Solution for March 2004

"That could be one approach."

Egad, this turns out to be a hard spectrogram, because there's a dearth of positive clues, and a lot of ambiguity. Welcome to the real world, gang.

Eth + Raising Sign
[ð̝], IPA 131 + 429
Well, there's not much here, except the one pitch-period-or-so of frication. But since the vowel here (from 200 to 325 msec or so) doesn't seem to start with creakiness, there must be *something* here. (Or this word would be vowel initial and get a glottal onset.) The frication (rather than a sharp release) suggests a raised fricative. Utterance initial eths often 'strengthen' to (inter)dental 'stops', but with fricative rather than transient releses. So safe bet.

[æ], IPA 325
Lowish vowel (higher than 500 Hz F1), not outrageously back (or the F2 would be lower), so this could be something lower than mid and fronter than, well, back. Good candidates are [æ] or [ɛ].

Lower-case T
[t], IPA 103
Nice little gap (okay mushy in the lower frequencies, but whatever), apparently voiceless with a nice sharp release. So this is almost undoubtedly a voiceless plosive of some varaiety. The release is a little ambiguous, being strongest in the F3 region. A little low for alveolar, a little high for bilabial or velar. The transitions aren't really helping. No obvious downtrends like bilabials, no obvious pinch like velar, no obvioust lift to F3 like alveoalrs. But the F2 doesn't seem to move at all, and it's overing sort of above 1500 Hz. This is near the locus for alveolar transitions, but it would be nice to ahve some serious positive evidence. How about this. How many words like 'thep' or 'thek' can you think of?

Lower-case K + Right Superscript H
[kʰ], IPA 109 + 404
Well, part of the problem with the previous plosive was that there's some coarticulation with this plosive going on. So this one is definitely a plosive too, but its cues are also sort of ambiguous. The F3 in the following vowel seems to point up (into the plosive, i.e. it may be falling. and F2 seems to point down, if anything, but not a lot of useful movement is going on. Okay, well, we know its plosive, and that 60 msec or so of aspiration clearly suggests voiceless (and aspirated, to be precise) The release burst is sort of long, possibly doubled, and centered, sort of, in the F2 region, more characteristic of velars than alveolars, but bilabials can't be ruled out either, except by the transitions. But speaking of the transitions, I'm really unhappy at the absolute *absence* of velar pinch in the aspiration. What's up with that?

[ə], IPA 322
Short vowel, mostly transition in F2. Call it schwa and move it on.

Lower-case D
[d], IPA 104
Well, if I didn't know better, I'd suggest this was a nasal. But it's not. It's not really resonant enough, considering how voiced it clearly is. But good guess. So if it isn't a nasal, it must be a plosive. I guess. Again, transitions aren't telling us a great deal, except that they're 'consistent' with alveolar. At least it can only be voiced.

Lower-case B
[b], IPA 102
Ditto this, as voicing is concerned. This is a long, long stretch for a single voiced consonant. There also seems to be some kind of amplitude discontinuity just before 700 msec, if that means anything. The release (at about 750 msec) is a little clean to be alveoalar, even though it seems to be broad band and concentrated, if anything, in the high frequencies. But the transitions into the following vowel are totally inconsistent with that. The F2 clearly starts quite low, and the F3 and F4 all definitely point down (toward the plosive, i.e. they rise into the vowel), which is most consistent with bilabials.

Lower-case I
[i], IPA 302
Well, the F1 doesn't move a lot, it seems to stay sort of low. So this is a fairly high vowel throughout. The F2 extremum (800 to 825 msec) is way high, at least 2200 Hz. The only thing that ever gets that front is [i]. And I don't usually produce offglides that front. So this is probably [i] and not [ei] or something like that, and the F2 movement is entirely coarticulation. That's my story.

Lower-case W
[w], IPA 170
Well, the swooping F2 can only mean extreme backness and rounding. The loss of ampoitude in the higher frequencies suggests initial /w/, although there's really nothing here to rule out a full-on [u], since a very tightly-rounded high [u] could damp the higher frequencies like this. The fact that there's another vowel on the other side might sway the decision.

Turned V
[ʌ], IPA 314
This is short, and could be another schwa, but the F1 definitely approaches the lower vowel range (compare it with the first vowel in this spectrogram), and the F2 doesn't rise they way it might. So this is farily back, even at the end. Lower-mid, and back.

Lower-case N
[n], IPA 116

This is a nice little nasal. Sharpish edges, but fully voiced and clearly resonating. Nice little zero at about 750 Hz, and a few more as you go up). The pole is up around 1200 Hz or so. This is a little low for my [n], but is a little high for my [m]. No apparent velar pinch, nor bilabial transitions, so in the end this is consistent with [n]. If it were a little shorter, it would look like a nasal flap, so it would definitely be alveolar.

[ə], IPA 322
Short vowel. Moving on.

Lower-case P + Right Superscript H
[pʰ], IPA 101 + 404
Another plosive, followed by a sharpish release, and some pretty obvious aspiration. So voiceless and probably word initial. The transitions from the preceding schwa suggest bilabial, which is consistent with the release information. So that's enough of that.

Turned R
[ɹ], IPA 151
Well, even with the aspiration, we can see a seriously low F3 at the left edge of this vowel. So there must be an /r/ here. Finally, something sreighforward. Please notice that moment just before 1400 msec, where the F2 reaches a local peak. At abou same moment, there's amplitude change in the space between F3 and F4, and I think a bandwidth change in F3. So that's the moment I segmented off the /r/. Something 'changes' here. Exactly what, I'm not sure, but it's as good a landmark as any.

Lower-case O + Upsilon
[oʊ], IPA 307 + 321
So starting at the moment before 1400 Hz where everthing changes, F1 seems to be mid, and drifts downward a little. So this vowel goes from mid to higher-mid. The F2 starts sort of low and gets lower. So theres movement from backish to backer, and/or roundish to rounder. Maybe both. Ignore the F3 transition, which is just transition. Mid and backish to higher and rounder. It's worth noting, I guess, that there really seem to be two targets here, and two different pitches of voice too.

Lower-case T
[t], IPA 103
Ah, gaps. Well, this one isn't bad, from the point of view of transitional information. There's a nice little closure transient, for once, between 1525 and 1550 msecs, and the good news is that the energy in F3 is definitely higher than it was when the harmonics in F3 seekmed to turn off. So what we have here is a rising transition. F2 rises as well, and that's pretty typical of alveolars. That and the following fricatve pretty much make this a dead cert.

[ʃ], IPA 134
Sibilant fricative, which by the way means it's high amplitude and high frequency, Actually, I think 'sibilant' means that it's produced by directing a jet of air at the teeth, but this is acoustic phonetics, not articulatory. This fricative, unlike the prototypical /s/ is not obviously highest in apmlitude in the very high freuencies, but definitely has lower frequency centeres, in the F2 and F3 region. Also the energy drops off sharply below the F2 region, leaving basically no energy below 1500 Hz or so. Typical of Esh. Note also the sharp onset upon release of the preceding plosive. So this and the preceding plosive together are an affricate [tʃ].