Institute of Art Design + Technology Dún Laoghaire
Ireland’s campus for the Creative Industries
On Show
 Close
Search
3D Animation 3D Design Modelmaking + Digital Art Animation Applied Psychology Art Character MakeUp Design Costume Design Creative Computing Creative Media Technologies Creative Music Production Design for Change Digital Marketing Film + Television Production Interaction + User Experience Design New Media Studies Photography Production Design User Experience Design Visual Communication Design




Conor Murphy 

BSc [Hons] Creative Computing

I am a final year Creative Computing student who interested in music, guitars and AI. Throughout my time at IADT, I have gathered knowledge and skills that can help me in the world of technology, with problem solving and working within a team. I hope to use what I have learned from IADT in a Full Stack Developer position in an innovative company.

This video displays a demo of the application. The user records a sample of a song they want to identify for 20 seconds. The user can pause and restart the recording or stop it whenever they want. To identify the sample they click the upload link to send the clip to be identified, when a match is found the results are displayed below the audio samples.
One method of music retravel systems is audio identification which uses audio fingerprint. It gathers the characteristic points of an audio file and places them in a constellation map as seen in diagram C. This image displays the progress of audio identification using The Beatles song "Act Naturally", the blue diagrams are from the database, while the red is a recording sample, diagram E displays the sample recording matching with the database recording.
This image displays the terminal window of when the JavaScript recording is sent to the serve to be identified. Due to an error of the JavaScript formatting, an accurate result is not given, as seen in the matched target zone, all values are 0.
This image displays the terminal window of when a Python audio file is uploaded to the server instead of the JavaScript recording. Here an accurate result is displayed, as seen in the matched target zone.
Project Description

This project is about music recognition which was developed with Python & Flask. The application allow users to use a sample recording of a song from their phone and play it into a microphone connected to a PC. The application will then try to identify the audio clip from the 6 songs that are stored in the database and show a result that is the closest match to the sample recording. The application will also show the user links that are used as lessons to play the song on guitar from YouTube and UltimateGuitar. Due to an error with the recording format, a work around is used to get an accurate result by uploading a file of the recorded sample to the application.


Download PDF

Thesis Title

The thesis of this project includes research that was carried out for the project. This included examining how music retrieval systems worked, the history of Shazam, API's & libraries that were founded that could help develop the application, and information about Flask. Other areas the thesis cover is the requirements of the project, the design of the project, how the application was implemented, testing the features, algorithm and usability of the application.


Download PDF
Creative Computing
All coursesClose
Top