The No.1 Website for Pro Audio
 All  This Thread  Reviews  Gear Database  Gear for sale     Latest  Trending
Free AI-based music analysis tool to understand the emotional effect of music
Old 25th June 2019
  #1
Here for the gear
 

Free AI-based music analysis tool to understand the emotional effect of music

Hello everyone!

On Thursday, we are launching the AI-based music analysis tool CYANITE in order to give an objective understanding of the emotional effect of music. For this, we are combining state of art the deep learning technology (CNN's & RNN's) with a one-of-a-kind dataset that we gathered ourselves in our music app Groovecat.

CYANITE can be used in different scenarios throughout the music and creative industry. You can either compare several song choices in their emotional profile or get in-depth insights in the dynamics of moods for every second in a specific song. Production music libraries or label catalogues can be indexed coherently, quick and easy. Decision makers in film, advertising, gaming as well as artists and producers can objectively make and communicate song choices that will match the desired mood.

Maybe there’s one or two people here who want to try it out. The Basic version is completely free and accessible for everyone, so if you want to find out how Music AI can actually be used as a supportive tool, this is your chance We appreciate any kind of feedback as we'll continue developing our product together with our users.

Enjoy trying out CYANITE & thank you!

https://cyanite.ai/welcome/
https://twitter.com/Cyanite_ai
https://www.youtube.com/watch?v=RZ9FGvtiJDQ (this demo is still for the Beta version)
Attached Thumbnails
Free AI-based music analysis tool to understand the emotional effect of music-screenshot-2019-06-11-14.22.04-kopie.jpg  
Old 25th June 2019
  #2
Lives for gear
 
VitaEtMusica's Avatar
 

Very interesting. Can you batch process? Do you have an API for building this into search engines/results? Do you have a function for keeping a database and finding similar tracks based on selected emotions or curves? How would one go about exporting metadata based on results for many songs? How does one tweak the algorithm if one does not agree with Cyanite or Cyanite creators' assessment of what makes a song happy, sad, melancholy, etc? Could one assign their own keywords to replace existing keywords? One man's melancholy is another man's nostalgic.

I'm interested in seeing how well the analysis works out of the box. Your YT demo seems to be a shotgun blast of emotion, none of main characteristics particularly stronger than the other. The Billie Eilish song, for instance, I would simply label as 100% depressed.
Old 26th June 2019
  #3
Here for the gear
 

Hi VitaEtMusica!

Thanks for your cool feedback and really good questions. Seems like you are working in the field. Questions on point! Happy to answer:
  1. Batch process and API: Not yet. We’re working on it! Hope to have updates by the end of next month. Also working with first partners on the problem.
  2. Similar Tracks: Also EOM
  3. Metadata Extraction: This is one of our premium features that we have to do case by case. Customizing outputs to perfectly fit the metadata requirements of our customers is very important.
  4. Continuous learning algorithm: Currently we are a very straight forward audio analysis platform but will slowly implement more and more features of actively managing you catalogues and metadata. We will implement means to customize your „own algorithm” meaning that it responds to changes you make in the tagging. I’ll keep you posted on the developments there. Changing tags however, will be possible very soon.
  5. Subjective perception (billie eilish): This is why we build the platform. Yes in that case, the song might seem very depressing to you. However, this is significantly biased by the marketing and image around the artist. In (production) music for audiovisual content, you usually do not have this priming. The artists and songs are unknown. A song is never just one emotion but rather a compilation of nuances tending towards certain emotions. That is why it is important to depict these nuances to get the full DNA of a song.

Hope this helps and is of your interest!
Cheers
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Forum Jump
Forum Jump