The typical sound of George Lucas’ laser blaster in the “Star Wars” series is well known. What does a laser blaster in “Star Wars” sound like, and why? Here we show a simple way to produce this sound by using low-cost lab material, like a spring or a Slinky (Fig. 1). Building on the work of Crawford, who analyzed the sound of a Slinky using oscilloscopes, we present a technique for analyzing the sound using mobile devices. For a deeper quantitative analysis, a PC with open-source software is used.

The typical sound of George Lucas’ laser blaster in the “Star Wars” series is well known. What does a laser blaster in “Star Wars” sound like, and why? Here we show a simple way to produce this sound by using low-cost lab material, like a spring or a Slinky (Fig. 1). Building on the work of Crawford,1 who analyzed the sound of a Slinky using oscilloscopes, we present a technique for analyzing the sound using mobile devices. For a deeper quantitative analysis, a PC with open-source software is used.

Fig. 1.

Slinky (above) and spring (below) connected to a can as resonator.

Fig. 1.

Slinky (above) and spring (below) connected to a can as resonator.

Close modal

Recently, many papers showed how to use smartphone applications effectively, targeting hands-on experiments for educational purposes, especially in mechanics2–4 and acoustics.5–7 These papers deliver an experimental approach to several acoustical concepts and phenomena. The underlying effect of the “Star Wars” sound, namely acoustic dispersion, was until now not under investigation. The experimental work we present here closes this gap by offering different methods and complexities for students to study the physics of this phenomenon:

  1. Elementary exploration of the phenomenon without any technical material.

  2. Qualitative and basic quantitative analysis using smartphone apps—“making acoustic dispersion visible.”

  3. Detailed analysis using a PC with Audacity and two microphones.

Figure 2 shows a schematic overview of the different methods and difficulty levels examining the acoustic dispersion with springs. For the experiments presented here, different types of springs can be used. The best acoustic as well as data collection results were produced with long metal pull springs with a circular cross section of the material,8 because they offer a longer propagation path for the wave (diameter of the material: 1.7 mm, diameter of spring: 34.8 mm, number of windings: 520, total length of material: 56.8 m).9 
Fig. 2.

Experimental setup (a) for exploration without any technical material, (b) for qualitative and basic quantitative analysis using smartphone apps, and (c) for detailed analysis using a laptop with Audacity and two microphones.

Fig. 2.

Experimental setup (a) for exploration without any technical material, (b) for qualitative and basic quantitative analysis using smartphone apps, and (c) for detailed analysis using a laptop with Audacity and two microphones.

Close modal

To analyze the sound of the spring or Slinky with a smartphone qualitatively, a tin can or plastic cup can be used as a resonator. The internal microphone of an ordinary smartphone is sufficient to record the sound, if it is located close to the resonator. Depending on the operating system, different apps are available.10 

For a detailed quantitative analysis, we used a laptop and the free and open-source audio software Audacity.11 To record the sound, two contact microphones are connected to the spring at the different endings. The connection to the laptop depends on the model. We recommend using an external sound card with line-in-input (details in the following section).

The spring (or Slinky) needs to be connected to a resonator (e.g., a tin or a plastic cup). An assistant holds this output end, while the experimenter holds the input end [see Fig. 2(a)]. The experimenter stretches the spring parallel to the floor, so it doesn’t touch the ground and adjacent turns of the spring don’t touch one another. If an assistant supports the spring in the middle, rubber bands can be used as a connection to the spring, so dampening of the vibrations are minimized at this point. With a pen or a fingernail, the spring is struck on the input end. This short impulse, produced by hitting the spring, consists of a continuous spectrum of frequencies. After the vibrations propagate down the spring, the resonator emits a significantly longer signal that changes from high to low pitch very quickly: “tiuu”— the sound of a laser blaster. When shortening the length of the spring by holding it at half-length, then at quarter-length and so forth, striking again, the sound changes with every shortening: from a stretched “tiuu” to a short sound, which resembles the initial “tack.”

Please note: Springs or Slinkies are usually used to demonstrate the propagation of waves, as a model for the propagation of sound, by compressing the coils of the spring and letting the compression travel along the spring. Though the setting seems quite similar, this experiment does not use the longitudinal waves of the spring, but the longitudinal compressional waves of the material when hitting the spring.

This exploration demonstrates that the phenomenon depends on the length of the spring and therefore the distance the sound travels through the material. The longer the material, the longer the sound lasts. From a learner perspective, two initial assumptions might be relevant:

  1. The initial sound, produced by hitting the spring, is made of various frequencies.12 

  2. High-frequency components travel faster in the spring than lower frequencies.

As Crawford stated, “[H]igh frequency components of the ‘delta function’ excited by the pencil tap propagate down the Slinky® delay line more rapidly than do the low frequency components.”1 These hypotheses can be analyzed by using mobile devices.

Crawford further found in his experiments with a Slinky, that “the spring tension has very little to do with the whistler we hear” (Crawford calls the “tiuu” sound “whistler”), which we could confirm in our experiments with the spring.

Crawford used in his experiments an oscilloscope and microphones to analyze the sound. For a more contemporary and hands-on implementation in a classroom situation, we’ll show an updated version of these experiments using mobile devices, using their internal microphone and appropriate applications for displaying and analyzing audio signals [see Fig. 2(b)].

To analyze the waveform of the sound we used the app phyphox,13,14 with its tool “audio oscilloscope” as shown in Fig. 3(a). When zooming in by changing the time scale [see Fig. 3(b)], it becomes obvious that the first wave packets have the shortest time period (high pitch) and the following wave packets’ time periods get consecutively longer—and therefore have a lower pitch.

Fig. 3.

(a) Waveform of the “tiuu” sound in the oscilloscope view, recorded with the app phyphox (250 ms record time). (b) Zoomed segments of the record (10 ms).

Fig. 3.

(a) Waveform of the “tiuu” sound in the oscilloscope view, recorded with the app phyphox (250 ms record time). (b) Zoomed segments of the record (10 ms).

Close modal

This (audible) shift of pitch can be visualized in an audio spectrum, which calculates and displays the frequencies of the sound by fast Fourier transform (FFT). However, the standard FFT spectrum doesn’t display time-dependent development of the frequency components. This important information can be visualized in an audio spectrogram (we used the app SpectrumView15). Figure 4 shows the spectrogram of the blaster sound: the declining sound is clearly visible, as well as the echoes, reflected at the ends of the spring. These echoes are not only reaching the output end later but also more stretched, as they propagate for a longer time in the material. The existence of echoes might be a surprising discovery, as they are not easily audible.

Fig. 4.

Spectrogram of the “tiuu” sound (main signal) recorded with the app SpectrumView. The starting signal of the strike against the spring and the first and second echo of the sound are visible as well. The FFT settings were chosen to get highest time resolution given by the app: FFT Order: 8 (256 samples), Audio Sample Rate: 48,000 Hz.

Fig. 4.

Spectrogram of the “tiuu” sound (main signal) recorded with the app SpectrumView. The starting signal of the strike against the spring and the first and second echo of the sound are visible as well. The FFT settings were chosen to get highest time resolution given by the app: FFT Order: 8 (256 samples), Audio Sample Rate: 48,000 Hz.

Close modal
For quantitative information about the propagation speed of the wave through the spring (speed of sound in the material) for different frequencies, it is crucial to get a starting signal of the input event on the spring. By using a tube instead of the pen to hammer on the spring, the air-conducted starting signal can also be recorded with the smartphone (see Fig. 4). Measuring the time t between this signal and the declining sound signal and knowing the distance s the sound travels along the spring, you can calculate the average speed of sound for different frequencies in the material:
(1)
This average speed v can be assumed as the group velocity vG of the specific frequency, which is a function of the frequency f:
(2)

This relation is given in dispersive media like metal.

The distance s (= 56.77 m) the signal travels on the spring was calculated by measuring the perimeter d (= 0.109 m) of one coil of the spring and multiplying it with the number of coils n (= 520).

Using the app SpectrumView, we were able to measure the time t the sound traveled in the material. As an exact display of time in the app is not possible (e.g., by zooming into the time axis), we used a screenshot of the recorded data and exported it on a computer (see Fig. 5).

Fig. 5.

Spectrogram recorded with SpectrumView. The traveling times for four frequencies are highlighted. The data were analyzed on a computer. Uncertainties in time measurements were high.

Fig. 5.

Spectrogram recorded with SpectrumView. The traveling times for four frequencies are highlighted. The data were analyzed on a computer. Uncertainties in time measurements were high.

Close modal

The time measurements for the different frequencies are of limited accuracy—in addition to the uncertaintiesof the errors in reading the data, the time of the initial input event is not very precise. In this setting, the detected input signal (“tack”) travels through the air from the input end to the microphone—this time is not included. This error is in the order of magnitude of 5 to 10% of the measured times (Fig. 10). This problem can be solved by including the echoes to the measurement or by using a more sophisticated setup with two microphones (see next section).

Figure 6 shows the results of the calculation of the group velocities vG for different frequencies. These results support assumption 2.

Fig. 6.

Group velocity vG for different frequencies, calculated on the basis of the spectrogram from SpectrumView (see Fig. 5). Errors come from uncertainties in measurement of length of spring and reading error of time from the spectrogram.

Fig. 6.

Group velocity vG for different frequencies, calculated on the basis of the spectrogram from SpectrumView (see Fig. 5). Errors come from uncertainties in measurement of length of spring and reading error of time from the spectrogram.

Close modal

The setup with a smartphone does not allow the gathering of direct information of the input event. The time between input event (striking the spring) and output event (hearing the sound) can only be measured indirectly. To verify if the method used above is correct, a setup using a laptop and microphones is used. The main differences of this setup are that signals are recorded on two different channels from two microphones respectively [see Fig. 8(a)]16 and the software Audacity is used for more accurate measurements and analysis.

Fig. 7.

Setup of the experiment with two microphones and external sound card.

Fig. 7.

Setup of the experiment with two microphones and external sound card.

Close modal
Fig. 8.

Typical signal detected with laptop and Audacity showing the (a) waveform and (b) spectrogram. Mic 1 is near the input end and mic 2 is near the output end. In (b) the amplitudes of the frequency content are coded by color (low=blue, high=red/white).

Fig. 8.

Typical signal detected with laptop and Audacity showing the (a) waveform and (b) spectrogram. Mic 1 is near the input end and mic 2 is near the output end. In (b) the amplitudes of the frequency content are coded by color (low=blue, high=red/white).

Close modal

To get data from the input end as well as from the output end in the same data set, two microphones were connected to the laptop [see Fig. 2(c)]. As the normal microphone jack is a mono-input, a line-in jack is needed. Some devices offer this option, but most don’t. In this case an external sound card with line-in jack is necessary.17 The two microphones are connected with a Y-cable (splitting right and left channel) to the line-in jack of the sound card, so in Audacity each microphone is on a separate audio track (see Fig. 7).

Using the external sound card, the line-in is selected as input. When starting the recording and hammering on the input end of the spring, microphone 1 detects a signal and after a short while (depending on the length of the spring) another signal on microphone 2 can be detected, which looks and sounds very different from the first signal [see Fig. 8(a)]. A conversion of the recorded data into a spectrogram reveals that this input signal consists of a continious frequency spectrum. This corresponds to assumption 1 listed under the elementary explorations. As assumption 2 stated, the frequency is declining over time—in a nonlinear fashion [Fig. 8(b)].

Spectrogram representation makes the reflected signals visible too: mic 1 shows a signal in the second part, which was reflected on the output end and detected at the input end (reflection 1). Note that the detected reflection on microphone 2 corresponds to echo 1 in Fig. 4.

The results of the spectrogram had been verified by analyzing the waveform of different wave packets: Measurements of the period of the waves (by counting the number of samples between two maxima) confirmed the frequencies in the spectrogram. As discussed above (“Elementary exploration”), varying the length s of the spring can modify the generated sound.

In order to demonstrate the antecedent, Fig. 9 shows the signal of mic 2 for three different spring lengths. In the spectrogram four different frequencies from 500 Hz to 7000 Hz are highlighted and the dashed line represents the input signal, detected on mic 1.

Fig. 9.

Spectrogram for three different lengths of the spring detected with mic 2 using Audacity. The run times of four frequencies (7 kHz, 3 kHz, 1.5 kHz, and 0.5 Hz) are highlighted. The dashed line marks the input, detected with microphone 1. Sample Rate: 44,100 Hz, FFT Order: 10 (1024 samples).

Fig. 9.

Spectrogram for three different lengths of the spring detected with mic 2 using Audacity. The run times of four frequencies (7 kHz, 3 kHz, 1.5 kHz, and 0.5 Hz) are highlighted. The dashed line marks the input, detected with microphone 1. Sample Rate: 44,100 Hz, FFT Order: 10 (1024 samples).

Close modal
Fig. 10.

Comparison of calculated group velocities for different lengths of the spring, based on Fig. 9. The shorter the spring, the more inaccurate the measurement. The mobile measurement (orange) and the computer measurement (blue) for the same length show higher accuracy for the computer measurement, confirming the mobile approach.

Fig. 10.

Comparison of calculated group velocities for different lengths of the spring, based on Fig. 9. The shorter the spring, the more inaccurate the measurement. The mobile measurement (orange) and the computer measurement (blue) for the same length show higher accuracy for the computer measurement, confirming the mobile approach.

Close modal

The results for the group velocities at different lengths were consistent and are shown in Fig. 10. The results for shorter springs (PC 14.2 m and PC 25.4 m) show much lower accuracy as the errors in reading the data carry more weight. Higher frequencies show a higher margin of error due to the shorter “time of flight” of the wave packet.

A comparison between PC measurement (PC 56.8 m) and mobile measurement (Mobile 56.8 m) confirms the systematic error of the mobile measurement described above: The calculated speeds of sound are significantly higher (>10%) when using mobile devices. As discussed previously, this is caused by the air-conducted and therefore delayed input signal.

The experiments shown above can help students to understand how the sound is produced and how it is related to the group velocity of sound in the spring. Measuring the group velocity for different frequencies gives the necessary data to get a dispersion relation for sound in the spring empirically. A theoretical model of this phenomenon was suggested by Crawford.1 On the assumption that the spring can be regarded as a long metal bar with a cross section of the diameter of the spring, he deduced a dispersion correlation (ωk) for the transversal waves for a “Slinky whistler,” which is given by:
(5)
Herein ω=2πf,k=2πλ, the material-dependent speed of longitudinal compressional waves c (for steel: c=5800 m/s), and r=d12, where d is the thickness of the material. Crawford is using Slinkies with a rectangular cross section (see Fig. 11). In our case the thickness is the diameter of the cross section of the spring material. From Eq. (5) one derives the frequency-dependent group velocity vG as:
(6)

For the spring we were using (d=0.17 cm), the theoretical model of Crawford does not seem to fit very accurately (see Fig. 12). As Crawford presumed, the effect of the spiraling material in the spring might cause this divergence.

A further empirical investigation of the Crawford model could be part of subsequent research, where Slinkies with different winding diameters are examined. A hypothesis might be that the larger this diameter, the less the spiraling of material, and the better the Crawford model with vGf0.5 fits. As the scope of this article lies in educational aspects of the experiments, further theoretical analysis beyond this point had been omitted.

Fig. 11.

Cross section of (a) a Slinky and (b) the spring, showing the thickness d of the material. Indicated are the direction of oscillation (blue) and the direction of propagation (green) of the transversal wave. Note: A Slinky with a rectangular cross section features two different modes for oscillation.

Fig. 11.

Cross section of (a) a Slinky and (b) the spring, showing the thickness d of the material. Indicated are the direction of oscillation (blue) and the direction of propagation (green) of the transversal wave. Note: A Slinky with a rectangular cross section features two different modes for oscillation.

Close modal
Fig. 12.

Comparison between the experimental results (for n=520 coils) and the theoretical model for the dispersion correlation. Shown are the measured group velocities for some frequencies (blue) and the theoretical group velocities based on Eq. (5). Both axes are in a logarithmic scale.

Fig. 12.

Comparison between the experimental results (for n=520 coils) and the theoretical model for the dispersion correlation. Shown are the measured group velocities for some frequencies (blue) and the theoretical group velocities based on Eq. (5). Both axes are in a logarithmic scale.

Close modal

The experiments shown above offer great potential for hands-on experiments about dispersion at high school as well as at the university level. By visualizing the audible phenomenon with mobile devices, it helps to understand the concept of dispersion and allows the cross connection to other dispersion phenomena like in optics (rainbow). The use of very low-level material in combination with mobile apps offers the potential for individual exploration. By developing and using a more sophisticated experimental setup with external microphones, we demonstrated the reliability of the mobile device setup. The measurement uncertainties are reasonably low for educational purposes. The systematic error in the mobile measurement can potentially be avoided by using the echoes of the signal to measure the time of flight. This will, on the other hand, make the measurement more complicated for students to understand. Therefore, we suggest discussing potential errors and more sophisticated analysis methods after conducting the experiments.

Further experimental studies should be done, for example, on the impact of the tension of the spring. Our studies showed evidence for Crawford’s assumption, that tension has no influence on the dispersion. A deeper experimental review on the theoretical model, which Crawford suggested for the “Slinky whistlers,”1 should be part of further investigation.

The reason that metal springs show dispersion—while those made of plastic do not—lies in the properties of the material: “There is no dispersion in homogeneous amorphous solids such as glass and plastic. … In heterogeneous materials such as cast iron, granite, concrete, etc., dispersion is very large. Considerable dispersion is also observed in most metals, even at a high degree of homogeneity.”18 

An alternative way to go further in the analysis of acoustic dispersion in everyday life is possible with ice: If you throw a stone on a frozen lake, a similar sound like the one of a laser blaster is audible. Furthermore we plan to make invisible sound phenomena visible in real time by using augmented reality technologies, which have already been proven suitable for physics education in other contexts.19–21 

In order to investigate the “tiuu” sound, we presented an updated version of the “Slinky whistler”1 experiment by using mobile devices to analyze the sound produced by a metal spring. This experiment offers great potential to investigate the phenomenon of dispersion with students at the high school and college level. The easily available material, as well as the additional sparking context of “Star Wars,” fosters an enthusiasm for the subject matter in class.

We discussed in this paper an experiment in which the well-known sound of the “Star Wars” laser blaster is produced. However, this analysis does not answer why laser blasters make a “tiuu” sound. The answer is pretty simple: Filmmaker George Lucas hired the sound designer Ben Burtt to produce the sounds for the “Star Wars” movies. Burtt is the inventor of the famous laser blaster sound and he came up with it during a backpacking trip by hitting the guy-wire of an AM radio transmitter tower with a hammer.22 A guy-wire is, like the metal pull spring, a long steel cable and shows the same phenomenon of acoustic dispersion.

1.
F.
Crawford
, “
Slinky whistlers
,”
Am. J. Phys.
55
,
130
(
Feb.
1987
).
2.
K.
Hochberg
,
S.
Becker
,
M.
Louis
,
P.
Klein
, and
J.
Kuhn
, “
Using smartphones as experimental tools – A follow-up: Cognitive effects by video analysis and reduction of cognitive load by multiple representations
,”
J. Sci. Educ. Technol.
29
(
2
),
303
317
(
2020
).
3.
K.
Hochberg
,
J.
Kuhn
, and
A.
Müller
, “
Using smartphones as experimental tools – Effects on interest, curiosity and learning in physics education
,”
J. Sci. Educ. Technol.
27
(
5
),
385
403
(
2018
).
4.
S.
Becker
,
A.
Gößling
,
P.
Klein
, and
J.
Kuhn
, “
Using mobile devices to enhance inquiry-based learning processes
,”
Learn. Instr.
69
,
101350
(
2020
).
5.
M.
Hirth
,
J.
Kuhn
, and
A.
Müller
, “
Measurement of sound velocity made easy using harmonic resonant frequencies with everyday mobile technology
,”
Phys. Teach.
53
,
120
121
(
Feb.
2015
)
6.
C.
Florea
, “
Brief analysis of sounds using a smartphone
,”
Phys. Teach.
57
,
214
(April
2019
).
7.
J.
Kuhn
and
P.
Vogt
, “
Analysing acoustic phenomena with a smartphone microphone
,”
Phys. Teach.
51
,
118
119
(
Feb.
2013
).
8.
The pull springs were ordered from a supplier of springs, http://www.federnzentrale.de.
9.
A cheaper version of this experiment can be done by using a metal Slinky (diameter of Slinky: 65.3 mm, diameter of material: 2.2 mm*0.75 mm, number of windings: 81, total length of material: 16 m). The toy called “Slinky®” is a spiral that is often used to demonstrate the propagation of acoustic waves. You need to make sure, that you work with a metal Slinky, as plastic springs don’t work for the acoustic phenomenon discussed in this paper.
10.
For iOS we used and recommend the app phyphox as a multi-tool, and SpectrumView as a specialized app for audio spectrograms. Both apps are freeware. SpectrumView offers a paid pro version with more functionality.
12.
An expert knows that the initial sound (“tack”) —produced by hitting the metal— is an impulse made out of various frequencies. From a learner perspective, this might not be obvious and therefore this hypothesis is a good start for an investigation.
14.
S.
Staacks
,
S.
Hütz
,
H.
Heinke
, and
C.
Stampfer
, “
Simple time-of-flight measurement of the speed of sound using smartphones
,”
Phys. Teach.
57
,
112
(
Feb.
2019
).
16.
With Audacity you can use the internal microphone of a laptop or plug an external microphone to record the sound. Many laptops don’t offer a separate microphone input jack, as it is a combined headset jack. In this case a Y-cable is needed, which allows you to plug a microphone in separately. Best results can be achieved by using contact microphones, which are attached on the spring.
17.
We used a CSL - USB 7.1 external sound card (for about 25€).
18.
Z.
Nazarchuk
,
Acoustic Emission
(
2017
), p.
52
19.
K.
Altmeyer
,
S.
Kapp
,
M.
Thees
,
S.
Malone
,
J.
Kuhn
, and
R.
Brünken
, “
Augmented reality to foster conceptual knowledge acquisition in STEM laboratory courses – Theoretical derivations and empirical findings
,”
Brit. J. Educ. Technol.
51
(
3
),
611
628
(
2020
).
20.
M.
Thees
,
S.
Kapp
,
M. P.
Strzys
,
F.
Beil
,
P.
Lukowicz
, and
J.
Kuhn
, “
Effects of augmented reality on learning and cognitive load in university physics laboratory courses
,”
Comp. Hum. Behav.
108
,
106316
(
2020
).
21.
M. P.
Strzys
,
S.
Kapp
,
M.
Thees
,
P.
Klein
,
P.
Lukowicz
,
P.
Knierim
,
A.
Schmidt
, and
J.
Kuhn
, “
Physics holo.lab learning experience: Using smartglasses for augmented reality labwork to foster the concepts of heat conduction
,”
Eur. J Phys.
39
(
3
),
035703
(
2018
).
22.
J. W.
Rinzler
,
The Sounds of Star Wars
(
Chronicle Books
,
2010
).