spectrum analyzer

Adrift in a Sea of Bells


My cohort Eleanor Mills introduced me to The Soundgarden at Central Park Elementary School several years ago. Eleanor goes there on a regular basis to “wake up the bells.” She has developed an intimate relationship with these bells, their interesting harmonics and how they all speak to and blend with each other. I have been privileged to play along with her on several occasions. Here is a short sample of Eleanor waking up the bells at a recent play date:

The Soundgarden was designed and constructed by Andrew Preiss in honor of Greg Taylor, a local musician and teacher at Central Park School who died in 2007.  It is made of steel cylinder tanks (often referred to as bottles) cut to varying lengths to produce a variety of tones low to high. There are 8 large tanks and 12 tank tops positioned along M shaped bar (see photo). As you can hear from the clip, these rough cut steel tubular “bells” send out a sweet and sour soup of tones. Eleanor has discovered a variety of techniques in her playing that pull a rich and interesting sonic landscape from them.

These days our group prefers to pop up and perform soundscapes in interesting spaces with little notice. So the Soundgarden is a perfect spot for us (once I discovered there was electricity available. As an electronic musician, electricity is a necessity🔌). In order to produce a soundscape that would compliment the bells, I wanted to analyze their harmonic character. So I took my tuner down to the Soundgarden and hit each bell and held up the tuner. Well, the diverse harmonics that spring forth from the bells were just too much for the tuner – it was all over the place and seemed inaccurate to my ear. For example, the two tanks on the right in the photo above are clearly a minor third apart to the ear. Yet the tuner registered A# to F, which would be a fifth. This was a puzzlement.

I found a more accurate method of analysis by recording each bell individually and studying them on a spectrum analyzer. This approach was revealing and somewhat tedious. However, the rewards made it well worth the time spent scrutinizing the spectrum analyzer to pinpoint precise frequencies as they arose and decayed in the bell tone. I was able to track overtones up into the 5th and 6th octaves above the fundamental tone of each bell.  Most of the overtones are enharmonic overtones (meaning the frequencies are not in a whole number integer relationship to the fundamental frequency of the bell), so they tend to be slightly more dissonant than consonant.

One interesting discovery was the presence of undertones in the bells. The two middle bells hanging on each side of the structure had tones that popped out underneath the perceived fundamental tone. The mystery of the heard minor third opposed to the measured perfect fifth was solved by this discovery. The fundamental tones of the two bells as seen on the spectrum analyzer and perceived by my ear are A# and C# – voila! the minor third. The bell that sounds a C# had an undertone of F, thus the tuner picked up the undertone. Interestingly, the center two bells on each side all had undertones and the tuner picked up on these undertones as confirmed by the spectrum analyzer.

So I had fun putting all the frequencies on charts to compare and contrast them. It was interesting to note how true (or not true) the bell overtones were to charted pitches. For example, a concert A is 440 hz but the bell tone frequencies that fell in the 400 block of A were closer to 432 hz (something to chew on for all the 432 hz tuning conspiracy theorists.) I charted all the over (and under) tones to see which tones were the most prominent. The most frequently appearing tone was a B, which is the tone of the natural world and deep space. The next most frequent were A,D,F and G. The least frequent tone was G#, but that one popped up alot in the tank tops.

In the final analysis, I used a pentatonic scale of BADFG for the soundscape called “Adrift in a Sea of Bells”- and it is a work in progress. Tonight we will play to a prerecorded track of the piece, but you will get a taste of it. My computer decided it was tired and started dropping audio as I was playing the soundcapes at our runthrough last Sunday. So we will not play “The Sound of Sirens” tonight because I really need to trigger and sculpt that piece with Ableton. We will perform that next time we play.

I believe the weather will turn in our favor, so please join us tonight at 7 pm for a Post-Moog sound offering. We will be at 724 Foster Street at the Sound Sculpture in front of the Central Park Elementary School. Bring your own chair and join us for an hour or so. Look forward to seeing you there!

Experiments in Audio Origami: Observing Sonic Folds

 

Folding/Unfolding at The Carrack Gallery in March was the first exploration in creating sound as origami in acoustic space. The soundscape accompanied Glenna Batson’s workshop Human Origami, which is conceived as a long form movement exercise in folding and unfolding the body in partnership with fabrics and textures. My approach to the soundscape was to create folds in the sounds through rising/falling tones, through voicings with longer decay, through amplitude ebbs/swells, and through acoustical comb filtering. Here are some samples of these effects from the recording of that day (I particularly enjoyed playing to the train “whistle” that came through at one point and playing with the creaks and groans as the dancers moved across the old wooden floors):

(The guitars in the last excerpt are from a recording of Lisa Means and Martha Dyer playing in the Sun(Ra) Room March 2016)

I analyzed the waveform of the soundscape using the Sonic Visualizer and a spectrum analyzer. These programs give me access to the amplitude measurements and the frequency measurements captured by the Zoom H2n microphone. One effect that appeared was an indication of folds in the amplitude created by swells and voicings with longer decay. Here is a picture of one section that highlights this effect:


I was very happy to see this folded impression in the amplitude waveform. This image confirms that these two techniques do create a kind of audio fold, so I will continue to explore with these techniques.

The frequency waves are the next layer of folding and can be observed through spectrum analyzers. This is more complex domain as frequency over time consists of fundamental tones with accompanying harmonic and enharmonic overtones. Looking at the soundscape as it unfolded in time and space, I was able to note a 40 hz fold as well as jumps in the fold at 14 Khz. When the scape was more percussive, the entire  spectrum behaved like a whip, with the jump at around 40 to 60 hz creating a wave effect that seemed to stop in the mid-range and then undullating up from 11 Khz to 14 Khz. Here is a video of that effect in motion (very noticeable on the last 15 secs of the video):

When the scape had more low pad drones with tonal voices such as woodwind or guitar providing the rhythmic momentum, the waveform was in a steady state across the spectrum with the tonal phrases creating quick blips in the wave. Since the next exploration of Human Origami will add ruptures and rips to the folding/unfolding process, this steady state frequency base with tonal ruptures in the waveform will be a technique to explore.

The comb filtering aspect of Audio Origami at The Carrack was not successfully rendered or captured. This was due in part to the unexpected large stack of boxes in the middle of the gallery on the day we presented the workshop. The boxes covered approximately 36 sq feet and went almost to the ceiling. They were part of the exhibit that was in The Carrack at the time. The boxes created an interesting shape for both the movers and the audio. I proceeded with the plan to point one speaker directly at the window to create strong early reflections, and those reflections were absorbed, refracted and diffracted around the space by the boxes. On one side of the room the boxes were between the larger speaker and the reflecting speaker, causing the early reflections to sort of travel around a corner. The boxes helped to make the experience of the soundscape quite different depending on where the listener was in the room. Given the way that sound moves through and around objects, I am certain there were myriad folds in the layers of tones as they reached the dancers.

The participant feedback on the soundscape was encouraging : “exquisite”, “primitive”, “a confection of music”, and (paraphrasing) supportive of the movement more often than leading the movement.  Several of the descriptors were very much a part of the intention of the soundscape, so I have a good grounding from which to move forward into the next experiment in Audio Origami. This workshop will take place May 15th at The Joy of Movement studios in Pittsboro, NC  from 2 – 4pm. The focus of this workshop is moving with paper, which holds its form in a way that fabric does not, and is also more prone to tearing and rupturing. The contrast will be very interesting for the movers and will be compelling to explore in creating the soundscape.

I hope you will join us in the journey.