Composition 1
In our first composition, our goal was to have low chance, fairly low indeterminacy for our group, and also possibly fairly low entropy. We did this by changing the selection of playback material, giving a choice instruction to each member of our group. Each member of the group then chose a song according to their interpretation of the instruction. The instructions were:
1. A piece with low, sustained instrumental sounds
2. A piece with high, sustained instrumental sounds
3. A piece with middle register, short to medium, vocal sounds
4. A piece with short percussive sounds that cover all registers and are soft
We performed the algorithm as follows:
Each of us generated a random number between 3 and 20 to be our individual “duration” for the 5 minutes we chose to record.
To begin, each of us random generated a 0 or 1 to indicate whether we play our song for the duration picked, then every “duration” number of seconds, we would again generate a 0 or 1 to indicate whether the song will play for the next “duration” time period. “0” indicates to pause the song, while a “1” indicates to play the song.
Our choices above meant low chance since we had a more limited pool of songs to choose from, lower indeterminacy since we had more of an idea what types of songs each of us would choose, and possibly lower entropy.
Our results are recorded in “Recording 1.MP3”. These results do reflect some chance, indeterminacy, and entropy, due to our algorithm and the fact that each of us did pick a song individually as opposed to as a group. However, this result most likely shows less chance, indeterminacy and entropy than if we had each randomly picked any song we had wanted. The songs chosen from the four categories make the final piece sound more harmonious, as all the songs happen to be soft music. And according to musician’s past experience as we discussed in class, the product of combining songs from these four categories will be less discordant.
Composition 2
In our second composition, we chose to attempt to keep chance fairly low (but increased), but increase indeterminacy quite a bit, and also entropy slightly.
We did this by allowing each group member to choose any song from the internet, but used the same algorithm to perform the composition as in the first one. There is still fairly low chance, though higher than in the first composition. We expect chance to remain low since certain songs are more popular than other ones, and we would be more likely to know them. However, chance would increase somewhat since we no longer have restrictions on type of song we can choose. Indeterminacy increases greatly since we now are choosing random songs without informing the other group members of our choices, so the resulting composition will be more uncertain. Entropy increases slightly since the song choices will be more random, so the resulting composition will be more varied.
Our results are recorded in “Recording 2.MP3”. These results do seem to reflect our predictions about chance, indeterminacy, and entropy as described above. Compared to “Recording 1.MP3”, recording 2 sounds more dissonant and contains more abrupt transitions as the songs we chose are more dissimilar, reflecting greater indeterminacy and entropy.
Composition 3
In our third composition, we chose to keep chance about the same, decrease indeterminacy, and possibly decrease entropy.
We used the same songs that we had chosen randomly for the previous composition, so this actually greatly decreased chance and indeterminacy for our group, since we each already knew what songs to expect. However, for another listener who had not yet heard the second composition, chance and indeterminacy would not be decreased.
In this case, though, we changed our algorithm. Instead of each player generating his or her own random duration, we generated one random duration for the whole group to use. This resulted in much less indeterminacy and entropy since start and stop times matched up more for each player.
Our results, recorded in “Recording 3.MP3” reflect these predictions since there is less seemingly random starting and stopping of songs. However, compared to “Recording 1. MP3”, recording 3 is still less harmonious and contains more bad transitions, as the songs we chose differ a lot.
Expected Results
Overall, we expect recording 1 to be the most smooth and harmonious, with low chance, low indeterminacy, and lower entropy. Next is recording 3, which we expect to be much more discordant, with increased chance, higher indeterminacy than 1, and maybe higher entropy than 1 as well. We expect recording 2 to be the most random, which should contain lots of random stops and sharp transitions. It should reflect a similar level of chance as recording 2, the highest indeterminacy, and the highest entropy.
Carrying out these processes made it clear to us that the smoothness and harmony of the final music highly depends on the levels of chance, indeterminacy, and entropy we attempted to create in the making process.