Thanassis Rikakis
Posted on 2014-09-22 22:01:08 -0400.
I like the changes you introduce between the three versions. It gives you and the audience a chance for informed comparisons amongst the three versions. I am not sure I fully agree with the chance, entropy examples as you present them but as you say; these are ambiguous terms. Thanks for the performance discussion. Interesting. Good work overall,
+0
Amanda Marano
Posted on 2014-09-23 02:04:24 -0400.
Skipping around the track is a really interesting way to change the algorithm. I have some questions about your methodology, though. Why didn't you all make sure your laptop volumes matched at the beginning instead of leaving it to chance, and why did you choose a youtube version of a song with a lot of silence at the beginning instead of a cleaner version? These questions would be interesting to answer in this writeup as well, since they contribute to the overall indeterminacy and entropy of the final product.
+0
Kristen Smith
Posted on 2014-09-23 08:35:43 -0400.
I think the idea of skipping to an arbitrary part of the song as a change to the algorithm is certainly interesting, but it introduces a lot more bias than the project should allow. People probably have a tendency to skip to the earlier part of the song when they have stayed on the later part of the song for several rounds of picking new numbers. It's the same way people are terrible at writing down values for heads and tails of a coin "randomly." They'll usually start writing something like heads, tails, tails, tails, then believe that there are too many tails in a row, so they'll write heads instead. However, when actually flipping a coin, there may be 10 tails in a row simply because the results of the first coin flip has no effect on the second. It seems to be the same concept. If you had used the random number generator to somehow pick a time in the song, the place in the song where you landed would have been left up to chance because there would be no human bias.
+0
Eunice Oh
Posted on 2014-09-23 17:05:55 -0400.
Kristen made a good point in how the third composition certainly did introduce some bias even though it seemed to be randomly selected. We were considering using a random number generator to select different times to skip to, but it would have been difficult with the short time we had to switch around.
As for Amanda's point, matching up the volumes was a little difficult. The phone mic had some trouble picking up some of the sounds, but overall I understand where you're getting at. Also, regarding the Youtube videos, you brought up a good point. The silence in the beginning did throw off the music but it also introduced more indeterminacy in the results.
Thanks for the feedback.
I like the changes you introduce between the three versions. It gives you and the audience a chance for informed comparisons amongst the three versions. I am not sure I fully agree with the chance, entropy examples as you present them but as you say; these are ambiguous terms. Thanks for the performance discussion. Interesting. Good work overall,
Skipping around the track is a really interesting way to change the algorithm. I have some questions about your methodology, though. Why didn't you all make sure your laptop volumes matched at the beginning instead of leaving it to chance, and why did you choose a youtube version of a song with a lot of silence at the beginning instead of a cleaner version? These questions would be interesting to answer in this writeup as well, since they contribute to the overall indeterminacy and entropy of the final product.
I think the idea of skipping to an arbitrary part of the song as a change to the algorithm is certainly interesting, but it introduces a lot more bias than the project should allow. People probably have a tendency to skip to the earlier part of the song when they have stayed on the later part of the song for several rounds of picking new numbers. It's the same way people are terrible at writing down values for heads and tails of a coin "randomly." They'll usually start writing something like heads, tails, tails, tails, then believe that there are too many tails in a row, so they'll write heads instead. However, when actually flipping a coin, there may be 10 tails in a row simply because the results of the first coin flip has no effect on the second. It seems to be the same concept. If you had used the random number generator to somehow pick a time in the song, the place in the song where you landed would have been left up to chance because there would be no human bias.
Kristen made a good point in how the third composition certainly did introduce some bias even though it seemed to be randomly selected. We were considering using a random number generator to select different times to skip to, but it would have been difficult with the short time we had to switch around.
As for Amanda's point, matching up the volumes was a little difficult. The phone mic had some trouble picking up some of the sounds, but overall I understand where you're getting at. Also, regarding the Youtube videos, you brought up a good point. The silence in the beginning did throw off the music but it also introduced more indeterminacy in the results.
Thanks for the feedback.
You must login before you can post a comment. .