contour_saliences_no_overlap, 270 | start_samples_no_overlap, num_samples): 271 | # accumulate pitch and salience 272 | pitch = np.array([0.] * num_samples) 273 | salience = np.array([0.] * num_samples) 274 | for i in xrange(0, len(pitch_contours_no_overlap)): 275 | start_samp = start_samples_no_overlap[i] 276 | end_samp = start_samples_no_overlap[i] + len( 277 | pitch_contours_no_overlap[i]) 278 | 279 | try: 280 | pitch[start_samp:end_samp] = pitch_contours_no_overlap[i] 281 | salience[start_samp:end_samp] = contour_saliences_no_overlap[i] 282 | except ValueError: 283 | warnings.warn("The last pitch contour exceeds the audio " 284 | "length. Trimming...") 285 | 286 | pitch[start_samp:] = pitch_contours_no_overlap[i][:len( 287 | pitch) - start_samp] 288 | salience[start_samp:] = contour_saliences_no_overlap[i][:len( 289 | salience) - start_samp] 290 | return pitch, salience 291 | 292 | @staticmethod 293 | def _remove_overlaps(start_samples, pitch_contours, contour_saliences, 294 | lens, acc_idx): 295 | # remove overlaps 296 | rmv_idx = [] 297 | for i in xrange(0, len(start_samples)): 298 | # print '_' + str(i) 299 | # create the sample index vector for the checked pitch contour 300 | curr_samp_idx = range(start_samples[i], start_samples[i] + lens[i]) 301 | 302 | # get the non-overlapping samples 303 | curr_samp_idx_no_overlap = list(set(curr_samp_idx) - 304 | set(acc_idx)) 305 | 306 | if curr_samp_idx_no_overlap: 307 | temp = min(curr_samp_idx_no_overlap) 308 | keep_idx = range(temp - start_samples[i], 309 | (max(curr_samp_idx_no_overlap) - 310 | start_samples[i]) + 1) 311 | 312 | # remove all overlapping values 313 | pitch_contours[i] = np.array(pitch_contours[i])[keep_idx] 314 | contour_saliences[i] = np.array(contour_saliences[i])[keep_idx] 315 | # update the startSample 316 | start_samples[i] = temp 317 | else: # totally overlapping 318 | rmv_idx.append(i) 319 | 320 | # remove totally overlapping pitch contours 321 | rmv_idx = sorted(rmv_idx, reverse=True) 322 | for r in rmv_idx: 323 | pitch_contours.pop(r) 324 | contour_saliences.pop(r) 325 | start_samples.pop(r) 326 | 327 | return start_samples, pitch_contours, contour_saliences 328 | -------------------------------------------------------------------------------- /notebooks/external_utilities/readme.txt: -------------------------------------------------------------------------------- 1 | Codes in this folder come from two other MTG repositories: 2 | 3 | predominantmelodymakam.py: 4 | https://github.com/sertansenturk/tomato 5 | 6 | converter.py, pitchdistribution.py: 7 | https://github.com/altugkarakurt/morty -------------------------------------------------------------------------------- /notebooks/formExpSubsets4ModeRecognition.ipynb: -------------------------------------------------------------------------------- 1 | { 2 | "cells": [ 3 | { 4 | "cell_type": "markdown", 5 | "metadata": {}, 6 | "source": [ 7 | "## Forming data subsets for mode and rhythm mode recognition experiments\n", 8 | "\n", 9 | "This notebook targets forming data subsets for mode and rhythm mode recognition experiments starting from the list of files (and meta data info) created and stored (using generateFileLists4Collections.ipynb) in a pickle file. For each recording the following information is available:\n", 10 | "- Files available for that recording\n", 11 | "- MusicBrainz id (mbid)\n", 12 | "- Mode information (raga, makam, etc)\n", 13 | "- Rhythm mode information (tala, usul, etc)\n", 14 | "\n", 15 | "This notebook reads this file and forms the subsets by grouping recordings with respect to mode or rhythm mode while also checking available files (ex: tonic annotation) for the recording. The outputs are json files for each culture (with the format of [this sample file](https://github.com/MTG/otmm_makam_recognition_dataset/blob/master/annotations.json)) which can be used in mode recognition implementations as in [this repo](https://github.com/emirdemirel/Supervised_Mode_Recognition)." 16 | ] 17 | }, 18 | { 19 | "cell_type": "code", 20 | "execution_count": null, 21 | "metadata": {}, 22 | "outputs": [], 23 | "source": [ 24 | "# Set your token here from https://dunya.compmusic.upf.edu/user/profile/\n", 25 | "token = '...yourAPITokenGoesHere...'" 26 | ] 27 | }, 28 | { 29 | "cell_type": "code", 30 | "execution_count": null, 31 | "metadata": {}, 32 | "outputs": [], 33 | "source": [ 34 | "import codecs\n", 35 | "import json, os, sys\n", 36 | "import numpy as np\n", 37 | "import pickle\n", 38 | "import csv\n", 39 | "import time\n", 40 | "import datetime\n", 41 | "import random\n", 42 | "from compmusic.dunya import docserver as ds\n", 43 | "from compmusic import dunya as dn\n", 44 | "from compmusic.dunya import conn\n", 45 | "import collections\n", 46 | "\n", 47 | "dn.set_token(token)#setting the token\n", 48 | "\n", 49 | "# Read metadata from the previous notebook\n", 50 | "with open(\"metaData_collections.pkl\", 'rb') as f:\n", 51 | " metaData = pickle.load(f)" 52 | ] 53 | }, 54 | { 55 | "cell_type": "markdown", 56 | "metadata": {}, 57 | "source": [ 58 | "## Most frequently used modes \n", 59 | "Modes in each collection ordered by the number of recordings that we have for each" 60 | ] 61 | }, 62 | { 63 | "cell_type": "code", 64 | "execution_count": null, 65 | "metadata": {}, 66 | "outputs": [], 67 | "source": [ 68 | "numModes = 20\n", 69 | "\n", 70 | "for collection, recordings in metaData.items():\n", 71 | " mode_counter = collections.Counter()\n", 72 | " for recording in recordings:\n", 73 | " if 'mode' in recording:\n", 74 | " mode_counter[recording['mode']] += 1\n", 75 | " print('Most frequently used modes in collection {}'.format(collection))\n", 76 | " common_modes = mode_counter.most_common(numModes)\n", 77 | " max_length = max([len(m) for m in dict(common_modes).keys()])\n", 78 | " for mode, count in common_modes:\n", 79 | " print('{mode:<{pad}} {count}'.format(mode=mode, pad=max_length, count=count))\n", 80 | " print('-'*50)" 81 | ] 82 | }, 83 | { 84 | "cell_type": "markdown", 85 | "metadata": {}, 86 | "source": [ 87 | "### Composing a mode recognition datasets for all collections\n", 88 | "\n", 89 | "Creating annotations.json file for each culture that can serve as an experimental dataset. We collect the tonic frequency for a random selection of recordings in each mode. These json files can be used as input to supervised mode recognition tests in [this repo](https://github.com/emirdemirel/Supervised_Mode_Recognition) " 90 | ] 91 | }, 92 | { 93 | "cell_type": "code", 94 | "execution_count": null, 95 | "metadata": {}, 96 | "outputs": [], 97 | "source": [ 98 | "# Take the top `numModes` modes. Randomly select recordings of this mode from the collection\n", 99 | "# until we have at least `numFilesPerMode` downloads for each mode.\n", 100 | "\n", 101 | "def download_tonic(mbid, collection):\n", 102 | " \"\"\"Retrieve the tonic value for a given recording\n", 103 | " \n", 104 | " Arguments:\n", 105 | " mbid: the recording MBID to retrieve\n", 106 | " collection: the name of the collection that this MBID comes from\n", 107 | " (used to choose the download method)\n", 108 | " \n", 109 | " Returns: The tonic of the recording, or None of this recording has no tonic computed\n", 110 | " \"\"\"\n", 111 | " try:\n", 112 | " if collection == 'makam': \n", 113 | " content = ds.get_document_as_json(mbid, 'audioanalysis', 'tonic')\n", 114 | " tonic = None\n", 115 | " if content:\n", 116 | " tonic = content['value']\n", 117 | " elif collection == 'carnatic' or collection == 'hindustani':\n", 118 | " content = ds.file_for_document(recording['mbid'], 'ctonic', 'tonic')\n", 119 | " tonic = content.decode()\n", 120 | " return tonic\n", 121 | " except dn.HTTPError as e:\n", 122 | " if e.args[0].response.status_code != 404:\n", 123 | " raise\n", 124 | "\n", 125 | "def get_tonics_for_recordings(recordings, collection, numModes, numFilesPerMode):\n", 126 | " # Count the modes in the recording list and group recordings by their mode\n", 127 | " mode_counter = collections.Counter()\n", 128 | " mode_recordings = collections.defaultdict(list)\n", 129 | " for recording in recordings:\n", 130 | " if 'mode' in recording:\n", 131 | " mode_counter[recording['mode']] += 1\n", 132 | " mode_recordings[recording['mode']].append(recording)\n", 133 | " selected_modes = dict(mode_counter.most_common(numModes)).keys()\n", 134 | " # for each mode, download tonic for `numFilesPerMode` random recordings\n", 135 | " collection_sample = []\n", 136 | " for mode in selected_modes:\n", 137 | " recordings = mode_recordings[mode]\n", 138 | " num_recordings = 0\n", 139 | " for recording in recordings:\n", 140 | " if num_recordings >= numFilesPerMode:\n", 141 | " break\n", 142 | " tonic = download_tonic(recording['mbid'], collection)\n", 143 | " # Some recordings may not have a tonic, only add those for which we do\n", 144 | " if tonic:\n", 145 | " recording['tonic'] = tonic\n", 146 | " collection_sample.append(recording)\n", 147 | " num_recordings += 1\n", 148 | " return collection_sample" 149 | ] 150 | }, 151 | { 152 | "cell_type": "code", 153 | "execution_count": null, 154 | "metadata": {}, 155 | "outputs": [], 156 | "source": [ 157 | "numModes = 10\n", 158 | "numFilesPerMode = 20\n", 159 | "\n", 160 | "tonics = {}\n", 161 | "\n", 162 | "for collection, recordings in metaData.items():\n", 163 | " print('Downloading Tonic values for collection {}'.format(collection))\n", 164 | " \n", 165 | " collection_sample = get_tonics_for_recordings(recordings, collection, numModes, numFilesPerMode)\n", 166 | "\n", 167 | " tonics[collection] = collection_sample" 168 | ] 169 | }, 170 | { 171 | "cell_type": "code", 172 | "execution_count": null, 173 | "metadata": { 174 | "scrolled": true 175 | }, 176 | "outputs": [], 177 | "source": [ 178 | "# Write tonic data to file\n", 179 | "for collection, recordings in tonics.items():\n", 180 | " with open('annotations_{}.json'.format(collection), 'w') as f:\n", 181 | " json.dump(recordings, f)" 182 | ] 183 | } 184 | ], 185 | "metadata": { 186 | "kernelspec": { 187 | "display_name": "Python 3", 188 | "language": "python", 189 | "name": "python3" 190 | }, 191 | "language_info": { 192 | "codemirror_mode": { 193 | "name": "ipython", 194 | "version": 3 195 | }, 196 | "file_extension": ".py", 197 | "mimetype": "text/x-python", 198 | "name": "python", 199 | "nbconvert_exporter": "python", 200 | "pygments_lexer": "ipython3", 201 | "version": "3.5.2" 202 | } 203 | }, 204 | "nbformat": 4, 205 | "nbformat_minor": 2 206 | } 207 | -------------------------------------------------------------------------------- /notebooks/generateFileLists4Collections.ipynb: -------------------------------------------------------------------------------- 1 | { 2 | "cells": [ 3 | { 4 | "cell_type": "markdown", 5 | "metadata": {}, 6 | "source": [ 7 | "## Generating file lists for various Dunya collections (for mode and rhythm mode recognition experiments)\n", 8 | "\n", 9 | "This notebook targets accessing Dunya data and collecting file lists for several collections.\n", 10 | "\n", 11 | "The main aim is to create subsets of data for mode and rhythm mode recognition experiments and the process is split into two notebooks. In this first notebook, we create a list of recordings and relevant metadata. For each recording the following information is included:\n", 12 | "- Files available for that recording\n", 13 | "- MusicBrainz id (mbid)\n", 14 | "- Mode information (raga, makam, etc)\n", 15 | "- Rhythm mode information (tala, usul, etc)\n", 16 | "\n", 17 | "Then, the second notebook (formExpSubsets4ModeRecognition.ipynb) reads this file and forms the subsets by grouping recordings with respect to mode or rhythm mode while also checking available files (ex: tonic annotation) for the recording" 18 | ] 19 | }, 20 | { 21 | "cell_type": "code", 22 | "execution_count": null, 23 | "metadata": {}, 24 | "outputs": [], 25 | "source": [ 26 | "# Set your token here from https://dunya.compmusic.upf.edu/user/profile/\n", 27 | "token = '...yourAPITokenGoesHere...'" 28 | ] 29 | }, 30 | { 31 | "cell_type": "code", 32 | "execution_count": null, 33 | "metadata": {}, 34 | "outputs": [], 35 | "source": [ 36 | "import codecs\n", 37 | "import json, os, sys\n", 38 | "import pickle\n", 39 | "import csv\n", 40 | "import time\n", 41 | "import datetime\n", 42 | "import collections\n", 43 | "\n", 44 | "import numpy as np\n", 45 | "\n", 46 | "import compmusic\n", 47 | "from compmusic import dunya as dn\n", 48 | "from compmusic.dunya import hindustani as hi\n", 49 | "from compmusic.dunya import carnatic as ca\n", 50 | "from compmusic.dunya import makam as ma\n", 51 | "from compmusic.dunya import docserver as ds\n", 52 | "from compmusic import musicbrainz\n", 53 | "from compmusic.dunya import conn\n", 54 | "\n", 55 | "dn.set_token(token)" 56 | ] 57 | }, 58 | { 59 | "cell_type": "markdown", 60 | "metadata": {}, 61 | "source": [ 62 | "### Collecting files of three collections: Carnatic, Hindustani and Makam\n", 63 | "\n", 64 | "In Dunya, data is stored according to a model specific to each culture. For cross-cultural studies (such as testing of a mode recognition algorithm for all Dunya collections), one needs to access all collections in some unified way. We access data from each culture collection and arrange it in a consistent format for further analysis. Further this list can be processed to create data subsets for automatic recognition experiments." 65 | ] 66 | }, 67 | { 68 | "cell_type": "code", 69 | "execution_count": null, 70 | "metadata": {}, 71 | "outputs": [], 72 | "source": [ 73 | "# Set to None to get all files\n", 74 | "maxNumFiles = None " 75 | ] 76 | }, 77 | { 78 | "cell_type": "markdown", 79 | "metadata": {}, 80 | "source": [ 81 | "These method get only the mode information from each collection that we require for this analysis. We rename the attributes to be consistent for all collections.\n", 82 | "We only consider the first values for each of these fields, in the case that several modes are available, you may like to alter the code to check all modes and treat those having more than one distinct mode in a different way " 83 | ] 84 | }, 85 | { 86 | "cell_type": "code", 87 | "execution_count": null, 88 | "metadata": {}, 89 | "outputs": [], 90 | "source": [ 91 | "def get_carnatic_metadata(maxNumFiles=None):\n", 92 | " \"\"\" Get Carnatic specific mode and rhythmic mode metadata for all recordings.\"\"\"\n", 93 | " carnatic_recordings = ca.get_recordings(recording_detail=True)\n", 94 | " if maxNumFiles:\n", 95 | " carnatic_recordings = carnatic_recordings[:maxNumFiles]\n", 96 | " \n", 97 | " # Get only the information that we want for each collection. Rename the attributes to be\n", 98 | " # consistent for all collections.\n", 99 | " # Carnatic\n", 100 | " # mode -> raaga, rhythmMode -> taala\n", 101 | " carnatic_metadata = []\n", 102 | " for r in carnatic_recordings:\n", 103 | " if r['raaga'] or r['taala']:\n", 104 | " data = {'mbid': r['mbid']}\n", 105 | " if r['raaga']:\n", 106 | " data['mode'] = r['raaga'][0]['common_name']\n", 107 | " if r['taala']:\n", 108 | " data['rhythmMode'] = r['taala'][0]['common_name']\n", 109 | " carnatic_metadata.append(data)\n", 110 | " return carnatic_metadata\n", 111 | "\n", 112 | "def get_hindustani_metadata(maxNumFiles=None):\n", 113 | " \"\"\" Get Hindustani specific mode and rhythmic mode metadata for all recordings.\"\"\"\n", 114 | " hindustani_recordings = hi.get_recordings(recording_detail=True)\n", 115 | " if maxNumFiles:\n", 116 | " hindustani_recordings = hindustani_recordings[:maxNumFiles]\n", 117 | " \n", 118 | " # Get only the information that we want for each collection. Rename the attributes to be\n", 119 | " # consistent for all collections.\n", 120 | " # Hindustani\n", 121 | " # mode -> raag, rhythmMode -> taal\n", 122 | " # The API for hindustani returns some MBIDs twice, we do a basic filtering here.\n", 123 | " seen_mbids = set()\n", 124 | " hindustani_metadata = []\n", 125 | " for r in hindustani_recordings:\n", 126 | " if r['raags'] or r['taals']:\n", 127 | " data = {'mbid': r['mbid']}\n", 128 | " if r['raags']:\n", 129 | " data['mode'] = r['raags'][0]['common_name']\n", 130 | " if r['taals']:\n", 131 | " data['rhythmMode'] = r['taals'][0]['common_name']\n", 132 | " if r['mbid'] not in seen_mbids:\n", 133 | " hindustani_metadata.append(data)\n", 134 | " seen_mbids.add(r['mbid'])\n", 135 | " return hindustani_metadata\n", 136 | "\n", 137 | "def get_makam_metadata(maxNumFiles=None):\n", 138 | " \"\"\" Get Turkish-makam specific mode and rhythmic mode metadata for all recordings.\"\"\"\n", 139 | " makam_recordings = ma.get_recordings(recording_detail=True)\n", 140 | " if maxNumFiles:\n", 141 | " makam_recordings = makam_recordings[:maxNumFiles]\n", 142 | " # Get only the information that we want for each collection. Rename the attributes to be\n", 143 | " # consistent for all collections.\n", 144 | " # Makam\n", 145 | " # mode -> makam, rhythmMode -> usul\n", 146 | " makam_metadata = []\n", 147 | " for r in makam_recordings:\n", 148 | " if r['makamlist'] or r['usullist']:\n", 149 | " data = {'mbid': r['mbid']}\n", 150 | " if r['makamlist']:\n", 151 | " data['mode'] = r['makamlist'][0]['name']\n", 152 | " if r['usullist']:\n", 153 | " data['rhythmMode'] = r['usullist'][0]['name']\n", 154 | " makam_metadata.append(data)\n", 155 | " return makam_metadata" 156 | ] 157 | }, 158 | { 159 | "cell_type": "markdown", 160 | "metadata": {}, 161 | "source": [ 162 | "This next step may take some time - these methods retrieve detailed information for all recordings in each collection, which requires a number of webservice requests" 163 | ] 164 | }, 165 | { 166 | "cell_type": "code", 167 | "execution_count": null, 168 | "metadata": {}, 169 | "outputs": [], 170 | "source": [ 171 | "print('Process start time: {}'.format(datetime.datetime.now()))\n", 172 | "print('Starting Hindustani: {}'.format(datetime.datetime.now()))\n", 173 | "hindustani_metadata = get_hindustani_metadata(maxNumFiles)\n", 174 | "print('Starting Carnatic: {}'.format(datetime.datetime.now()))\n", 175 | "carnatic_metadata = get_carnatic_metadata(maxNumFiles)\n", 176 | "print('Starting Makam: {}'.format(datetime.datetime.now()))\n", 177 | "makam_metadata = get_makam_metadata(maxNumFiles)\n", 178 | "\n", 179 | "metaData_collections = {'hindustani': hindustani_metadata,\n", 180 | " 'carnatic': carnatic_metadata,\n", 181 | " 'makam': makam_metadata}\n", 182 | "\n", 183 | "print('Process end time: {}'.format(datetime.datetime.now()))\n", 184 | "\n", 185 | "# Save data to file\n", 186 | "pickle.dump(metaData_collections, open('metaData_collections.pkl', 'wb'))" 187 | ] 188 | } 189 | ], 190 | "metadata": { 191 | "kernelspec": { 192 | "display_name": "Python 3", 193 | "language": "python", 194 | "name": "python3" 195 | }, 196 | "language_info": { 197 | "codemirror_mode": { 198 | "name": "ipython", 199 | "version": 3 200 | }, 201 | "file_extension": ".py", 202 | "mimetype": "text/x-python", 203 | "name": "python", 204 | "nbconvert_exporter": "python", 205 | "pygments_lexer": "ipython3", 206 | "version": "3.5.2" 207 | } 208 | }, 209 | "nbformat": 4, 210 | "nbformat_minor": 2 211 | } 212 | -------------------------------------------------------------------------------- /notebooks/symbolicDataPro_symbTr.ipynb: -------------------------------------------------------------------------------- 1 | { 2 | "cells": [ 3 | { 4 | "cell_type": "markdown", 5 | "metadata": {}, 6 | "source": [ 7 | "## Symbolic data processing, Turkish music case\n", 8 | "\n", 9 | "In this notebook, the SymTr data (Turkish Makam Music Symbolic Data Collection) (http://compmusic.upf.edu/node/140 , https://github.com/MTG/SymbTr) is used. Here we demonstrate accessing pieces in a specific form and makam, reading a specific a section of the form and ploting the melodic curves. The makams are chosen to apply the same scale but different melodic progressions ('seyir'). \n", 10 | "\n", 11 | "\n", 12 | "The following steps are carried:\n", 13 | "- Downloading SymbTr data from the github repo\n", 14 | "- Printing a list of makams and forms sorted with respect to the number of files in those categories\n", 15 | "- Plotting melodic curves from the first sections in saz-semaisi form in two makams: rast and mahur.\n" 16 | ] 17 | }, 18 | { 19 | "cell_type": "code", 20 | "execution_count": 1, 21 | "metadata": {}, 22 | "outputs": [], 23 | "source": [ 24 | "#Imports\n", 25 | "import urllib.request\n", 26 | "import zipfile\n", 27 | "import os, sys,shutil\n", 28 | "import numpy as np\n", 29 | "import matplotlib.pyplot as plt" 30 | ] 31 | }, 32 | { 33 | "cell_type": "markdown", 34 | "metadata": {}, 35 | "source": [ 36 | "### Downloading SymbTr data" 37 | ] 38 | }, 39 | { 40 | "cell_type": "code", 41 | "execution_count": 2, 42 | "metadata": {}, 43 | "outputs": [ 44 | { 45 | "name": "stdout", 46 | "output_type": "stream", 47 | "text": [ 48 | "Data downloaded and unzipped to: ../data/compMusicDatasets/turkishMakam/SymbTr-2.0.0\n" 49 | ] 50 | } 51 | ], 52 | "source": [ 53 | "dataDir = '../data/compMusicDatasets/turkishMakam/'\n", 54 | "downloadData = True#set to False if you already downloaded the data\n", 55 | "\n", 56 | "if downloadData:\n", 57 | " if not os.path.exists(dataDir):#creating the directory\n", 58 | " os.mkdir(dataDir);\n", 59 | " url = 'https://github.com/MTG/SymbTr/archive/v2.0.0.zip'\n", 60 | " filename = 'SymbTr-2.0.0.zip' \n", 61 | " #Downloading the zip file from the url\n", 62 | " urllib.request.urlretrieve(url, filename)\n", 63 | " #Unzipping to a specific folder\n", 64 | " zip_ref = zipfile.ZipFile(filename, 'r')\n", 65 | " zip_ref.extractall(dataDir)\n", 66 | " zip_ref.close()\n", 67 | " os.remove(filename)#Removing the zip file\n", 68 | " final_data_dir = os.path.join(dataDir, filename.replace('.zip',''))\n", 69 | " print('Data downloaded and unzipped to: %s' % final_data_dir)\n", 70 | "else:#assuming data was downloaded in a previous call\n", 71 | " final_data_dir = os.path.join(dataDir, 'SymbTr-2.0.0')" 72 | ] 73 | }, 74 | { 75 | "cell_type": "markdown", 76 | "metadata": {}, 77 | "source": [ 78 | "#### Grouping files with respect to makams, printing a short sorted table\n" 79 | ] 80 | }, 81 | { 82 | "cell_type": "code", 83 | "execution_count": 3, 84 | "metadata": {}, 85 | "outputs": [], 86 | "source": [ 87 | "#Function definitions\n", 88 | "def printSortedList(file_dict, numCategories = 5):\n", 89 | " '''Prints list of categories sorted in terms of number of files contained\n", 90 | " \n", 91 | " Args:\n", 92 | " file_dict (dict): dictionary containing categories mapped to file list\n", 93 | " numCategories (int): number of categories \n", 94 | " '''\n", 95 | " categories = list(file_dict.keys())\n", 96 | " \n", 97 | " numFiles_category = np.zeros((len(categories),), dtype=int)\n", 98 | " for index in range(len(categories)):\n", 99 | " numFiles_category[index] = len(file_dict[categories[index]])\n", 100 | " \n", 101 | " sortedIndexes=np.flipud(np.argsort(numFiles_category))#highest to lowest sorting\n", 102 | " for index in range(numCategories):\n", 103 | " print(categories[sortedIndexes[index]],' :\\t', numFiles_category[sortedIndexes[index]])\n", 104 | "\n", 105 | "def constructFileDict(fileList, tokenIndex):\n", 106 | " '''Constructs/returns dictionary that maps categories to file lists\n", 107 | " \n", 108 | " The file name is split into tokens accorindg to the format of SymbTr slug:\n", 109 | " makam--form--usul--title--composer\n", 110 | " \n", 111 | " Args:\n", 112 | " fileList (list): file list\n", 113 | " tokenIndex (int): category to be used for grouping the files\n", 114 | " Example: tokenIndex=0, the dictionary returned contains files grouped\n", 115 | " in makam categories (keys: makam name, values: file list for the makam).\n", 116 | " Outputs:\n", 117 | " file_dict (dict): dictionary makking categories to file lists \n", 118 | " '''\n", 119 | " file_dict = dict()\n", 120 | " for txtFile in fileList:\n", 121 | " tokens = txtFile.split('--')\n", 122 | " category = tokens[tokenIndex]\n", 123 | "\n", 124 | " files4category = file_dict.get(category)\n", 125 | " if files4category == None:\n", 126 | " files4category = [txtFile]\n", 127 | " else:\n", 128 | " files4category.append(txtFile)\n", 129 | " file_dict[category] = files4category\n", 130 | " \n", 131 | " return file_dict\n" 132 | ] 133 | }, 134 | { 135 | "cell_type": "code", 136 | "execution_count": 4, 137 | "metadata": {}, 138 | "outputs": [ 139 | { 140 | "name": "stdout", 141 | "output_type": "stream", 142 | "text": [ 143 | "-------------------------------------\n", 144 | "Most frequently used modes in score collection \n", 145 | "hicaz :\t 160\n", 146 | "nihavent :\t 128\n", 147 | "ussak :\t 119\n", 148 | "rast :\t 110\n", 149 | "huzzam :\t 96\n", 150 | "segah :\t 93\n", 151 | "huseyni :\t 91\n", 152 | "mahur :\t 88\n", 153 | "hicazkar :\t 79\n", 154 | "kurdilihicazkar :\t 70\n", 155 | "-------------------------------------\n", 156 | "Most frequently used forms in score collection \n", 157 | "sarki :\t 991\n", 158 | "turku :\t 295\n", 159 | "seyir :\t 156\n", 160 | "kupe :\t 120\n", 161 | "pesrev :\t 93\n", 162 | "sazsemaisi :\t 82\n", 163 | "aranagme :\t 73\n", 164 | "ilahi :\t 41\n", 165 | "beste :\t 39\n", 166 | "yuruksemai :\t 38\n" 167 | ] 168 | } 169 | ], 170 | "source": [ 171 | "#Constructing file lists and printing\n", 172 | "txtFilesList = os.listdir(os.path.join(final_data_dir, 'txt'))\n", 173 | "makamsFileList = constructFileDict(txtFilesList, 0)\n", 174 | "formsFileList = constructFileDict(txtFilesList, 1)\n", 175 | "\n", 176 | "print('-------------------------------------')\n", 177 | "print('Most frequently used modes in score collection ')\n", 178 | "printSortedList(makamsFileList, 10)\n", 179 | "print('-------------------------------------')\n", 180 | "print('Most frequently used forms in score collection ')\n", 181 | "printSortedList(formsFileList, 10)" 182 | ] 183 | }, 184 | { 185 | "cell_type": "markdown", 186 | "metadata": {}, 187 | "source": [ 188 | "### Plotting melodic curves\n", 189 | "Below we plot first sections in saz-semaisi form in two makams: rast and mahur. " 190 | ] 191 | }, 192 | { 193 | "cell_type": "code", 194 | "execution_count": 5, 195 | "metadata": {}, 196 | "outputs": [], 197 | "source": [ 198 | "def readInstrumentalSection(filePath, sectionStartStr = '1. HANE'):\n", 199 | " '''Reading notes of an instrumental section\n", 200 | " \n", 201 | " Args:\n", 202 | " filePath (str): file path\n", 203 | " sectionStartStr (str): section start label. \n", 204 | " The section boundaries are marked in the lyrics column with specific indicators.\n", 205 | " Example: In saz-semaisi, '1. HANE' refers to the A section\n", 206 | " Outputs:\n", 207 | " notesSequence (list): note sequence as a list of tuples (midi number, duration) \n", 208 | " '''\n", 209 | " file = open(filePath,'r')\n", 210 | " content = file.read()\n", 211 | " lines = content.split('\\n')\n", 212 | " #Reading content\n", 213 | " notesSequence = []\n", 214 | " inSection = False\n", 215 | " for line in lines[1:]:#skipping the first line\n", 216 | " tokens = line.split('\\t')\n", 217 | " if(len(tokens) > 8):\n", 218 | " lyrics = tokens[11]\n", 219 | " if (not inSection) and (sectionStartStr in line):\n", 220 | " inSection = True\n", 221 | " elif inSection and len(lyrics) > 0:#any lyrics non-empty is considered a section-end if section is already started\n", 222 | " inSection = False\n", 223 | "\n", 224 | " if inSection:\n", 225 | " intervalInHc = int(tokens[4])\n", 226 | " if intervalInHc > 0:\n", 227 | " midiNumber = intervalInHc*(1200/53)/100#conversion from Holderian commas to midi no\n", 228 | " else:\n", 229 | " midiNumber = -1#pause represented with midi no = -1\n", 230 | " notesSequence.append((midiNumber, int(tokens[8])))\n", 231 | " return notesSequence\n", 232 | "\n", 233 | "def melodicCurveFromNotesSequence(notesSequence, totalNumPoints = 1000):\n", 234 | " '''Constructing melodic curve from note sequence with a fixed size \n", 235 | " \n", 236 | " Duration is scaled accordingly to match totalNumPoints\n", 237 | " \n", 238 | " Args:\n", 239 | " notesSequence (list): note sequence as a list of tuples (midi number, duration)\n", 240 | " totalNumPoints (int): total number of points to represent the melodic curve\n", 241 | " Outputs: \n", 242 | " numpy array containing samples(in midi numbers) of a melody sequence\n", 243 | " '''\n", 244 | " totalDuration = 0\n", 245 | " for (note,duration) in notesSequence:\n", 246 | " totalDuration += duration\n", 247 | " \n", 248 | " durationUnit = totalDuration/totalNumPoints\n", 249 | " melograph = []\n", 250 | " for (note,duration) in notesSequence:\n", 251 | " if(note > 0):#pauses are discarded\n", 252 | " numPoints = int(duration/durationUnit)#due to this rounding, totalNumPoints will not be exactly matched\n", 253 | " melograph += ([note]*numPoints)\n", 254 | " return np.array(melograph)" 255 | ] 256 | }, 257 | { 258 | "cell_type": "code", 259 | "execution_count": 6, 260 | "metadata": {}, 261 | "outputs": [ 262 | { 263 | "name": "stdout", 264 | "output_type": "stream", 265 | "text": [ 266 | "Files in makam rast and form: sazsemaisi\n", 267 | "['rast--sazsemaisi--aksaksemai----benli_hasan_aga.txt', 'rast--sazsemaisi--aksaksemai----kantemiroglu.txt'] \n", 268 | "\n", 269 | "Files in makam mahur and form: sazsemaisi\n", 270 | "['mahur--sazsemaisi--aksaksemai----gazi_giray_han.txt', 'mahur--sazsemaisi--aksaksemai----nikolaki.txt', 'mahur--sazsemaisi--aksaksemai----refik_talat_alpman.txt', 'mahur--sazsemaisi--aksaksemai--bahar_1--goksel_baktagir.txt'] \n", 271 | "\n" 272 | ] 273 | }, 274 | { 275 | "data": { 276 | "text/plain": [ 277 | "" 278 | ] 279 | }, 280 | "execution_count": 6, 281 | "metadata": {}, 282 | "output_type": "execute_result" 283 | }, 284 | { 285 | "data": { 286 | "image/png": "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\n", 287 | "text/plain": [ 288 | "
" 289 | ] 290 | }, 291 | "metadata": { 292 | "needs_background": "light" 293 | }, 294 | "output_type": "display_data" 295 | } 296 | ], 297 | "source": [ 298 | "def formFileListForMakamAndForm(fileList, makam, form):\n", 299 | " '''Selects/returns a file list ( as a subset from the whole list: 'fileList') \n", 300 | " given the filters: makam(str) and form(str)'''\n", 301 | " selectedFiles = []\n", 302 | " for txtFile in fileList:\n", 303 | " if((makam == txtFile.split('--')[0]) and (form == txtFile.split('--')[1])):\n", 304 | " selectedFiles.append(txtFile)\n", 305 | " return selectedFiles\n", 306 | "\n", 307 | "#Let's pick two makams only differing in seyir and plotting melodic curves \n", 308 | "# for first sections in the sazsemaisi form\n", 309 | "makams = ['rast','mahur']\n", 310 | "form = 'sazsemaisi'\n", 311 | "sectionStartStr = '1. HANE'#section's start indicator (in lyrics)\n", 312 | "\n", 313 | "plt.figure(figsize = (13, 3))\n", 314 | "colors = ['k','r--','b.','g.']#plotting color options for different makams\n", 315 | "minInterval = 400#a high interval value to be updated with min value observed\n", 316 | "maxInterval = 0#a low interval value to be updated with max value observed\n", 317 | "for index in range(len(makams)):\n", 318 | " makam = makams[index]\n", 319 | " selectedFileList = formFileListForMakamAndForm(txtFilesList, makam, form)\n", 320 | " print('Files in makam ', makam, ' and form: ', form)\n", 321 | " print(selectedFileList, '\\n')\n", 322 | " for symbTrFile in selectedFileList:\n", 323 | " #Reading notes of a section (section starting with sectionStartStr in lyrics, ending with any non-empty lyric)\n", 324 | " notesSequence=readInstrumentalSection(final_data_dir + '/txt/' + symbTrFile, sectionStartStr)\n", 325 | " if(len(notesSequence) > 0):\n", 326 | " #print(symbTrFile)\n", 327 | " #print('-----------------------')\n", 328 | " melograph = melodicCurveFromNotesSequence(notesSequence)\n", 329 | " plt.plot(melograph,colors[index], label = makam)\n", 330 | " minInterval = min(np.min(melograph), minInterval)\n", 331 | " maxInterval = max(np.max(melograph), maxInterval)\n", 332 | "\n", 333 | "plt.ylim([minInterval, maxInterval])\n", 334 | "plt.ylabel('midi number')\n", 335 | "plt.xlabel('time')\n", 336 | "plt.legend()" 337 | ] 338 | }, 339 | { 340 | "cell_type": "markdown", 341 | "metadata": {}, 342 | "source": [ 343 | "#### Note that the midi numbers are fractional. Here is part of a note sequence" 344 | ] 345 | }, 346 | { 347 | "cell_type": "code", 348 | "execution_count": 7, 349 | "metadata": {}, 350 | "outputs": [ 351 | { 352 | "name": "stdout", 353 | "output_type": "stream", 354 | "text": [ 355 | "Sample note sequence:\n", 356 | "List of (midi number, duration), midi=-1 refers to silence \n", 357 | "\n", 358 | "[(79.0188679245283, 1000), (74.0377358490566, 500), (74.0377358490566, 125), (76.0754716981132, 125), (77.88679245283019, 125), (79.0188679245283, 125), (81.0566037735849, 125), (82.86792452830188, 125), (84.0, 125), (86.03773584905659, 125)]\n" 359 | ] 360 | } 361 | ], 362 | "source": [ 363 | "print('Sample note sequence:')\n", 364 | "print('List of (midi number, duration), midi=-1 refers to silence','\\n')\n", 365 | "print(notesSequence[:10])" 366 | ] 367 | } 368 | ], 369 | "metadata": { 370 | "kernelspec": { 371 | "display_name": "Python 3", 372 | "language": "python", 373 | "name": "python3" 374 | }, 375 | "language_info": { 376 | "codemirror_mode": { 377 | "name": "ipython", 378 | "version": 3 379 | }, 380 | "file_extension": ".py", 381 | "mimetype": "text/x-python", 382 | "name": "python", 383 | "nbconvert_exporter": "python", 384 | "pygments_lexer": "ipython3", 385 | "version": "3.5.2" 386 | } 387 | }, 388 | "nbformat": 4, 389 | "nbformat_minor": 2 390 | } 391 | -------------------------------------------------------------------------------- /notebooks/tuningAnalysis_SetOfRecordings.ipynb: -------------------------------------------------------------------------------- 1 | { 2 | "cells": [ 3 | { 4 | "cell_type": "markdown", 5 | "metadata": {}, 6 | "source": [ 7 | "## Tuning analysis for a set of recordings from the same mode\n", 8 | "\n", 9 | "Here, we demonstrate tuning analysis of a set of recordings in several modes. The names of the modes and number of files used in each mode is set in the second code cell. \n", 10 | "The following steps are carried:\n", 11 | "- A group of recordings from the set of files listed in the [annotations.json file](https://github.com/MTG/otmm_makam_recognition_dataset ) in the given modes are downloaded as the first step.\n", 12 | "- Then pitch analysis and pitch distribution computation is carried\n", 13 | "- Tonic annotation is accessed from Dunya and interval distributions are computed (with respect to tonic)\n", 14 | "- Octave folding is applied and mean of all distributions are computed\n", 15 | "- Automatic peak picking applied to detect scale degrees' distance to the tonic\n", 16 | "- Interval list stored in the Scala format (http://www.huygens-fokker.org/scala/) to be able to sonify the intervals using computer keyboard. A Scala file(.scl) is created for each mode in the modes list. \n", 17 | "\n", 18 | "To be able to download sounds from Dunya, you would need to create a user and obtain an API authenticaion key(token). Please create your user from: https://dunya.compmusic.upf.edu/developers/ In order to get your API token you have to log in to Dunya, access your profile, you will find your token there. " 19 | ] 20 | }, 21 | { 22 | "cell_type": "code", 23 | "execution_count": 4, 24 | "metadata": {}, 25 | "outputs": [], 26 | "source": [ 27 | "# Set your token here from https://dunya.compmusic.upf.edu/user/profile/\n", 28 | "token = '...yourAPITokenGoesHere...'" 29 | ] 30 | }, 31 | { 32 | "cell_type": "code", 33 | "execution_count": 5, 34 | "metadata": {}, 35 | "outputs": [], 36 | "source": [ 37 | "import os \n", 38 | "import json\n", 39 | "import matplotlib.pyplot as plt\n", 40 | "import matplotlib.ticker as ticker\n", 41 | "import numpy as np\n", 42 | "from compmusic.dunya import docserver as ds\n", 43 | "from compmusic import dunya as dn\n", 44 | "from external_utilities.predominantmelodymakam import PredominantMelodyMakam\n", 45 | "from external_utilities.pitchdistribution import PitchDistribution\n", 46 | "from scipy.spatial import distance\n", 47 | "\n", 48 | "from compmusic import dunya\n", 49 | "dn.set_token(token)\n", 50 | "\n", 51 | "%matplotlib inline\n", 52 | "CENTS_IN_OCTAVE = 1200\n", 53 | "REF_PITCH = 220" 54 | ] 55 | }, 56 | { 57 | "cell_type": "markdown", 58 | "metadata": {}, 59 | "source": [ 60 | "### Downloading the recordings with mode 'Huseyni'" 61 | ] 62 | }, 63 | { 64 | "cell_type": "code", 65 | "execution_count": 6, 66 | "metadata": {}, 67 | "outputs": [ 68 | { 69 | "name": "stdout", 70 | "output_type": "stream", 71 | "text": [ 72 | "Downloading mp3 files in ../data/compMusicDatasets/turkishMakam\n", 73 | "15 files downloaded\n", 74 | "15 files downloaded\n", 75 | "15 files downloaded\n", 76 | "Download finished!\n" 77 | ] 78 | } 79 | ], 80 | "source": [ 81 | "fileType = 'mp3'\n", 82 | "numFiles = 15\n", 83 | "collectionName = 'makam'\n", 84 | "modeType = 'makam'\n", 85 | "# modes for which we will estimate a scale\n", 86 | "modes = ['Huseyni', 'Saba', 'Huzzam']\n", 87 | "dataDir = os.path.join('..', 'data', 'compMusicDatasets', 'turkishMakam')\n", 88 | "\n", 89 | "# Load dataset files in mode Huseyni\n", 90 | "with open(os.path.join(dataDir, 'annotations.json')) as json_data:\n", 91 | " collectionFiles = json.load(json_data)\n", 92 | "\n", 93 | "#Create directories for modes and download one recording for each\n", 94 | "modeFilesInfo = {}\n", 95 | "print('Downloading mp3 files in {}'.format(dataDir))\n", 96 | "for mode in modes:\n", 97 | " os.makedirs(os.path.join(dataDir, mode), exist_ok=True)\n", 98 | "\n", 99 | " mbidList = []\n", 100 | " tonics = {}\n", 101 | " fileCnt = 0\n", 102 | " for file in collectionFiles:\n", 103 | " if file[modeType] == mode:\n", 104 | " mbid = file['mbid'].split('http://musicbrainz.org/recording/')[-1]\n", 105 | " try:\n", 106 | " content = ds.get_document_as_json(mbid, 'audioanalysis', 'tonic')\n", 107 | " tonic = content['value']\n", 108 | " except dunya.HTTPError:\n", 109 | " tonic = None\n", 110 | " \n", 111 | " # If tonic info is available from Dunya, add to set of recordings\n", 112 | " if not tonic:\n", 113 | " print('Tonic could not be read for {}, skipping this file'.format(mbid))\n", 114 | " else:\n", 115 | " tonics[mbid] = float(tonic)\n", 116 | " mbidList.append(mbid)\n", 117 | " \n", 118 | " #Download mp3\n", 119 | " name = '{}.{}'.format(mbid, fileType)\n", 120 | " mp3FileURI = os.path.join(dataDir, mode, name)\n", 121 | " # Download file if it has not been previously downloaded\n", 122 | " if not os.path.exists(mp3FileURI):\n", 123 | " contents = ds.file_for_document(mbid, fileType)\n", 124 | " open(mp3FileURI, \"wb\").write(contents)\n", 125 | " fileCnt += 1\n", 126 | " if fileCnt >= numFiles:\n", 127 | " print('{} files downloaded'.format(numFiles))\n", 128 | " break\n", 129 | " modeFilesInfo[mode] = (mbidList, tonics)\n", 130 | " \n", 131 | "print('Download finished!')" 132 | ] 133 | }, 134 | { 135 | "cell_type": "markdown", 136 | "metadata": {}, 137 | "source": [ 138 | "### Pitch extraction and distribution computation" 139 | ] 140 | }, 141 | { 142 | "cell_type": "code", 143 | "execution_count": 7, 144 | "metadata": {}, 145 | "outputs": [], 146 | "source": [ 147 | "def tonic_aligned_octave_wrapped_dist(mode, mbidList, tonics, pd_params, numBins, dataDir):\n", 148 | " '''Gathering tonic aligned octave wrapped distributions \n", 149 | " \n", 150 | " Args:\n", 151 | " mode (str): name of the mode ('Huseyni', 'Saba', etc)\n", 152 | " mbidList (list): list of musicbrainz ids\n", 153 | " tonics (dict): dictionary mapping mbids to tonic info in Hz\n", 154 | " numBins (int): number of bins in octave for the distribution\n", 155 | " dataDir (str): path info\n", 156 | " Outputs:\n", 157 | " pcds (numpy array(2D)): 2D array(pd size * number of file) \n", 158 | " containing pitch distribution for each file \n", 159 | " '''\n", 160 | " pcds = np.array([]).reshape(0,numBins)\n", 161 | " #Pitch extractor definition\n", 162 | " extractor = PredominantMelodyMakam(filter_pitch=True)\n", 163 | " print('Pitch analysis of files in mode ', mode)\n", 164 | "\n", 165 | " for mbid in mbidList:\n", 166 | " name = '{}.{}'.format(mbid, fileType)\n", 167 | " mp3FileURI = os.path.join(dataDir, mode,name)\n", 168 | " #Setting file names for writing analysis results\n", 169 | " pitchFile = os.path.join(dataDir, mode, '{}.pitch'.format(mbid))\n", 170 | " histFile = os.path.join(dataDir, mode, '{}.pitch_hist.json'.format(mbid))\n", 171 | "\n", 172 | " #If pitch file exists, read it, if not run extractor and create the pitch file\n", 173 | " if not os.path.exists(pitchFile):\n", 174 | " #running pitch extraction \n", 175 | " results = extractor.run(mp3FileURI)\n", 176 | " pitch = results['settings'] # collapse the keys in settings\n", 177 | " pitch['pitch'] = results['pitch']\n", 178 | " # Write pitch data to text file, \n", 179 | " # you can use it together with SonicVisualizer to view in sync with the spectrogram of the mp3 file\n", 180 | " pitchSeriesHz = []\n", 181 | " file = open(pitchFile,'w')\n", 182 | " for p_triplet in pitch['pitch']:\n", 183 | " file.write(str(p_triplet[0])+'\\t'+str(p_triplet[1])+'\\n')\n", 184 | " pitchSeriesHz.append(p_triplet[1])\n", 185 | " file.close()\n", 186 | " pitchSeriesHz = np.array(pitchSeriesHz)\n", 187 | " else:\n", 188 | " pitchData = np.loadtxt(pitchFile)\n", 189 | " timeStamps = pitchData[:,0]\n", 190 | " pitchSeriesHz = pitchData[:,1]\n", 191 | "\n", 192 | " #Pitch distribution computation\n", 193 | " #Computing pitch distribution with default reference frequency = REF_PITCH\n", 194 | " pitch_distribution = PitchDistribution.from_hz_pitch(pitchSeriesHz,REF_PITCH, **pd_params)\n", 195 | " pitch_distribution.to_json(histFile)\n", 196 | "\n", 197 | " #Computing pitch distribution with reference frequency = tonic\n", 198 | " tonic_Hz = tonics[mbid]\n", 199 | " pitch_distribution_tonicRef = PitchDistribution.from_hz_pitch(pitchSeriesHz,tonic_Hz, **pd_params)\n", 200 | " pitch_distribution_tonicRef.to_json(histFile.replace('.pitch_hist.json','.pitch_hist_wrtTonic.json'))\n", 201 | "\n", 202 | " #Creating octave folded distribution\n", 203 | " pcd = np.zeros(numBins)#initializing pitch class distribution\n", 204 | " for index_pd in range(len(pitch_distribution_tonicRef.bins)):\n", 205 | " bin_pd = pitch_distribution_tonicRef.bins[index_pd] #get bin in pitch distribution\n", 206 | " index_pcd = int(np.mod(bin_pd,CENTS_IN_OCTAVE)/pd_params['step_size']) #corresponding index in pitch class distribution\n", 207 | " pcd[index_pcd] += pitch_distribution_tonicRef.vals[index_pd]\n", 208 | " pcds = np.vstack((pcds,pcd))\n", 209 | "\n", 210 | " return pcds\n" 211 | ] 212 | }, 213 | { 214 | "cell_type": "code", 215 | "execution_count": 8, 216 | "metadata": {}, 217 | "outputs": [ 218 | { 219 | "name": "stdout", 220 | "output_type": "stream", 221 | "text": [ 222 | "Pitch analysis of files in mode Huseyni\n", 223 | "Pitch analysis of files in mode Saba\n", 224 | "Pitch analysis of files in mode Huzzam\n", 225 | "Pitch files are stored in ../data/compMusicDatasets/turkishMakam\n", 226 | "You can use Sonic Visualizer at this step to check pitch extraction quality\n" 227 | ] 228 | } 229 | ], 230 | "source": [ 231 | "# Running extraction and gathering of data\n", 232 | "# pitch distribution extractor parameters in cents\n", 233 | "pd_params = {'kernel_width': 5, 'step_size': 5}\n", 234 | "# distribution bins for octave-folded histogram\n", 235 | "bins = np.linspace(0, CENTS_IN_OCTAVE, CENTS_IN_OCTAVE/pd_params['step_size'], endpoint=False)\n", 236 | "numBins = len(bins)\n", 237 | "\n", 238 | "# Gathering tonic aligned, octave wrapped distributions for each mode\n", 239 | "modePcds = {}\n", 240 | "for mode in modes:\n", 241 | " mbidList, tonics = modeFilesInfo[mode]\n", 242 | " pcds = tonic_aligned_octave_wrapped_dist(mode, mbidList, tonics, pd_params, numBins, dataDir)\n", 243 | " modePcds[mode] = pcds\n", 244 | "\n", 245 | "print('Pitch files are stored in {}'.format(dataDir))\n", 246 | "print('You can use Sonic Visualizer at this step to check pitch extraction quality')" 247 | ] 248 | }, 249 | { 250 | "cell_type": "code", 251 | "execution_count": null, 252 | "metadata": {}, 253 | "outputs": [], 254 | "source": [ 255 | "#Plotting all distributions\n", 256 | "for mode in modes:\n", 257 | " pcds = modePcds[mode]\n", 258 | " plt.figure(figsize=(13, 3))\n", 259 | " for pcd in pcds:\n", 260 | " plt.plot(bins,pcd)\n", 261 | "\n", 262 | " plt.title('Octave folded pitch distributions for '+str(len(pcds))+' files in mode '+mode)\n", 263 | " plt.xlim([0,1200])\n", 264 | " plt.grid()\n", 265 | " ticks = np.arange(0,1200,100)\n", 266 | " plt.vlines(ticks, 0, np.max(pcds), color='y',lw=1)\n", 267 | " plt.ylabel('Relative freq. of occurence')\n", 268 | " plt.xlabel('Distance to tonic(cents)')" 269 | ] 270 | }, 271 | { 272 | "cell_type": "markdown", 273 | "metadata": {}, 274 | "source": [ 275 | "### Plotting average distribution and extracting scale information" 276 | ] 277 | }, 278 | { 279 | "cell_type": "code", 280 | "execution_count": null, 281 | "metadata": {}, 282 | "outputs": [], 283 | "source": [ 284 | "#Function definition for automatic scale-interval detection from pitch distribution\n", 285 | "def peakLocationDetection(pcd):\n", 286 | " '''A simple peak detection implementation for demonstration purposes\n", 287 | " Thresholds are manually set for this demo\n", 288 | " '''\n", 289 | " windowSize=15#should be odd\n", 290 | " midPointIndex=int(windowSize/2)\n", 291 | " threshold=np.max(pcd)*0.05\n", 292 | " peakIndexes=[]\n", 293 | " for index in range(len(pcd)-windowSize):\n", 294 | " frame=pcd[index:index+windowSize]\n", 295 | " if (np.argmax(frame)==midPointIndex) and np.max(frame)>threshold:\n", 296 | " peakIndexes.append(index+midPointIndex)\n", 297 | " return peakIndexes" 298 | ] 299 | }, 300 | { 301 | "cell_type": "code", 302 | "execution_count": null, 303 | "metadata": {}, 304 | "outputs": [], 305 | "source": [ 306 | "#Extracting intervals and plotting together with mean distributions\n", 307 | "modeIntervals = {}\n", 308 | "for mode in modes:\n", 309 | " pcds = modePcds[mode]\n", 310 | " \n", 311 | " pcds_array = np.array(pcds)\n", 312 | " mean_pcd = np.mean(pcds, axis=0)\n", 313 | "\n", 314 | " plt.figure(figsize=(13, 3))\n", 315 | " ticks = np.arange(0,1200,100)\n", 316 | " plt.vlines(ticks, 0, max(mean_pcd), color='y', lw=1)\n", 317 | " plt.plot(bins,mean_pcd)\n", 318 | "\n", 319 | " plt.title('Octave folded mean pitch distribution for mode {}'.format(mode))\n", 320 | " plt.xlim([0,1200])\n", 321 | " plt.grid()\n", 322 | " plt.ylabel('Relative freq. of occurence')\n", 323 | " plt.xlabel('Distance to tonic(cents)')\n", 324 | "\n", 325 | " # Detect intervals from pitch distribution and plot them on the figure\n", 326 | " intervals = np.array(peakLocationDetection(mean_pcd))*pd_params['step_size']\n", 327 | " plt.vlines(intervals, 0, max(mean_pcd), color='r', lw=2)\n", 328 | " print('Intervals computed: {} (cents with respect to tonic)'.format(intervals))\n", 329 | " modeIntervals[mode] = intervals" 330 | ] 331 | }, 332 | { 333 | "cell_type": "markdown", 334 | "metadata": {}, 335 | "source": [ 336 | "### Creating the Scala file\n", 337 | "Writing the scale to .scl file which can be loaded in Scala with which one can sonify the estimated scale " 338 | ] 339 | }, 340 | { 341 | "cell_type": "code", 342 | "execution_count": null, 343 | "metadata": {}, 344 | "outputs": [], 345 | "source": [ 346 | "for mode in modes:\n", 347 | " intervals=modeIntervals[mode]\n", 348 | " scalaFile=dataDir+mode+'_scale.scl'\n", 349 | " file = open(scalaFile,'w')\n", 350 | " file.write('! autopeak.scl\\n!\\nFile created by tuningAnalysis\\n'+str(len(intervals)+1)+'\\n!\\n')\n", 351 | " #First octave\n", 352 | " for interval in intervals:\n", 353 | " file.write(str(float(interval))+'\\n')\n", 354 | " file.write(str(float(CENTS_IN_OCTAVE))+'\\n')#octave\n", 355 | " file.close()\n", 356 | "\n", 357 | "\"\"\"\n", 358 | "scalaFile = os.path.join(dataDir, '{}_scale.scl'.format(mode))\n", 359 | "with open(scalaFile, 'w') as fp:\n", 360 | " fp.write('! autopeak.scl\\n!\\nFile created by tuningAnalysis\\n'+str(len(intervals)+1)+'\\n!\\n')\n", 361 | "\n", 362 | " #First octave\n", 363 | " for interval in intervals:\n", 364 | " fp.write(str(float(interval))+'\\n')\n", 365 | " fp.write(str(float(CENTS_IN_OCTAVE))+'\\n')#octave\n", 366 | " fp.close()\n", 367 | "\"\"\"" 368 | ] 369 | }, 370 | { 371 | "cell_type": "markdown", 372 | "metadata": {}, 373 | "source": [ 374 | "### Loading the estimated scales in Scala\n", 375 | "Initiate a synthesizer your Scala software can communicate with for synthesis (for example simplesynth). Open Scala and click 'Open' to choose the .scl file this code has created in your local folder: 'mode'_scale.scl. A scl file is created for each mode.\n", 376 | "\n", 377 | "Scala would display the set of pitches of the loaded scale as a list and set the keyboard layout to start the scale with C. If you would like to set tonic to some other note and frequency, click 'Opts.' on the top menu. You can set the tonic frequency('frequency for 1/1') and the offset (deafult is C.0)\n", 378 | "\n", 379 | "Now you can click 'Play' on the top menu to start experiment with your new keyboard playing the scale automatically extracted by the analysis above. Enjoy!" 380 | ] 381 | } 382 | ], 383 | "metadata": { 384 | "kernelspec": { 385 | "display_name": "Python 3", 386 | "language": "python", 387 | "name": "python3" 388 | }, 389 | "language_info": { 390 | "codemirror_mode": { 391 | "name": "ipython", 392 | "version": 3 393 | }, 394 | "file_extension": ".py", 395 | "mimetype": "text/x-python", 396 | "name": "python", 397 | "nbconvert_exporter": "python", 398 | "pygments_lexer": "ipython3", 399 | "version": "3.5.2" 400 | } 401 | }, 402 | "nbformat": 4, 403 | "nbformat_minor": 2 404 | } 405 | --------------------------------------------------------------------------------