For our first composition, we decided on a length of time, 4 minutes, and each member of the group selected a piece that was about that length, without telling the others what they had chosen. We then each generated our own random number to determine if we would play, or not play, our song when we started. We also each generated our own number between 3 and 20 to determine the intervals we each would play our song for. Laura got (0,8), Kim had (0,5) and Anna generated (1,12). Because of a misunderstanding of the instructions, Kim also generated either a 0 or 1 at each interval to determine whether she would play during the next interval, rather than simply alternating as the other two did. We muted the songs when we were not playing, so that all of our songs ended about the same time. This had low chance, since each of us chose our own songs and we limited it to a song that was around 4 minutes in length. It had high indeterminacy since none of us knew what the other was picking and we had no knowledge of how the finished piece would sound. The result had surprisingly low entropy, because our choices actually complemented each other well and the intervals made them weave in and out in what feels like a very natural way. Interestingly enough, during this silence you can sometimes hear the muffled sounds of other groups in the background, which adds entropy as well as indeterminacy and chance, since we had no idea you could hear others and we had no control over what they were saying.
For our second piece, we decided we would all pick pieces of a similar type: subdued and mainly instrumental. Then we used our generator to pick an interval between 7 and 12. We got 10, so we decided every 10 seconds each of us would generate a 0 or 1 to decided if we should play or not. We also randomly generated the length we wanted our piece to be, between 4 and 7 minutes. We got 5 minutes and 19 seconds, but Laura was reading the timer upside down and stopped at 3:19, so it ended up being shorter. Since we picked the type of music we each would select, this decreased the chance even more than our first, since we were limited even more than just or personal song preference to a specific genre. The indeterminacy was also lower for the group members since we had a better idea of what the others were going to be selecting. The entropy was also lower since the instrumental songs seemed to clash less with each other and since all changes happened every 10 seconds, there was less abrupt changes.
The third piece we wanted to be the most random, so we decided to each select songs based on shuffling our entire libraries. If the first song we got was shorter than the interval, we would simply let it continue to the next shuffled song once it ran out. We each generated a 0 or 1 to decide if we would play or not at the beginning, and then alternated between playing and not. We also generated the length of each interval (between 3 and 20) that we played or muted. We also muted our song when we weren’t playing so that we didn’t know what part of the song was going to play next. Anna was using her phone to play the songs, which only has a slider to control the volume; the inconsistency in volume contributes to the entropy. As in the second composition, we randomly generated a time between 4 and 7 minutes and got exactly 6 minutes. This time we had higher chance than in our previous pieces because besides the limitations of each of our personal music collections, we did not select what we would be playing, or what would come after it. It also, as a result, had higher indeterminacy since no one really knew what was coming next or where we were in the songs we were playing. The resulting entropy is much higher than our other two compositions because the types of songs were very different and the muting means that even each song isn’t continuous, causing an experience that changes frequently and has mismatched sounds competing with each other.