Think you know your flute players??

Irish Flute
1

Hi and firstly thanks to everyone (29 of you!!) who filled out my last survey on computer composed tunes. I got some great feedback even though the quality of the tunes needs some work and my paper was accepted for AICS. Here is the final version of the paper:

http://www.comp.dit.ie/bduggan/Research/A%20Combinational%20Creativity%20Approach%20to%20Composing%20Traditional%20Irish%20Reels.pdf

Anyway, my main (PhD) project is to do with how humans understand musical style and can identify a musician based on their style of playing (and can a machine use the same processes). The purpose of this experiment is to take phrases of music played by three well known flute players, in quite distinctive styles and try and figure out what are the characteristics of style that a human listener uses to identify the musician.

After listening to each phrase of music, please try and identify the musician playing the phrase. The phrases you will hear are all from field recordings of the same tunes (The Golden Keyboard, Ambrose Maloney’s and Jackson’s reel) played by Eamonn Cotter, Catherine McEvoy and Harry Bradley.

Secondly and more importantly, can you please take some time to reflect on the process you went through to identify the musician and write down what was it about the phrase that led you to conclude that it was distinctively in a particular musicians style. There are 30 samples to evaluate. If you find after listening to a certain number of samples, you don’t feel like filling out the comments thats ok, but please try and identify the style and leave the comments blank.

Some ideas you might consider in your answers include timbre, use of ornamentation, timing of ornamentation, timbre of ornamentation, tempo, variation/interpretation of phrases, use of accidentals, legato/staccato attacks breathing/phrasing. Feel free to add your own characteristics, be as detailed as possible and if possible identify the particular part of the phrase that gave it away.

There are thirty phrases in all (ten on each page) and after you fill out the survey, you can hear each of the field recordings (each musician’s interpretation of the three tunes) in their entirety.

Again, dont worry about filling out comments for each of the samples.


http://www.bryanduggan.com/phpESP/public/survey.php?name=CognitionStyle

Thanks again!

Bryan

2

That was a tough one, Bryan. Some were quite distinct, but other clips, being like two seconds from the middle of the phrase were rather hard, I might say. ;D

Interesting anyhow… Good luck with the PhD and hope to see you soon!

3

Thanks for sharing your paper. I’ve only recently started reading this forum, but my Ph.D. was on computational creativity (storytelling in my case) so I found your work very interesting. If you haven’t read my book (http://tinyurl.com/397gkv) or the previous work in computer-based storytelling (e.g., Meehan) you might find it interesting. It may not be directly relevant to your work but there are some parallels. (Ah, I see your advisor has referenced me, so presumably you’re aware of it.)

Some years ago I did some work on composition of trad tunes very similar to what you did, although I only ever looked at the N-gram level and took a somewhat different tack than you did. I’ve been thinking for some time of revisiting it, and I’m curious to hear what language and/or packages you used for your programming. I didn’t really want to write a whole ABC interpreter from scratch, and at the time I wasn’t overly taken with the ones that were available.

One of the things that I did in my early work was to look at the most common fingering changes for tinwhistle/flute across a broad corpus of tunes. I then took the ten most common fingering changes and composed them into a practice “tune”:

X:1
T:Practice Fingerings for D Tinwhistle
M:C|
K:D
FGAB AGEF | DED2

Anyway, thanks again for sharing the paper – very interesting! Feel free to contact me offline if you want to discuss your work on a more technical level.