Friday, October 21, 2011

A thesis post-mortem

During the last couple of years I have been planning, writing and developing the thesis for my Masters in Sound and Music Computing. In this post I want to highlight some of the lessons I learned on the road, and (hopefully) provide some useful tips from my experiences for anyone who might be interested in doing a thesis under a related field and die in the process... or oh well, no... better say: enjoy the process ;)

My motivations are also influenced by the theses from my Masters colleagues, and the feedback I received. I think there is a real need for doing more theses in the fields related to Game Audio, Interactive Audio, Digital Media, Sound Design and Analysis/Synthesis/Procedural technologies. Looking at emerging conferences, workshops and industry events, I feel there is a potential to continue working towards what I believe we will see in the next few years not only in games, but in various industries, education programs, academic spin-offs and novel fields of research.

What went wrong
  • The art of planning, generally speaking, is hard to master. Even more if you don't have deadlines. Additionally, planning things that are totally new to you, or novel, is even more difficult. I tried to do several estimations of what would take to finish certain tasks, and I failed at most of them. Though it wasn't hugely critical for this project, I had to remember to myself that the aim was to showcase an idea or the potential of some ideas, and not a shippable product with marketing costs.
  • I feel I didn't collaborated with enough people. A problem I somehow faced was finding the right people to collaborate. Failed at that for some parts of the thesis, but I was also constrained by the fact that sometimes I need to isolate and find the time to focus my energy on getting things completed alone.
  • Project pitches. I also realized that pitching ideas more to colleagues and people closer to me would had improved the way of how some material was presented.
  • Did once, and forgot forever. Some of the important basics were set at the beginning of the project, and then sometimes I forgot to bring them out to the table again to justify some of the decisions. I think it is always good to keep track of where you are coming from and where you want to go with your work.
  • Fail early and often. This is a general tip I forgot in some implementations. If you are trying things that would take an unexpected amount of time and don't know if they will work or not, it is better to make them fail at the beginning of the project, and don't wait for some deadlines approaching.
What went right
  • Open source code, libraries and tools are your friends and provide a good (and free) way to learn. Try not to crack a proprietary software that will give some extra push to the project if you can. I was surprised by the amount of free resources that can be found online. Some of them helped to develop my work, and when I had some doubts about how to implement certain things I had some code to take as reference.
  • Contacting professionals and industry people. I think I was quite lucky with this one, so got response from a good range of people working at companies and other academic institutions. Their comments were really handy and a great complement to what I had in my University.
  • Accomplishing what I felt it was necessary for the project. Sometimes it got quite tough to filter out what I wanted to do with the project because while I was reviewing the state of the art, I was discovering very good resources for ideas. At the end I think that my supervisor, advisor and myself found a good packaged balance of contents for the thesis.
  • State of the art was a fun puzzle that worked out well in the end. I was also happy to find out that some of the known research done already at my University was being used in games and related projects while I was doing the thesis. So, for me it really was a good indicator for the state of the art and helped to support my review of technology, algorithms and methods available. It also helped to work on a topic that appealed to me a lot and where I had some experience to bring out.
  • The acceptance of a co-authored white paper and my attendance to the Audio Mostly conference was also a good experience. There I met some like-minded people, had the chance to pitch and talk again about my thesis, and received more feedback about it.
Some lessons learned, tips and conclusions...
  • Being positive and focused. Though at some points I felt like an outcast with the project, looking at it from outside and asking for feedback apart from the University, gave fresh air to the work I was doing and kept me motivated.
  • Be aware of the scope of academic projects and technology. There will be times when you will not have the time you think you have, neither the expected working (bug free) technology at your disposal. Expect technical problems, planning delays and also be aware that writing things to Latex, Word or plain paper... also takes time and is also an important part of the process.
  • There are chances that you will be mainly alone doing this. Your tutor, supervisor or advisor would be of great help but you will likely be the lead of the project so, try to prepare for that. Maybe this is not that critical for a Bachelor's project, but when doing a Masters, you will likely be expected to take things to the next level.
  • Share your knowledge and tools with your peers. Maybe you are using a library or framework that would be of great help for another colleague. Don't keep the guns for yourself. Maybe you can also discover an useful tool for your project on the road...
  • Ask people, people near you, and people far from you, for feedback. If you target to bring out some novel stuff to the masses, be sure it is also going into an appealing direction for yourself. No pressure, no sweat... just try to be sure what you will be working on is challenging enough and (if possible) novel to the community, so to feel it is worthy of doing to keep you continuously motivated.
  • Don't forget to include a future work section in your thesis final report. If the thesis has a potential for others to continue the path you have opened, don't be shy to write it! Use your imagination here.
  • Have fun and make time for yourself. Writing a thesis can be absorbing but by no means you should get rid of time to rest and for yourself. I know it could get tough some times when you have quite a bit of pressure over your shoulders but, make your homework, and try to find time for you.
  • Again, fail early and often. You can, and should make errors. If you are in academia, errors should be supported and encouraged (well, this is also really appreciated in industry as well...). Making errors and learn from them is an important part of the learning process.
  • Try to develop your intuition. Being into a learning process means that there are things for sure you will not know. But I think that by no means this should be taken to diminish or ignore the previous experience you may have, which is also valuable.

As a side note, I have to remark that all of what I wrote above is highly influenced and biased by the University where I did the thesis, the Masters (this was a MSc, a Master of Arts or MBA would be different, I guess), the country where I did it, the people I have collaborated with and my general experiences. Nevertheless, I think some of the tips mentioned can be applied to different contexts.

So, thats it! Please feel free to email me if you think I could be of any help to info [at] jorgegarciamartin [dot] com

And by the way, if you have time I do really appreciate receiving feedback!

Hopefully my next blog post will be a final entry telling a bit of a story from my current job search, and - spoiler in 3, 2, 1... - it will also conclude this blog :)

No comments:

Post a Comment