Anytune Project Overview
For the final project in the Usability Research class I took at UW (this project was not sponsored by Anytune), I worked on a 3-person team to evaluate the usability of the looping function on the app Anytune. For this project, our team developed research questions, studied Anytune’s functionality, conducted 5 user interviews, and analyzed findings from user interviews to guide design recommendations.
I specifically was responsible for being the subject matter expert, recruiting participants, tailoring a set of tasks for those participants, conducting interviews, coding and analyzing notes, developing design recommendations, and contributing to the final report.
What is Anytune and what is Transcribing?
Anytune is a music practice tool that provides useful tools for musicians learning music by ear. Specifically, users can slow down recordings while maintaining pitch integrity and sound quality, isolate or ignore certain pitch frequencies, change the key of a recording, loop certain sections, and more.
Transcribing, as it is known in the jazz community, is the act of listening to a piece of music, whether it be a tune, someone’s solo, or otherwise, and proceeding to learn that piece of music by ear and get to the point where you can play it on your instrument. Transcribing is an essential process in learning jazz improvisation.
Methodology & Research Question:
As a jazz musician, I am an avid user of Anytune, and I frequently use it to loop and slow down certain sections of recordings that I am transcribing. When finding out about the quarter-long project for the Usability Techniques class I took, I thought that Anytune would be a great app to analyze for the project due to its complicated interface:
A key of the relevant buttons on Anytune's interface. Take note of these buttons, as I will refer to them in this case study.
Going into this project, our group wanted to analyze the usability of Anytune’s tools to change speed, change keys, and loop certain sections of music. I then devised a list of tasks for 5 test users that I recruited to help us do this:
Navigate to a certain section
Loop the music over a certain interval
Slow the recording by a certain amount
Change the key of the recording
Undo all speed and key changes they have made and transcribe the section they looped
Loop another part of the recording
It became clear early on in the interviewing process that the looping task was by far the most complicated and yielded the most interesting results among the test users.
We ultimately focused on this research question:
“What is the learnability of the loop function of the Anytune application?”
Findings
After coding our data and discussing our thoughts amongst ourselves, we present the following findings:
Similar Iconography Creates Confusion
The way to create a repeating loop in Anytune is as follows: You scroll on the audio file on top, press the A and B Markers where you want the loop to start and end, and then press the Loop Mode button to make Anytune loop the music between the A and B marks.
This may seem simple on paper, but in practice, users in the study frequently confused the A and B slider buttons for the A and B Markers, often pressing the slider buttons first likely due to the fact those two sets of buttons look similar. This problem is also present with the Loop Mode and Add Loop Mark buttons. In our interviews, users repeatedly pressed the incorrect button that looked similar to the proper button they were supposed to press, which added to their confusion. Users who had clicked on the incorrect counterparts of the proper buttons struggled to complete this task noticeably more than the others. The similar iconography was detrimental to the learnability of the looping function.
Lack of Feedback in the UI
Anytune’s UI does not give enough feedback to reaffirm or educate users about what they are doing. This is especially problematic when inexperienced users are trying to complete a multi-step task (i.e. creating a repeating loop).
In summary, there was little way for the users to know whether or not they were doing the right thing due to the lack of feedback given to them by the app. There were instances when participants pressed the correct button, but due to the lack of feedback, they thought they had made a mistake. Those users proceeded to play around with the interface, and only after they circled back after some time did they get it right. Furthermore, a participant clicked on the incorrect button and eventually figured out how to get the loop working. This participant finished the task thinking that pressing the incorrect button was a necessary step. Learnability was negatively impacted severely due to the lack of UI feedback.
Similarly Functioning Elements Create Confusion
Anytune’s current UI has both different ways to do the same thing and different buttons that do something only slightly different. An example of the former occurring is when opening up an audio file in Anytune, there is an A and B marker already present at the start and end of the song respectively. One participant saw this, and proceeded to drag the A and B markers to where they needed to set up their loop, despite scrolling on the audio file to the appropriate place and pressing the A/B markers being a more efficient way to complete this task.
An example of the latter is most prominent with the Add Loop Mark button. Pressing it adds A and B marks where the current A and B marks are located, but they don’t function the same way. To make things worse, the words “Added Loop Mark” appeared on the screen after pressing this button, which misled multiple users.
Recommendations
Interaction Feedback for all Functions
This recommendation is almost self-explanatory, and thus needs little explanation. Anytune in its current state gives very little feedback to users when they engage with the interface. When pressing the A Marker button, even a small message like “Loop Starting Point Set” message would be extremely helpful for newer and experienced users alike.
Clear Iconography
This is another simple fix that would go a long way to improve Anytune’s learnability. Some iconography, specifically the Slider buttons and the Add Loop Mark button look too similar to the A/B Markers and Loop Mode buttons. As previously stated, test users repeatedly had issues clicking incorrect buttons that looked similar to the correct ones. Updating the iconography on the buttons would significantly clarify Anytune’s UI and improve learnability.
Removing Slider Buttons
The biggest indictment against these two buttons is that their intended use itself does not do enough to warrant their existence. The way these buttons are intended to work is that when the user presses the A or B slider button, they can use the scrolling wheel above the Loop Mode button to adjust the location of the A and B marks. The more intuitive way to do this is to either scroll on the audio file to the desired place and press the A/B button or drag the markers on the audio file directly.
Removing these buttons de-clutters the UI and removes an unnecessary way to move the A/B markers.
Active Guides with Functionality & Processes
While the process of making a loop is fairly easy to get the hang of, the fact that the process of making a loop can be done in different orders can be overwhelming and confusing to someone who is trying to figure it out for the first time.
To streamline this process, it would be beneficial to add a setting that has users select where they want the A and B Markers and press the Loop mode button. This would help newer users navigate Anytune’s complex UI.
Limitations
Narrow Scope & Quick Turnaround Times
This project was very fast-paced, and we only had a few weeks to go through the entire process. We only had enough time to focus on one main feature and leave out other more complex features as a result.
The Structure of the Study
The study had users perform specific tasks, like setting up a repeating loop, while we observed and noted their actions. This approach was due to time constraints and our preconceived ideas about Anytune's core functions. However, these directed tasks may not represent how users naturally learn the app. Small language choices in our instructions, such as "create a loop" instead of "set up a loop," may have influenced participants' actions.
Narrow Population
While we consider our study's findings well-documented, there's a notable limitation related to the selected musicians that we find important to address. All participants are jazz musicians with varying degrees of familiarity with transcribing. Their prior experience may have influenced their understanding of looping music, especially since they've used different software for similar tasks before. Additionally, because they share similar musical backgrounds, it may not be reasonable to assume that musicians from different genres would perform similarly in our tests. Therefore, our sample may not be representative, limiting the generalizability of our findings. However, it's uncertain how much a user's musical background would impact their interaction with the application.