Bird song sonographs show distinct drawing patterns

I don&#x27;t want to detract from the images, or the work that went into this, but, er, there are a <i>lot</i> of types of birds. Just among passerines (songbirds), we&#x27;re talking over 6,000 species.<p>So any sweeping statement like &quot;birds don’t seem to bother to create a complex multi-layered harmonics pattern&quot; is practically guaranteed to be wrong. And so it is. Lots and lots of birds sing incredibly harmonically complex songs. Browse any of these (<a href="https:&#x2F;&#x2F;www.remoteenvironmentalassessmentlaboratory.com&#x2F;explore&#x2F;birdsong&#x2F;" rel="nofollow">https:&#x2F;&#x2F;www.remoteenvironmentalassessmentlaboratory.com&#x2F;expl...</a>) if you&#x27;re interested - it&#x27;s a tiny sample of birds, and many, many of them do in fact have harmonically complex songs.

It makes sense to me that bird song would have few harmonics. Although we may appreciate the beauty of bird song they don&#x27;t sing for aesthetics. They are signaling. Concentrating energy into a narrow band delivers greater range for the same energy than harmonically rich signals. We do likewise with electromagnetic communication.

&gt; <i>However it would be a mistake to to call flute sound simple: as you see, every level has its own regular pattern that can’t be recreated with a simple mix of sinusoidal tones.</i><p>The resonance of the flute&#x27;s air chamber is driven by the noise of forced air stream, which is why it contains high frequencies at all. Of course you can&#x27;t recreate the flute sound just with some periodic sinusoidals, because the noise component has to be present. The noise is complex and is itself filtered by the flute&#x27;s chamber, in a different way depending on which valves are closed. You can play a recognizeable scale on a flute without getting a tone out of it, just using air noise, the same way you can produce a musical tune using chhhh sounds out your mouth. Those notes, or something like them, are still there in the background when a tone is being produced. There is no flute tone without them.

Those are intersting. They remind me of the sonographs we collected of bat echolocation calls when we were doing bat surveys with anabat detectors. Just by the shape of the calls you could narrow the bat down to a likely species and determine whether it was just flying around, actively hunting or about to eat a hapless insect.<p>It was a pretty awesome device.

&gt; I don’t know if it’s a feature or a defect of the instrument.<p>It&#x27;s a feature<p><a href="https:&#x2F;&#x2F;monoskop.org&#x2F;images&#x2F;a&#x2F;af&#x2F;Gough_C_2007_The_violin_Chladni_patterns_plates_shells_and_sounds.pdf" rel="nofollow">https:&#x2F;&#x2F;monoskop.org&#x2F;images&#x2F;a&#x2F;af&#x2F;Gough_C_2007_The_violin_Chl...</a>

Um, I&#x27;d like to be able to parse this article. but I don&#x27;t know what it is saying. Can anyone explain the gist of it to me in simple terms?<p>What does this mean?<p>&quot;These sonograms are remarkably different from other sounds, as if birds “draw” with sound something that’s flying backwards in time.&quot;

For a moment I thought they would hypothesize that the patterns might be an actual “rendering” of what the birds were looking at, especially when I saw the 7th image which looks remarkably similar to a rodent.<p>Apparently I&#x27;m pretty far off, but just this idea of animals communicating images with sound waves was something I&#x27;d otherwise never have.

Dupe: <a href="https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=27030171" rel="nofollow">https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=27030171</a>

This is amazing. The GitHub repo has more information too, about how the common explanation of how our ears work (isolating sounds in an FFT-like-process) may be incorrect:<p>&gt; Contrary to what you might think, our ears don&#x27;t seem to rely on an FFT-like process to extract isolated frequencies. Instead, our ears detect periodic parts in the signal, although in most cases those periodic parts closely match the FFT frequencies. There is a simple experiment that proves this point:<p><a href="https:&#x2F;&#x2F;github.com&#x2F;soundshader&#x2F;soundshader.github.io#why-acf-matches-our-perception-of-sound" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;soundshader&#x2F;soundshader.github.io#why-acf...</a>

I choose to imagine they’re API endpoints communicating lists of entities with similar data causing the patterns.

Slight related tangent: For all those interested in birding, I highly recommend the BirdNet app available on Android and iOS! You can record and send audio data, and it will try to classify the bird for you based on the recording.<p>In the process of recording, it will show you the bird&#x27;s spectrograms, which is really cool!

Snek! <a href="https:&#x2F;&#x2F;soundshader.github.io&#x2F;hss&#x2F;gallery&#x2F;bird&#x2F;3.png" rel="nofollow">https:&#x2F;&#x2F;soundshader.github.io&#x2F;hss&#x2F;gallery&#x2F;bird&#x2F;3.png</a>

I wanted to add that the nature of those thin horizontal layers and the patterns they form is similar to diffraction. A laser beam passing thru a thin slit forms a pattern with the sinc-wave distribution of intensity, and so is FFT of a rectangular window used in these spectrograms makes a sinc-wave shape that &quot;diffracts&quot; the &quot;true&quot; sound frequency.

HSV color scheme looks beautiful, but since it doesn&#x27;t seem to have perceptually uniform color change we won&#x27;t see it used in papers much :(<p>I think the rule for audio papers is linear change in color according to energy?<p>So this rep may lead us to hallucinate our conclusions?

Are there AI projects working on mapping bird song into possible meanings to figure out a language?