Experiment III – Fishing String / X/Y H4n + 2 Piezo Microphones

Date of Experiment: September 19, 2011

Purpose + Introduction:

The purpose of this experiment is to detect the differences between the 2 different sized piezo and the X/Y H4n microphones before attempting further experiments with these microphones.

Result + Observations + Data:

I came across some difficulties when attempting my first setup.  In one instance, I thought the piezo microphone’s weight and size would be negligible to the vibration of the
string.  So I tried attaching the microphone directly to the string; the string did not resonant.  Another difficulty I found was that the piezo microphones were extremely sensitive and the X/Y H4n microphones were extremely insensitive.  The recording levels were set to 35 for the piezo microphones and 100 for the X/Y H4n microphones.  The results below will show why.

The link below is the original recording done during a regular studio session.

Notice how clear it is to hear distant footsteps and the clicking of the computer mouse.  Now compare it graphically with the spectrogram.

It is clearly shown in the spectrogram that recording at a level of 100 will pick up a lot of unwanted background noise – the mass of random blue dots.  The background noise will be more evident when it is compared to the piezo microphone recording below.

Input 1 (25/32″ diameter)

Input 2 (1 1/16″ diameter)

Notice how Input 1 (first sound) produces a higher frequency than Input 2 (second recording).  Now to compare it graphically with the spectrogram.

After analyzing the 2 different spectrograms, I noticed some obvious similarities and some shocking differences.

Similarities I found were:

  1. Resonating tones at 1k, 2k and 4k Hz.

Differences I found were:

  1. Input 1 there is a cut off at around 12k Hz, while Input 2 there is a cut off at around 7k Hz.
  2. Input 1 has another frequency at 8k Hz, but is just an overtone from the matching 1k, 2k, 4k Hz
  3. Input 2 has another set of frequencies at 2.5k Hz and 5k Hz.

Ironically, even with the extra set of frequencies, the sound output from Input 2 was more accurate to the X/Y H4n recording.

Airborne and vibration sounds are very similar in sound colour.  However, there are pros and cons to both.

X/Y H4n microphone Piezo Microphone
  • Records want humans hear
  • Records resonance from
    the aural atmosphere
  • Records just the sound
    from the string
  • More accurate to the
    resonance of just the instrument
  • Very quiet compared to piezo microphone / hard to hear
  • Picks up external noise
  • Extremely sensitive –
    picks up sound from any equipment, setup and procedure errors
  • Need to find consistency
    for recording (placement of microphone and size of microphone)


In conclusion, attaching the piezo microphones to the bulldog clips rather than attaching directly to the strings was a more effective recording method.  The recording from the larger piezo microphone was also more accurate than the smaller peizo microphone.  Next, I will experiment with the piano to identify the differences in sound from the strings and the soundboard.  This should provide me with a better understanding to passive amplification and the change in sound

Experiencing Architecture

When reading ‘Experiencing Architecture, Chapter X – Hearing Architecture’, one of the first things I thought to myself was “YES!  I am so glad I am not the only one who appreciates the acoustics of architecture in the earlier ages”.  I was particularly intrigued with the barrel-vaulted passage in Copenhagen.  I can just imagine how resonant a space like that would be.  I just want to have someone stand in the centre of the arch on one end and I would stand in the centre of the arch on the other end and listen to the many different ways on how I could experiment with a space like that.

This also brings me back to my post ‘first thoughts’ under inventions and areas of research.  Even though I am interested in the passive amplification with materials and instruments, such as the soundboard of a piano and the tuba, I am also as interested with the forms required to amplify sound.  The two shapes I continue to think about are the circle and the parabola.

Some experiments I would like to do in the near future with shapes would be to take a circular, spherical or a parabolic shape and hear the different ways how sound can be passively amplified through these.  The most readily available objects I can think off the top right now are spotlight bowls and satellite dishes.

Colour of Sound – Timbre – Quality of Sound

After reviewing and analyzing the different spectrogram graphs through the experiments, I still did not understand why the graph did not look like a single horizontal line as it was playing one frequency.  I understood that overtones were graphed, but I did not understand how they were produced.  This reminded me of Edison’s phonograph, and the tin-foil film and how it would be indented with a needle on a diaphragm. But I asked myself how could it tell the difference between the sound of a voice and of a cornet if both were singing and playing the same frequency.  I now found the answer.

Though the spectrogram graphs I used in my experiments generated through audacity are different from the graph from Musical Instruments, they can be read the same way.  In Musical Instruments, the higher the vertical wave, the higher the amplitude.  In Audacity, ranging from blue-red-white, the whiter the colour, the higher the amplitude.  In the first graph shown below there are 2 amplitude peaks at different points of frequency, or strong frequency regions.  These are called formants.

Even though 2 different instruments can be playing the same frequency, there can be a distinction between the 2.  Refer to the diagrams below and how the components along the frequency axis are evenly spaced apart.  This spacing is the fundamental frequency, hence the same note or pitch.  However, from different people and instruments, there will a different formants and different wave-forms.  These differences determine the difference in colour of sound, timbre and quality of sound.  At the moment, I cannot find any difference between these 3 terms.

References and Reading material

  1. Fauvel, John, Flood, Raymond, and Robin Wilson. Music and Mathematics,  New York, Oxford University Press, 2004. p 55-56
  2. Campbell, Murray, Greated, Clive, and Myers, Arnold. Musical Instruments History, Technology & Performance of Instruments of Western Music, New York, Oxford University Press, 2007. p 35-36

Schematic Device 1

As I came during lunch to get a spotlight bowl from Karen Schellenberg, the Theatre Technician, I waited outside and had nothing better to do than to sketch.  Relating to my ‘Experiencing Architecture’ post, I was particularly intrigued with the parabola and its focus point, hence the need to get a parabolic form, or spot light.  My first thought was similar to Edison’s phonograph where there would be a diaphragm with a needle (or piezo) in the middle.  This needle would be located at the focus point so that when any type of wave form come towards the parabola, it would engage the diaphragm and the needle into vibration.  The clear disadvantage is that the waves must come perfectly perpendicular to the parabola or the effect is lost.

Next, I thought about having a network of strings moving through the diaphragm then continuing to reach out of the parabola.  In this case, the vibrations on any of the strings would send a wave towards the focus then resonant that wave back out in the direction the parabola is facing, and potentially vice versa.

As for the diaphragm, discussing about colour of sound – timbre – quality of sound, it is in my thoughts to have a variety of diaphragm materials producing different colours of sound, timbre and qualities of sound.  Such materials should be flexible and thin enough to vibrate freely.  At the same time, the material needs to have enough tensile strength and needs to be aurally permeable.  Potential materials are:

  1. Brass
  2. Pine
  3. Plastic
  4. Paper
  5. Rubber

In fact, I just thought about having a whole network of parabolic forms attached to one another, and to a ‘master parabola’.  Some ways I can could take advantage of the ‘only works if perfectly perpendicular’ case are:

  1. Have some parabolic objects in plan and others only in section
  2. Have parabolic objects move around to different ‘focus’ areas

Experiment IV – Piano

Date of Experiment: September 19, 2011

Purpose + Introduction:

To graphically identify the difference in colour of sound  and amplitude from the vibration of the strings and soundboard to what we hear.

Results + Observations + Data:

The recording below is of an upper C played on the piano split up into 3 different channels:

  • Input 1 (piezo microphone on strings)

  • Input 2 (piezo microphone on soundboard)

  • X/Y H4n recorder microphone (airborne sound).

Shown below are graphs generated from Audacity which show different types of information.

One aspect that is not surprising on all the spectrograms above is the fundamental frequencies are the same, since they all are recording the same note.  Most of the differences that do occur between the 2 piezo recordings.  Note how Input 2 has a stronger ‘white amplitude’ over Input 1.  This is most likely from the better resonance effect of the soundboard over the string.  Input 1 also shows a strange horizontal line and wavers on and off at around 1.5 kHz.  I do not know the reason for that, but I can guess hypothesize that another object, like the pin holding the string, is vibrating at another frequency and is causing deconstructive and constructive waves.

I was very surprised when I saw the difference graphically in the amplitude of the waveform in Input 2.  This is most likely the case of ‘hitting the sweet spot’ – the exact location of where the bridg elies on the soundboard.

Again, I’m not very surprised that the X/Y H4n waveforms look almost exactly the same, but after listening to the audio file I was expecting to see similar waveforms.  I was wrong.  Again, I do not know the answer to this, since all of the audio clips sound like a string vibrating

However, some of these differences in amplitude from one piezo recording to the other is that one is connected to a few strings, while the other is connected to the soundboard, which is in contact with all strings.

Also, taping down the piezo microphone is not the best attachment method for recording.  This may cause additional vibration from the microphone itself, and may not have full contact with the vibrating object.

Experiment V – Tube length and resonance

Date of Experiment: Sept 22, 2011

In this experiment, I played starting from the bass clef-bottom staff Bb down to an E.  I was looking at the effects of resonance in relation to the length of the tube.  Each length of tube corrisponds to a different note (or frequency) because each note has a different wavelength.  When the length of tube corresponds to this wavelength, the sound is most resonant.  However, in this experiment, it shown that a half step (ie. Bb to A) is still playable with the same length of tube.

In the following audio file, each interval will play the following:
Note corresponding to the tube length, pitch-bend note half-step lower with same tube length, extend tube length to corresponding note

Sympathetic Resonance

After Friday’s crit, 2 types of devices were suggested. One was an annunciator / recorder that could also be multiplied.  The other was a device that would cause repetitive vibrations along a surface, like a motor with a rope.  I’m still not too certain with my exact device, but I know I am interested and was directed towards a more creative side about the immaterial phenomenon of vibrations.

At the end of the day, I still didn’t have a scientist to look at, but was just focused towards vibrations, but I felt the urge to find someone, so I started with Helmholtz.  What made me first fascinated in his work was by the way he was able to graphically represent vibrations through the vibration microscope.

Then I was able to find a few of his experiments based on sympathetic resonance.  The first was based on the resonance piano strings.  If one was to press the key of a note, but not play it, that string is free from any dampers.  After, if one would play a note that is of the corresponding pitch (or pitches) loud enough, that free string will be able to resonate to the note that was just played.

Another experiment I came across is the tuning fork resonance box.  This is very much the same concept as the piano strings.  However, this time the tuning fork resonance box can point to sound to a specific direction so even across a room, with another box of the same frequency, one box is able to put the other in resonance purely by the vibration of sound.

But I wanted to see and hear this to another level of intensity…I did not only wanted to see motion through matching frequencies, but more of an explosive action: the breaking of wine glass through the resonance of sound.  And through MythBusters, it is confirmed to be true.

The fact that resonance can be produced and reproduced in so many ways and the power it has to put objects into motion is truly amazing. How I apply this to a device, I don’t know yet, but this is one method that draws me back to the first device of the annunciator / recorder.


Helmholtz, Hermann von, and Ellis, Alexander John. On the Sensations of Tone as a Physiological Basis for the Theory of Music, London: Longmans, Green, and Co., 1885

Making Sound

Though the next few experiments are not based off of sympathetic resonance, there is still a strong connection between the two phenomenons.

The ‘Singing Rod’ is an experiment that uses friction to create sound.  This is based on the concept of nodes, and how a simple amount of frictional force can cause an aluminum rod to resonate so freely.

The ‘Sound hose – Whirly Hose’ is an experiment that uses corrugated pipes to create vortices with the suction of air molecules that finally creates a whistling sound.

The ‘Singing Pipes’ is an experiment that uses the movement of hot air inside, even though the air movement is very slow, to vibrate the metal and create sound.

These three experiments have exposed to creating natural sounds with very little force.  This is starting to convince me that creating my device does not need the use of electronics.

Lastly, ’Sympathetic Resonance’, by Joshua Kirsch, is a really cool piece of sculptural-music art.  It does not use sympathetic resonance to perform its functions, but uses electronics instead.  However, what amazes me is the gentle touch on a metal key can create a natural looking stroke of the mallet.

After watching this, I wonder if this piece could still be functional without the use of electronics and work purely by the matching or corresponding frequencies of an intricate network of finely tuned strings.

Making the Device

The last few days have heavily into designing and building the devices.  I already knew what I wanted to make, but I just had to plan it out so I would work as planned.  At first, the trickest part was the electronics component since I haven’t soldered a circuit since grade 8.  Another were the attachments and connections so that the vibrations would feed through the entire box and figuring out the dimensions so it would match the frequency.  I also thought about other additions to a simple circuit, but then I remembered about keeping the circuit simple.  Especially since this is my first time.

I first started off with making the resonance boxes.  Not too much of a problem building, and it seems to work for the first box, but the real test is with the combination of 2 to test the sympathetic resonance.  Though I am a bit worried it sounded a bit quieter than I thought.

When the shop closed, I had to break through my barrier of lack of soldering.  In the end, I was able to successfully solder together my first circuit board.  Though I am still waiting for my vibration motors, most likely to come on Monday, it did work with other motors.  Also, with the help of a tuner, I am able to tune the motor to an ‘A’ using the potentiometer.  This should keep the tuning fork going for a sustained amount of time rather than just hitting it once.  If the vibration of the motor does not keep it going, I also had an idea of putting on a rubbery type arm to ‘hit’ the tuning fork.

I also started drawing together a electric box on CAD to be set for laser cutting. This should be ready by the beginning of next week.

Though the devices are not completely yet, I have some plans on how to use the tuning fork resonant boxing on site.

  1. Measuring the interuption of activites between resonant boxes.
  2. Measuring the sympathetic resonance of the surface.
  3. Finding a ‘whispering’ area of where the sympathetic resonance is most effective by the amplitude of the recording.

Experiment VI – Resonance Box

I was so glad to finish 4 resonance boxes within 12 hours.  The following video shows the testing of 2 resonant boxes with 2 piezo microphones on the left resonance box for audo recording.

Unfortunately, I came across many troubles when trying to make this work.  Well, technically, it’s not working the way I want it to.  The only reason I can think it is not working is that the initiated resonance box is not loud enough, which is most likely the case for the failed attempts of sympathetic resonance.  However, after the 2nd hit with the mallet, there seems to be a very quiet sound that is vibrating the left box.  Here are some of my thoughts on how to improve the device:

  1. Use longer and thicker tuning forks
  2. Create a slot for the tuning fork through the top of the box rather than just the attachment
  3. Create a hammer for a mtor

This shows a video how industry standard resonance boxes work: