(Subscribe to this discussion)

Musical instrument sounds have distinct timbral and emotional characteristics that can change when audio processing is applied. This paper investigates the effects of MP3 compression on the emotional characteristics of eight sustained instrument sounds using listening tests. The experimental paradigm involved a pairwise comparison of compressed and uncompressed samples at several bit rates over ten emotional categories. The results showed that MP3 compression strengthened neutral and negative emotional characteristics such as Mysterious, Shy, Scary, and Sad, and weakened positive emotional characteristics such as Happy, Heroic, Romantic, Comic, and Calm. Angry was relatively unaffected by MP3 compression, probably because the background “growl” artifacts added by MP3 compression decreased positive emotional characteristics and increased negative characteristics such as Mysterious and Scary. Compression effected some instruments more and others less; trumpet was the most effected and the horn the least.

Authors: Mo, Ronald; Choi, Ga Lam; Lee, Chung; Horner, Andrew
Affiliations: Department of Computer Science and Engineering, Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong; United Overseas Bank, Singapore(See document for exact affiliation information.)
JAES Volume 64 Issue 11 pp. 858-867; November 2016
Publication Date: December 1, 2016

Download Now (292 KB)

This paper is Open Access which means you can download it for free.

(Comment on this paper)

Comments on this paper

Paul Jessop	Comment posted December 22, 2016 @ 16:55:23 UTC (Comment permalink) For this report of experimental findings to be useful, it would be necessary to know whether the quoted data rates are per channel or for a stereo signal (it would also be important to know how the encoder was configured: constant or variable bit rate).  According to the Wikipedia (and my understanding which is why I checked), Spotify does not use MP3 coding as the paper states. It uses Ogg Vorbis. It is not clear to me that an extrapolation from single instrument to real world music is justified. Although John Coltrane's music does indeed feature a prominent saxophone, the encoder is not encoding the saxophone notes in a vacuum - it has to encode all the other instruments as well and the artefacts imposed on the saxophone sounds will be affected by the other sounds to be encoded. Perhaps a further study using real music, with a modern codec and realistic data rates would be useful. (Respond to this comment)
Author Response Andrew Horner	Comment posted December 30, 2016 @ 21:15:51 UTC (Comment permalink) Thank you very much for your comments. Constant bit rates were used in our study. Since all instrument sounds were in mono, the bit rates we picked were per channel.   We agree that a better wording would be that "music providers such as Spotify use MP3 or related compression algorithms".   Apparently, a saxophone solo in the middle of a piece also contains some other instruments like the rhythm section. It is not clear how mp3 compression changes the perceived emotion of real world music. However, our results give an insight into how MP3 compression may change the emotional characteristics of a music piece. We also agree that testing MP3 compression on musical excerpts is a very promising area for further work.   Thank you so much for your suggestions. Ron Mo and Andrew Horner (Respond to this comment)

Subscribe to this discussion

To be notified of new comments on this paper you can subscribe to this RSS feed. Forum users should login to see additional options.

Join this discussion!

If you would like to contribute to the discussion about this paper and are an AES member then you can login here:
Username:
Password:

If you are not yet an AES member and have something important to say about this paper then we urge you to join the AES today and make your voice heard. You can join online today by clicking here.