Soufee

In the intersection of mood and sound comes a very interesting area which I’m willing to explore for the purpose of attempting to enrich our life using digital design in particular.

The main question I am trying to answer in this project is:

"How to find the sound that suites our mood?"

I started my research by looking at the average music consumer’s journey, then I looked at the past and the current state of music services. I gathered insights about music & mood psychology, after which I conducted a questionnaire about listening to music which included 80 individuals..

These findings helped me create a good perspective through which I was able to see what a new way of finding suitable sound might be.

I found that the need for a new way of finding music comes from few things:

1. The increasing huge amount of audio content produced everyday around the globe.

2. The time music lovers spend searching for suitable content.

3. The advances in the filed of artificial intelligence and the capabilities in offers.

Adaptive music is a form of music which changes depending on the situation, it’s popular in gaming, like adaptive soundtracks that react to the player’s actions. This gives inspiration for making a non-static changing experience, as our personal mood changes during the day and it varies from one state to another.

_______________________________________________________________________

Mood Theory

In mood repair techniques I found that there are two modes in which music can be used. Firstly there is the mood-congruent listening approach, which simply means that a person would listen to music that symbolizes their current mood. Secondly there is the mood-incongruent listening approach which means that a person would listen to what helps achieve the desired mood (most commonly feeling good).

University of Alabama 2007

The solution’s requirements

These are the things that will determine the solution itself

I was able to find these particular requirements through analyzing: (The state of music services, mood & sound psychology, Questionnaires & surveys, General observations)

Purpose: Enhancing the user's mood through sound.

Form: Since most people listen to music via smartphone the solution should be available in the form of a quick to access mobile device application and not a website.

Accuracy: multiple “moody music” solutions didn’t offer very accurate result including Spotify playlists, so this new solution should offer an accurate result which could mean that it should be based on a better technology.

Speed: It might be a fruitful idea to explore music online randomly or by typing something in the search bar to find results that matches what we are looking for, however this is time-consuming and for many people not fruitful in the case of not knowing how to communicate what’s in their mind that they are trying to search for, so this new solution has to be quicker.

Connectivity: The content that users will play exists not in the service but in other major services, like Spotify and SoundCloud, where the music exists online in huge amounts, so the app should be connected and able to import music from those services.

Suitability: I think that the solution should be adaptive to both kinds of listeners: there are two kinds of people, those who listen to what suits their mood and those who listen to what makes their desired mood, and finding that in mood repair techniques labeled as mood-congruent or mood-incongruent.

Language: referring to not just the language of the user interface but the language of the content, the soundtracks that contain vocals are in various languages and the user's preferences regarding the language in the music are different, so the solution should be customizable in a way that allows everybody to find the right language in their music.

Price: People's first complain about music services is advertisements, by excluding it from the app people will be happier with the experience, however this might indicate that the app most likely won’t be free to download.

The solution

My research has led me to create an app concept, I decided to call it Soufee, The name comes from simply mixing two words, Sound & Feeling. Because this solution comes in the intersection of those two.

The purpose of the app is to give people a more accurate and quick way to finding sounds that suits their personal mood at any giving moment

The way it’s meant to function

Each music file can be analysed to find audio data such as:

(Tempo of the track, Key in which a song is written, vocals availability degree, sound volume, layers of sound frequencies, sound texture, richness, sound dynamics and lyrics)

Using (Essentia) “Open-source library and tools for audio and music analysis, it can help arrange music and soundtracks into useful categories” So for example, a track with fast tempo and rich/high sound dynamics indicates that the track is energetic and if it was written in B Major then it has joyful and happy melody emotion in it, which makes it ideal for working out in a sunny day for instance. That example shows how a track’s mood can be defined automatically and sorted out for the users to find quickly, using a program that exports content from both Spotify and SoundCloud then analyzing the files to later sort them out using an advanced system for the user to listen to.

However a permission to bring audio from Spotify & SoundCloud or major record labels to Soufee is required in order for such an app to function because they hold the rights for legal distribution.

Another important part of a song are the lyrics. A lot can be understood about the song by listening to the lyrics, so a simple digital text analysis can be performed to scan words that compose the lyrics of a song to determine automatically what general feelings the song gives or what atmospheres it creates for the listeners. There are many libraries of song lyrics online where the info about lyrics can be taken from to be later analyzed and associated with a song (e.g. Genius).

Soufee's logo in inspired by the different types of sound waves, 4 of them are arranged in a way which reflects the app's UI.

Communication is essential between the app and the user, since the app has to know user's mood. For the user to communicate his/her mood to the app in an accurate way text input and audio input are not good good enough. It won’t be quick for the user to input text each time, like in most search bar based music services, and the same thing goes for the audio method, because saying ie. “I feel happy and a bit relaxed right now” won’t work well looking at the huge amount of music out there which has to be matched with the user's mood.

We are able to determine the degree of something we see more than something we read, that is why they use statistics visualizations in most companies. Using a visual way of communicating mood is more ideal, as it helps more accurately define the degree or the amount of something.