I Made a Song with Scrum

When I was 15, I received a free copy of Acid Music Studio while skipping out on Third Eye Blind at an alternative music festival in Chicago.  I spent weekends in the basement with my friends, arranging pre-made loops with delusions of being the next Chemical Brothers.  For my 16th birthday, I convinced my grandmother to buy me a Roland MC-505, leading to hours spent with the Groovebox sequencing bass lines and tweaking filters.  I continued writing music and djing through college, eventually teaming up with Morgan Hendry on a DJ show for USC’s college radio and sporadic live electronic performances.  I had spotty motivation and no great talent at finishing.  Generously, I would say that I spent a lot of time sketching.

During college, I began my professional career in the video game industry after landing a bottom rung job as a production intern and tester at Pandemic Studios.  After college, I moved up to San Francisco to pursue game design & production full time, my music output dwindling from a tiny, backyard creek to the maddening drip of a shower valve that won’t quite close.   I would occasionally finish DJ mixtapes for my own listening pleasure, but that was it.

In the past two years at BioWare San Francisco, I have become heavily steeped in Scrum Methodology as a project management tool.  We use scrum to turn the creative, team oriented, constantly shifting software development process into something that is clear, measurable and accountable.  Somewhere along the way, I felt the music itch and had an interesting thought.  Could scrum be used to help me write music?

I imagined applying the process to music for a long time, but it was little more than a pleasant fantasy I to turned to when stuck in traffic for over a year.   This January, I finally took the plunge.  I bought a copy of Ableton Live Suite and dedicated a week of vacation to writing and finishing one song, using scrum to manage the project.  The week was equal parts frustration and invigoration.  I faced setbacks and came close to giving up on multiple occasions, but in the end was able to finish the song shared above.  In true scrum fashion, this article is a postmortem on my sprint.

What went Right

1. Clear Goals

I began the sprint by writing a work brief outlining my idea for the song “One of Us”.  I had never approached a song by asking myself “What do I want to accomplish?” the way I regularly did as a game developer.  Setting a clear intention and goal state built a foundation for the project.  Without setting an intention, I would never have finished a song, likely spending a few days sketching randomly as in my early electronic music days.

2. Sprint Commitments

I wanted to give up.  A lot.  A week of vacation time is precious, and I strongly considered abandoning my project and spending the rest of my vacation on the far more attainable goal of beating Uncharted on the PS3.  But, a cornerstone of the scrum process is that you make a commitment to your team to finish your tasks during the sprint.  Reminding myself of this commitment kept me from throwing my laptop at the wall, and drove me forward each time I wanted to give up.

3. Ableton Live

It had been several years since I last learned to use Logic Pro Audio as part of a college course on electronic music.  Learning a new audio suite and writing a song in a single week was a daunting amount of work.  Luckily, Ableton Live is easy to pick up if you are familiar with audio sequencing, and Ableton hosts some great getting started videos.   Ableton is a very powerful program and I was able to learn the basics and write a professional quality song with minimal heartache, including writing probably my all time best drum loops.

What went Wrong

1. Environment Setup and Production Pipeline

I spent more time during the sprint learning the software, getting my production environment setup, running to Guitar Center and wrestling with hardware then actually writing music.  This is partially to blame on the poor decision to try and get an old, Mbox 2 working instead of purchasing a new audio interface in the middle of the week.  But, I should have spent a sprint before this project acquiring all necessary hardware, getting my production environment set up and learning the program and pipeline.  I would have written a much better song if I had split up the onboarding tasks and creative tasks into two separate weeks.  In Scrum terms, I should have identified and resolved all of my dependencies before starting my song writing tasks.

2. Vocals

My vocals were not good.  Scoping a 3:30 minute song with recorded vocals and mixed audio was overly ambitious given that it has been years since I sang on a daily basis and was in any sort of shape to record vocals.  Also, much of the time spent wresting with hardware was to enable the recording of vocals.  If I had set a less ambitious goal, I would have had more time to focus on the quality of the final song.  Instead, I feel like I am sharing a decent first draft at a song that needs a lot of work.  I wish I had programmed a better lead patch, had found a catchy lead line, had programmed velocity, embellishments and effects into the track, etc.  Instead, I spent a large portion of the sprint writing and recording vocals that are not worth sharing.

3. Not Quite Dubstep

I set out to write a dubstep song with NIN style vocals.  I missed this high concept pretty drastically.  In addition to unsharable vocals, the track that I did write is almost dubstep, but not quite.  It really feels more like some late 90s drum and bass with an updated, warble on that bass line.  Drums are historically my weakest aspect of production and, in a vacuum, I am very happy with the drums in this song.  But they are not dubstep drums.  Also, with more time programming I could have made the song sound more traditionally dubsteppy by changing the speed of modulation on the bass warble.

Conclusion

Relative to the goals I set out to achieve, I would rate my song a 2/10.  However, I view this project as a remarkable success.  I finished a song for the first time in years.  You can listen to this song and share it on the internetz.  I learned a new piece of software and now have the skills and knowledge to pursuit my musical hobby with a renewed vigor.

Most importantly, I know that scrum works when applied to creative projects outside of game development.  Without the clear, measurable goals, daily check in and focus on commitment that the scrum process enforced, writing electronic music would have remained a fantasy for traffic jams and boring meetings.  Thanks to scrum, I finished a song and hopefully, will finish more in the future.

Stats

  • Sprint length – 5 days
  • Story completion rate – 75% (6 of 8 stories)
  • Hour completion rate – 80% (64 of 80)
  • Project Cost – $930.42
  • Song Length – 03:42

Further Reading

What is Scrum?

Scrumification of Music – Day 1

Scrumification of Music – Day 2

Scrumification of Music – Day 3

Scrumification of Music – Day 4

Scrumification of Music – Day 5