 Let's identify the sound segments exhibited in the following spectrogram. Maybe you want to pause the video here to use the printout and try on your own first. Okay, and this is how I would approach the analysis. The spectrogram obviously exhibits two parts separated by a considerable portion of silence. Both parts have a final voiceless fricative with high frequency friction noise, thus they are likely to be voiceless alveolar fricatives. Each part involves a relatively long vowel of about 250 milliseconds. The first vowel has a low F1 value and a high F2 value, so it looks like an E. The second vowel must be a low one, that is R or O, since its first two formants are close together and can be located at around 600 and 1000 hertz. The initial elements of both parts must be voiceless plosives, which are both aspirated. They both exhibit a significant burst. The burst frequency of both plosives involves a wide frequency range, but the first one is slightly lower than that of the second, so they must be different. And the first one looks like a bilabial plosive, the second one must be alveolar. This frequency difference can also be seen in the different F2 origins of the vowel that follows. And then there is another plosive before the final fricative. It involves a high frequency burst and must thus be vela. So the result is clear. I said peace talks.