Case Study 1

Youtube Feature: AI Mixer

Interactive Mixing made built in and Simple

Overview:

Many musicians/creators rely on 3rd-party tools to overlay YouTube content. YouTube Mixer simplifies everything to be included within the browser itself.

Problem:

Youtube is the main source of music where musicians come to learn and listen to music that they need to perform.

However many musicians rely on 3rd party applications to customize settings, such as raising and lowering volume on specific instruments or even changing the keys of the particular song.

Role:
UI/UX Designer

Duration:
4 Weeks | August 2025

Process:
User Research and synthesis, Solutioning, Prototyping, Usability Testing

  • Description text goes here

Design Process

Emphathize - Design Process

Research Process

I conducted research to compare YouTube with some of its competitors in the field. Spotify, Apple Music, Pandora, and Melon (Korean website) are music distributors that I decided on. The goal of my research was to see what were some of the technologies that these competitors were using and what are advantages that YouTube already have.

Advantages of YouTube:

  • YouTube has a vast catelog of original music along with covers done by different artist.

  • Because of YouTube’s reputation, most artists will upload their music on YouTube, having almost every music.

Disadvantages:

  • Youtube has more compressed audio than other streaming platforms

  • Some features that are hidden away because of Premium features

  • If I were to do this again, one aspect I would have changed in my research was look into more of the 3rd party applications that are being used for the competitive analysis.

User Interviews

I conducted interviews with 6 users who generally use YouTube for their music needs and practices (mostly musicians and people who use YouTube to practice their said instrument).

Things that I have found through my initial research:

  1. 3 users found that there is a time consuming element when it comes to finding the right key to practice. Often having to use 3rd party applications to try and get the sound that they desire

  2. Many users found they really like certain parts of songs that they would like to learn, but sometimes cannot hear the part and having freedom to change and manipulate certain parts of the track would improve their practice.

  3. Almost all of the users explained wanting some sort of way to see sheet music or some sort of help learning songs.

  • After gathering the user interviews, I found users also wanted functionality of sheet music/tabs as well as the mixer function. At this point I considered adding this feature, do to the user interviews and proceeded to try and implement this feature as well as the mixing.

After Finishing the User Interviews, I made some personas that would help to solidify what the target for this project is and how I can move forward to problem solve these Persona’s needs and challenges

With the personas made, I then went to the user flow step. Because of the feature being implemented is pretty simple and adds to existing YouTube,

Defining

Design

Low Fidelity Wireframes:

I created some low fidelity sketches from some of the ideas that were mentioned during the user interviews. The idea initially was to try and implement this into the UI of the YouTube videos themselves.

However, after some feedback sessions, I decided to shift the idea into a button because of the limitations of the slider.

  • After gathering the user interviews, I found users also wanted functionality of sheet music/tabs as well as the mixer function. At this point I considered adding this feature, do to the user interviews and proceeded to try and implement this feature as well as the mixing.

Mid Fidelity Wireframes:

After some rethinking, I realized by having a button like the ones that are found next to the like and dislike button, I can create a dropdown option that would implement all the ideas (Audio mixing, Transposing, and Chords) through the use of AI could be accomplished.

Early iterations of the button logo were inspired by the use of this feature being used for music along with the use of AI to pick out the instruments in the music selected.

Early iterations of the button logo were inspired by the use of this feature being used for music along with the use of AI to pick out the instruments in the music selected.

Design/Mood Inspirations

YouTube has an established pattern and color scheme to their website. Part of my design goals were to make sure to keep the familiarity of YouTube, but incorportate a new mechanism into YouTube seemlessly.

Testing and Iterations:

After testing and finding the designs some of the things that I ended up changes as a result of those results:

Iterations to the Mixing Element:

  1. Added instrument Icons to indicate which instrument volume is being changed

  2. Through user testing found users asking if music had more than 1 instrument how to would there be a differenciation,

    1. Added a descriptin of the instrument

  3. Wanted clearer way to indicate when an instrument is fully muted and made the icon also turn red and changed to say “unmute” when muted

  4. Added a volume and mute in place of - and + through user testing finding it reduntant.

Iterations to the Transposing Element:

  1. Through user testing, found that keeping color simple between the two functions were key.

    1. many users found that having the red felt like something was "quite” because of the mixing color being red when muted

Iterations to the AI Mixing Button:

  1. Many users indicated having a simplier design would be better.

    1. the square design was too busy with the design, and so I simplified the button to become an AI star with a music note.

  • While doing user testings for the AI Chord generator function, I found users being very confused about the functionality, indicating “"not knowing what it was suppose to be at first glance".”

    After some more user testing and iterations of the design. I found myself deciding, for the scope of this project, to omit this function and focus on Mixing and Transposing functions.

  • After some user testings, many musician testers asked why there were little customization (EQ capability, gain stage controlability)

    I considered making the design more elaborate, however through one user testing (who likes to dance and teach dance) explained that this simple layout was great for her teaching classes [example is raising the drums so that students can hear better the beat pattern clearer]

    I decided (with the time limit of the project) to keep the simple design, and after launch and more feedback, that the next iterations can include more mixing capabilities.

High Fidelity Prototype:

High Fidelity Prototype:

Figma Prototype

Takeaways/Reflections

I had a wonderful time with this project. I love playing music and to try and problem solve a problem that I had for a while was something that grew my passion for the project. Some take aways from this project:

  1. Sometimes user needs and my ideas may not always align. But navigating through those things and continuing to be open mided is key to navigate through these times.

  2. Although there were many directions to take with this project. Sometimes once things are doing its things correctly and functions well, that is good enough.

  3. Many elements go into adding a feature, color matching, familiar elements. Research and data collecting go a long way to helping create a familiar but new element.