Monthly Archives: June 2015

Almost out of the Country

The intertia of a group seems to increase exponentially with its size. Shari and I finally managed to herd the cats to airport and through checkin, security, and food. Now we wait for boarding and post random ramblings to blogs. 

I am proud of Max and Zoe though. When I kissed Shari on her neck, they said “you’re being like Riker on that Q episode.”

Star Trek – The Next Generation is so educational. 


Wait But Why

I found a really smart website, called “Wait But Why” that delves into big issues and explains them really well. The author, Tim, tries to dig past the fuzzy parts and the politics, and largely succeeds.

I enjoyed the Elon Musk articles:

  1. Elon Musk – The World’s Raddest Man
  2. How Tesla will change your life

And the AI articles:

  1. Artificial Intelligence Revolution: The Road to Superintelligence
  2. Artificial Intelligence Revolution: Our Immortality or Extinction

The articles are long, but excellent.

Scuffy The Tugboat – The Process

I find that I learn best when learning is a side effect of actually making something. It’s what makes me an engineer, I guess. So in an effort to learn more about composing music, I decided to attempt to write some music for the Wind Band that I joined a couple of years ago.

The conductor, Danika, was open to the idea and mentioned that the theme to an upcoming concert was Songs of Sea and Sky. I love listening to soundtracks, so I wanted to do a soundtrack. I could take a scene from a movie and write some music to it, but that has the problem of having existing music that will influence whatever I might do. So I decided to take a children’s book with lots of pictures and a story involving water, then make a slide show  of the pictures in the book and write a soundtrack to that. The book I chose was Scuffy the Tugboat“; a book I’ve read many times to my kids.


Danika gave me a good specification for the piece of music. So I started work.

After doing a little reading about the process of orchestrating, or instrumentation, it seemed that the normal way to approach this is to write a reduced score and then expand this into a full score. A reduced score is essentially something like a piano part that contains all the information you need to expand the music into a full score.

My first step was to muck around and come up with a bunch of tunes and write them down in a piano score form, with different ideas for accompaniments and counterlines and stuff. I wrote a different section of the piece for each page of the book. It took a little while, but wrote down lots of ideas in a long rambling score in noteflight. I refined this reduced version in about 6 iterations until I had something that seemed to match the book. I threw out a lot of ideas.

So I started the process of arranging this for Concert Band. I added all the instruments that I needed to the noteflight score. Here is where I hit the first snag. Noteflight is great, and easy to use. I’ve been using it for a few years to write a bunch (gaggle?, murmuration?) of music. It is a web based application and runs in the browser. When I added all the instruments, it started running like a lethargic sloth stuck in cold honey. So I needed to buy a proper professional notation package. After much searching of the internet, I settled on Sibelius and bought it. It’s the most expensive software purchase I’ve ever personally made. But, it saved so much time and was worth it.

There was a learning curve to start using Sibelius, but it wasn’t too bad, especially as I was used to Noteflight. One of the major time savings was that the sounds built into SIbelius produce a half decent attempt at playing the music. Also, It allowed easy sharing of the score via YouTube. Here is one of the first versions I made.

The early versions were very early drafts, and had almost no percussion. I needed to set the music to a slide show so that I could show it to people and they would have some idea as to what I was trying to achieve. I wanted to do a slide show with the effect used in documentaries where they pan a camera slowly over a photo, just to fill in time. Some more searching revealed that this is called the “Ken Burns Effect“. Then I dug around and found a free program that let me do exactly that, called PhotoFilmStrip. So I made the official Version 1 video.

Now I needed some critical feedback. This is not as easy to get as you might think. Everyone wants to be encouraging and doesn’t want to say “this bit is terrible”. But I really value this kind of feedback because it is very easy to get too close to it. So I got some good feedback from my wife, Shari, and a few people in the band.

One of the main early comments was that the different sections of the piece sounded disjointed and unrelated. So I changed a couple of the sections to be variations on other themes from the piece. This certainly helped.

By the time Version 4 rolled around, I decided that I needed to do something about the percussion, which I didn’t know a lot about. Here is where percussionist Andrea Leong was amazingly helpful. She sent me an ever so subtle hint that I should ask her what to do. So I did, and got some awesome advice, including what instruments were available.

After a few more versions, I printed out the music and took it along for it’s first play through. I was a bit nervous, and with good reason. The first rehearsal exposed a few things about the music that are not so obvious when you are working with the computer, and also when you are very familiar with the story.

I wrote a different section of music for each page of the book, and this had the effect of lots of sections that were not easy transition when you are playing live. I wrote it in a key that was tricky to play in. The parts of the music that I had written in 11/8 confused the hell out of everyone. Also, I’d missed a bunch of dynamic markings. Also, we added a few extra dynamic flourishes. Some of the notation was unclear in its articulation.

However, on the bright side, using SIbelius meant that all the transpositions were error free, and the harmonies sounded correct.

After the first  rehearsal, I transposed the entire piece down a tone to make the key signature much easier. This did help in the next rehearsal.

Percussion still needed some work, and my saviour was Andrea, who took home the percussion part and re-notated it so it was notated how she liked it, and in the process modified it a little. She made it sound great and, more importantly, playable with only two hands and two feet. I also read a great book on percussion around this time.

The next tricky part, was the section with the cows.


Even though I said it was to be played with a “Bovine swagger”, 11/8 swing was proving unfamiliar and difficult for everyone to play. So I converted it to 4/4 swing. It made all the difference. After 3 runs through and they had it.

It was the second last rehearsal before the concert and for the first time, the band finally played Scuffy The Tugboat all the way through without needing to stop. I was thrilled. I knew then that  it would work on the day.

I recorded the last run through at rehearsal before the concert and redid the timing on the slideshow to match it.

All was ready for the world premier. I was a little bit nervous.

On the day of the concert, Scuffy was in the second half. The pre-concert rehearsal went well and it matched the slideshow really well. We had a couple of technical hassles getting the slideshow to work, but once that was sorted out. It was performed, and it went well!

I was so happy to have it performed after such a lot of work.

It is the biggest, longest, most complicated piece of music I’ve written (so far :-)). And I couldn’t have done it without the support of the whole band, particularly Danika and Andrea. I’ve learned a massive amount from this project. Now for the next one…

Thank You to all of the Inner West Community Band.

Make It Quick #4 – Stupid AI for Tic-Tac-Toe

It doesn’t sound like much, but I added a random computer player to the Tic-Tac-Toe game I made in an earlier MIQ. I had to refactor the entire application to be less of a monumental kludge and more object oriented. Then I could make the computer select a random move from the list of possible moves. Again, this took closer to 30 minutes than 15.

Here it is: