warning: this is a linguisticky post!
O_O...
i am about to go insane from all these spectrograms. argh. why didn't i listen to everyone who told me to get the bulk of the experiments done during the summer? :( :( :(
30 words x 4 lists x 40 people = 4800 words that i have to look at. i think i'm about halfway through, a bit less, give or take. maybe around 2000.
and here are the things i have learned so far from staring at spectrograms all day for the past couple days:
(1) i like male data; it's sooo much easier to find and measure vowel formants. i think, if i ever have to look at formants again, i would try to only recruit male subjects. would this be considered sexist? :(
(2) i don't actually need to depend on the auto-formant tracker! i can find them with my own eyes!
(3) i'm getting pretty damn good at differentiating between [i], [u], [ae], and [a] without having to listen to the actual files.
(4) did you know that sometimes english speakers actually voice the initial voiced consonants? i've found negative vot for some of these!
(5) i hate finding f1 for [u] and [i]. it's all mixed up with f0. :(
(6) i hate [u], period. there's a serious lack of common english words that have [u].. where speakers will actually produce a [u] and not some weird [i-u] sound. :p and i made -such- a bad judgement call by putting in the words 'cool' and 'pool' and 'ghoul' as part of my word lists... D: oh well.
okay, that's all. back to my endless pile of spectrograms.
i am about to go insane from all these spectrograms. argh. why didn't i listen to everyone who told me to get the bulk of the experiments done during the summer? :( :( :(
30 words x 4 lists x 40 people = 4800 words that i have to look at. i think i'm about halfway through, a bit less, give or take. maybe around 2000.
and here are the things i have learned so far from staring at spectrograms all day for the past couple days:
(1) i like male data; it's sooo much easier to find and measure vowel formants. i think, if i ever have to look at formants again, i would try to only recruit male subjects. would this be considered sexist? :(
(2) i don't actually need to depend on the auto-formant tracker! i can find them with my own eyes!
(3) i'm getting pretty damn good at differentiating between [i], [u], [ae], and [a] without having to listen to the actual files.
(4) did you know that sometimes english speakers actually voice the initial voiced consonants? i've found negative vot for some of these!
(5) i hate finding f1 for [u] and [i]. it's all mixed up with f0. :(
(6) i hate [u], period. there's a serious lack of common english words that have [u].. where speakers will actually produce a [u] and not some weird [i-u] sound. :p and i made -such- a bad judgement call by putting in the words 'cool' and 'pool' and 'ghoul' as part of my word lists... D: oh well.
okay, that's all. back to my endless pile of spectrograms.