(prior page: Musicality)
When I say a bird song is noisy, I don’t just mean that it is loud. Instead I’m saying that the sound contains a very specific acoustic attribute called noise: aperiodic, essentially random sound that contains energy at many audible frequencies without discernible pattern. Noise makes up the sounds of waterfalls and television static. On a spectrogram, noise appears as a smudgy darkness that covers a wide vertical frequency band:
The call of the Steller’s Jay is made up primarily of noise, and that is why it sounds “harsh” or “rough.” It is unmusical because the sound energy covers so many frequencies that there is no way to ascribe a pitch to it. Therefore, I would describe this sound as basically toneless.
Extra credit: The call of the Steller’s Jay is made mostly of noise, but it is not made of white noise. That is, the sound energy is not evenly distributed across all frequencies. On the spectrogram you’ll notice that inside the call of the Steller’s Jay are some darker bands and some lighter bands; the darker parts are louder (more energy), the lighter parts softer (less energy). Some of the darker bands are slightly upslurred (that is, they rise from lower left to upper right), and you may be able to hear this as a vaguely upslurred voice inside the noise.
Here is a bird sound that has a little bit more voice and a little bit less noise:
Note the dark downslurred bands towards the beginning of the Red-tailed Hawk’s scream. They provide a more musical, tonal quality to the beginning of the scream, which thereafter becomes mostly noisy. However, the downslurred intonation of the scream is noticeable throughout, both by ear and on the spectrogram.
The drumming of a woodpecker is made entirely of noise: one extremely brief burst of noise for every beat of the bill against the tree. These bursts of noise show up as vertical lines on the spectrogram (truly vertical lines, unlike the California Quail calls we saw on the previous page). Note again that some frequencies are darker than others.
It should already be clear that noise is not an all-or-nothing quality, but for good measure, let’s look at a bird whose song starts out clear (that is, free of noise) and ends up noisy:
Notice the gradual transition from clear notes to noisy notes in the song of the Guianan Warbling-Antbird (Hypocnemis cantator). The middle notes have significant noise content, but also a significant tonal whistled component. This gives them a “scratchy,” “hoarse,” or “screechy” quality.
Extra credit: The Warbling-Antbird isn’t the only thing making noise on the recording! Note these three other sources of noise:
- The faint band of constant noise in the background between 6 and 7 kHz is the noise of insects such as crickets and cicadas, a very common sound in the tropical rainforest, where this recording was made.
- The dark band of noise running along the very bottom of the spectrogram is low-frequency background rumbling that could be caused by distant trucks, waterfalls, thunder, ocean waves, or any number of other sources. The lower-pitched a sound, the farther it carries, and sounds below the pitch of human hearing carry for dozens or even hundreds of miles; that is why it is very difficult to get away from background rumbling anywhere on the planet. (But that’s also how elephants and cassowaries communicate with each other long-distance.)
- The faint “smearing” of noise just to the right of each note of the antbird’s song is echo. This bird was singing in a dense forest, so each of its notes bounced off many tree trunks and branches at many distances, creating this characteristic “smear” on the spectrogram. It’s difficult to get a recording free of echo unless you are recording an unobstructed bird silhouetted against clear sky at close range.
If you go back and look again at the spectrograms from earlier pages, you’ll notice other examples of background noise and echo. Remember: every sound you hear is going to show up on the spectrogram somewhere!
(next page: Trills and Beats)