The first five sound pieces on this CD were originally presented at the Good Works Gallery in Spokane, Washington during November of 2004. The narrative from that show is reproduced below and should be understandable even outside the gallery setting. Three additional pieces (6, 7 & 8) are on this CD as well. Only song eight requires some additional explanation. That's added on at the end of the narrative.


What’s going on? Some background for those of you who are interested in such things.

While reading a book about sound it struck me that changes in atmospheric pressure are really very similar to the sounds we hear every day. Our hearing is definitely sensitive enough to hear it, but we don’t. The difference is that the force atmospheric pressure exerts on our ear drums occurs too slowly for us to hear. In other words, the atmosphere doesn’t vibrate fast enough.

A once per day change in the barometer (pressure) is something like 0.00001 cycles per second (1/86,400 seconds per day). About the lowest we hear even in our prime is 20 cycles per second. So atmospheric vibrations would have to be speeded up about two million times (20/0.00001) for us to hear them. To make the sound of middle C on the piano (261 cycles per second), pressure changes would need to be speeded up over 26 million times.

But hey, no problem. I could just simulate a speeded up atmosphere on my computer to hear what it would sound like. What are computers for anyway? And doesn’t the government collect all this weather data and post it on the internet?

Well to make a long story short, this problem turned out to be bigger than the both of us. Me and my computer that is. I’ll skip the gory details of how I came to this conclusion. Hint: obscure problems require obscure software with not so obscure prices. But rather than giving up completely I did what I think is the next best thing. I “sonified” 1800 days (5 years) of pressure data.

Let me explain. Turns out that I’m not the only one who has thought about this type of problem. A discipline of sonification has grow up in the last several years brought about by the availability of more powerful personal computers for processing sound data and the increased complexity of the information available today. So what’s sonification?

The clicks of a Geiger counter and the pings of sonar are examples that predate the word "sonification”. The beeeeeeeep of the heart rate monitor flat lining is a good example that’s frequently used in the movies. To put in the simplest terms, sonification is defined as the use of non-speech audio to convey information. If you are applying for a grant, The National Science Foundation definition works better: sonification is the transformation of data relations into perceived relations in an acoustic signal for the purposes of facilitating communication or interpretation. Sonification tends to be interdisciplinary, integrating concepts from human perception, acoustics, design, the arts, and engineering.

While I was at it, I also collected heaps of temperature data and sonified that as well. My rationale? Our ears are almost sensitive enough to hear individual molecules of air hitting our ear drum. In fact, if we were just a bit more sensitive we could actually hear changes in temperature, which makes molecules move faster. “Gee, SOUNDS too hot out to mow the lawn”. I got caught up in this data frenzy and collected even more and sonified that to: total minutes of sunshine per day and average daily precipitation even though these don’t have the same direct link to hearing.

Now the project began to take on a little different shape, as projects are prone to do. And I, as I am sometimes prone to do, resisted until I realized the new direction was pretty interesting. Maybe even more interesting than a buzzing atmosphere, even one at middle C.

And new questions came up. Doesn’t all this data really represent what we call weather? What does Spokane’s weather sound like? What does Seattle’s sound like? Washington D.C.? Well, never mind that one. Can you HEAR the difference between Spokane and Seattle weather? Could there be a weather CD comparing different cities? The earth also vibrates very slowly. It’s called earth hum. What does that sound like? But that's yet another story.

The next step was to organize all the data I collected. For example, 1800 days worth (5 years) of Spokane’s daily barometric pressure are shown on the long graph over the top of the speakers. This thing even looks like a sound wave. All four weather measures were taken from the same day so for each pressure data point on the graph there is corresponding temperature, sunshine, and precipitation data that is not shown.

Then unique sounds, in this case musical instruments, were assigned to each of the four weather elements.

Pressure was assigned to a bass (or mapped onto the bass as they say in the sonification biz). The greater the atmospheric pressure (barometer reading), the higher the bass’s pitch.

Maximum daily temperature was mapped onto a grand piano. Higher temperatures correspond to higher pitched notes on the keyboard.

A high C whistle (like a flute) was mapped onto total daily minutes of sunshine. The more sunshine, the higher the pitch.

Metal mallets (like a cowbell) were mapped to total daily precipitation. Again, the more rain and snow, the higher the pitch.

The sequence of “songs” on the demonstration CD that is playing goes like this. The first song is Spokane weather speeded up 657,000 times. It starts with each of the instruments just described fading in to give an idea of what to listen for. Then all four play at once like really happens with the weather.

The next cut is Seattle. It’s also speeded up 657,000 times and the instruments are mapped identically to Spokane.

Then both the songs are repeated, Spokane then Seattle, but slowed down to only 175,000 times faster than normal weather.

The final cut is a more interpretive version of the first Spokane song. This time different instruments take turns on playing “sunshine”, which is treated like the lead part for the purpose of this song.

It’s worth stressing that these rhythms are not manipulated other than being speeded up and having a sound assigned or mapped onto them. No “notes” have been added or subtracted and the time value was not varied--they were consistent for all songs all the way through at one data point per day. Only the weather data changed over time: barometric pressure, temperature, sunshine, and precipitation. There was no attempt to bias the songs with titles like “Mold in the Mist” or “Sunburn and Frostbite” although I do have my favorite.

After some practice it is possible to hear the difference between Spokane and Seattle. Here are some things you may want to listen for. Seattle pressure is higher and more variable than Spokane’s and there is twice as much precipitation in Seattle compared to Spokane. Also listen for passages where you hear the combined higher pitch of the whistle, piano and the bass along with the absence of many mallet sounds. These are long periods of clear, hot, high pressure summer days.

After listening for a short while certain instruments will not stand out as much even though the physical loudness is constant throughout (loudness was not mapped although it could be). To me it seems analogous to how our interest in the weather waxes and wanes depending upon the nature of the outdoor activities we have planned.

So this might be what our weather would sound like at warp speed. Beam me up Scotty. It might even be more like what insects hear since their sensory range and sensitivity is vastly different than ours. How long is a cricket year anyway?

When I started this project I thought I might find some interesting patterns that aren’t readily apparent at normal speed. I wasn’t disappointed. The longer you listen to these sequences, the more you hear. In this sense these natural sound patterns have the characteristics of good music. They grow on you.

To the degree these weather sonifications sound like music (jazz?) they bring up questions about why there is a similarity of musical rhythms across various cultures. Could it be we have some fundamental sense of these nature rhythms buried deep in our genetic code? Or does our brain (or maybe our musical experience) lead us to impose patterns where very little pattern really exists?

Addendum: On Spokane v. Seattle: Rain (song 8) each city's precipitation data series is played at the same time for each day by different instruments. The statistical correlation between the two series is 0.48, low but statistically significant. Does the correlation sound lower or higher than this or does it sound about right? What would a 0.0 statistical correlation sound like?

For more information or to make comments and suggestions you can email me via my web page, To try your own hand at sonification log onto Bruce Walker's Sonification Lab page at And thanks for listening.

Tom Dukich

back to songs page