April 30, 2024

Skylight Webzine

Online since 2000

Testing the similarity of your music with Machine Learning

2 min read

Andrson, is an AI music discovery platform, using machine learning to analyze the music released by new artists and then compares it with commercial artist recordings.

“It offers percentage-based recommendations based on sonic similarity,” explains co-founder Zach Miller-Frankel on Music Week magazine.

 For instance, if you are looking for an artist who sounds at least 70% like Coldplay, the app will deliver the results. This is an analysis of the track “Venus” by the artist “Billy Yfantis” using machine learning, displaying the similarity of the music to other artists.

According to the analysis, the track “Venus” sounds:

65% like Tenderlonious

63% like Jacob Collier

63% like Frivolous

59% like Melodysheep

59% like Heatbeat

59% like Ypsis

58% like Sofasound

58% like Khotin

58% like City Girl

57% like Jessica Audiffred

57% like Duval Timothy

57% like Deftones

55% like Slugbug

55% like Four Tet

55% like What So Not

55% like Tonti Guerito

54% like Kid Cudi

54% like Padé

54% like Hans Zimmer

54% like RÜFÜS DU SOL

53% like Silver Sphere

52% like TheFatRat

52% like Axel Boman

52% like Das Kope

52% like Ecco2k

52% like Shiey

52% like Mazoulew

51% like ULYSSE

51% like Tinlicker

51% like Gorillaz

51% like Getter

51% like The Avalanches

51% like In Love With a Ghost

50% like Throttle

50% like VIZE

50% like Victor Leksell

50% like Tears & Marble

50% like Kygo

49% like George Gretton

49% like Kyoto Love Hotel

49% like 4ЯR

49% like Bullion

49% like Jonas Aden

49% like Touch Sensitive

49% like SAULT

49% like Goth Babe

 Frankly, I am not familiar with many of the mentioned artists and it is difficult to admit that these artists inspired me. However, it is interesting to find out how a current A&R agent of a record label may perceive my music compared with what is current popular in the music industry. Moreover, another rising issue is that this method discourages the A&R personnel from finding new music trends because everybody will look for new music that sounds like established artists. It is a discouraging factor for exploring deeper the Underground music scene to find new trends and artists. It is just a safe solution for the record labels who seek to build a roster of “potentially” successful artists. I hope that this app will improve its algorithm by embedding new trend features.

Information: Get Discovered by Your Sound – Digital A&R App – Andrson

About Author