Change search
ReferencesLink to record
Permanent link

Direct link
Music Information Retrieval. A case study of MIR in modern rock and metal music
Blekinge Institute of Technology, Faculty of Engineering, Department of Applied Signal Processing.
Blekinge Institute of Technology, Faculty of Engineering, Department of Applied Signal Processing.
2014 (English)Independent thesis Advanced level (degree of Master (Two Years))Student thesisAlternative title
Musik Informationssökning. En fallstudie av MI i modern rock och metal musik (Swedish)
Abstract [en]

"Music Information Retrieval (MIR) is a multidisciplinary research endeavor that strives to develop innovative content-based searching themes, novel interfaces, and evolving networked delivery mechanisms in an effort to make worlds vast store of music accessible to all" (Downie, summer 2004). The purpose of this study is analysis of some MIR algorithms and their efficiency on modern rock and metal songs. To do so, one algorithm for four of most commonly MIR uses was chosen, described and evaluated on full songs or fragments of them. Music separation part proved that it is possible to effectively separate repeating background. Audio fingerprinting experiments show that higher quality of records generate significantly better results. Modern metal music turned out to be difficult for tempo estimation and chord recognition was working. All results show that modern rock and metal music, even though complicated, can be analyzed but still needs more research in this field to make it effectively.

Abstract [sv]

Detta dokument beskriver fyra musik informationssökning algoritmer och deras effektivitet på modern rock och metal musik. Spåra separation, ljudfingeravtryck, tempo uppskattning och ackord erkännande lösningar har undersökts med hjälp av olika inställningar och ljudfiler för att avgöra om de kan användas för analys av modern rock och metal musik. Experiment visat att algoritmer designade för mer populära typer av musik kan användas effektivt med rätt inställningar men kan förbättras.

Place, publisher, year, edition, pages
2014. , 52 p.
Keyword [en]
Music Information Retrieval, Track Separation, Audio Fingerprinting, Tempo Estimation, Chord Recognition
National Category
Signal Processing Media and Communications
URN: urn:nbn:se:bth-2713Local ID: diva2:830004
Social and Behavioural Science, Law
Available from: 2015-04-22 Created: 2014-10-22 Last updated: 2015-06-30Bibliographically approved

Open Access in DiVA

fulltext(655 kB)92 downloads
File information
File name FULLTEXT01.pdfFile size 655 kBChecksum SHA-512
Type fulltextMimetype application/pdf

By organisation
Department of Applied Signal Processing
Signal ProcessingMedia and Communications

Search outside of DiVA

GoogleGoogle Scholar
Total: 92 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

Total: 206 hits
ReferencesLink to record
Permanent link

Direct link