Lending an Ear in the

Courtroom: Forensic Acoustics

Robert C. Maher Forensic acoustics deals with acquisition, analysis, and evaluation of
audio recordings to be used as evidence in an official legal inquiry.
Postal:
Department of Electrical and Introduction
Computer Engineering In most circumstances, the
Montana State University memorable sounds of the world
610 Cobleigh Hall around us include the routine
Bozeman, Montana and ephemeral sounds of civili-
59717-3780 zation in our urban or rural sur-
USA roundings; the sociable speech
Email: of friends and coworkers; the
rob.maher@montana.edu pleasurable notes of music; and
the occasional barks, chirps,
croaks, purrs, and thunderous
Figure 1. A typical courtroom, as pictured here, may
rumbles of the biophony and
differ in many respects from venues such as the class-
rooms and professional conference halls more familiar geophony. However, there are
to ASA members, but acoustical forensic science can play some circumstances in which
a key role in both civil and criminal legal proceedings. the sounds around us become
the subject of a law enforce-
ment investigation, an accident review, or some other legal proceeding that ends
up in a courtroom. Although most acousticians might reasonably prefer to stay out
of a courtroom (Figure 1), except perhaps to improve the architectural acoustics of
the facility, there are surprisingly many circumstances in which the knowledge of
acoustical scientists can be helpful to legal and investigative proceedings.
Does your coworker's voice mail message confess to a crime?
Was the emergency alarm actually loud enough to be heard when the
engine was running?
Was that subtle background sound due to a telltale malfunction emerging
in your passenger planes airframe?
Does that music playing in the bar infringe on a valid copyright?
Was the threatening message on your voice mail the utterance of
Mr. Smith or was it Mr.Jones?

Welcome to the Field of Forensic Acoustics!

Forensic science has become a common subject in the media due to dramatic court
cases in the news and the emergence of fictional entertainment series on US televi-
sion like CSI: Crime Scene Investigation and NCIS on the CBS network and quasi
documentaries such as Forensic Files on the TLC and Headline News networks.
Even the rather arcane specialty of audio forensics sometimes finds its way into
the headlines, particularly in sensational cases. Many people listened to online
copies of the 911 call center audio evidence related to the tragic shooting of un-
armed black teenager Trayvon Martin in 2012, and recent reports have described
the recovery and interpretation of cockpit voice recorder black box audio from

22 | Acoustics Today | Summer 2015 , volume 11, issue 3 2015 Acoustical Society of America. All rights reserved.
Germanwings Flight 9525 that crashed in
the French Alps in March 2015. Forensic
acoustics comes into play also when there
is a dispute about the likelihood that an
acoustical product presents a hearing
hazard or interferes with the audibility of
an aural emergency notification alarm.
The domain of forensic acoustics even
includes forensic musicology for cases of
alleged copyright infringement of musi-
cal style and intellectual property (Be-
gault et al., 2014). The impact of forensic Figure 2. Example of a forensic audio recording, with manual annotation, displayed with
musicology was evident in the recent suc- a sound software package. It depicts the amplitude vs. time waveform of approximately one
minute of audio recorded at an emergency dispatch center and includes utterances from the
cessful litigation award of US$7.4 million land mobile radio system and the local voice of the dispatcher. Figures like this allow experts
in damages a court ruled must be paid in forensic acoustics inform and educate attorneys and "triers of fact" (e.g., judge, jury,
by entertainers Robin Thicke and Phar- designated administrator) who are likely inexperienced with forensic audio interpretation.
rell Williams to the heirs of the late singer
Marvin Gaye.
Despite its occasional high profile, forensic science has come interpretation of tangible audio recordings, forensic acous-
up against some serious scrutiny and soul searching in re- tics may also treat questions of audibility in the context of
cent years, including an influential 2009 report from the litigation or criminal prosecution, such as civil annoyance
US National Research Council that criticized many forensic complaints from an outdoor performance venue, noise lev-
fields, including audio forensics, for lacking scientific evalu- els produced by takeoffs and landings at an airport that vio-
ation of reliability and error rates (National Academy of Sci- late local statutes, or whether or not a scream or other sound
ences, 2009). Thus, it is increasingly important that the sci- was likely to have been detectable under the circumstances
ence in forensic investigations be based on unquestionably claimed by an ear witness.
objective interpretation and not merely subjective opinions, Forensic acoustics experts who deal with recorded evidence
as has sometimes been the case. are most often consulted about three concerns: authenticity,
I begin with a description of several key elements, terms, and enhancement, and interpretation (Maher, 2010).
historical origins of forensic acoustics, primarily in terms of Authenticity denotes the acceptance of a recording as being
the judicial system in the United States. Next, I consider the unaltered and true to its source and chain of custody. Crim-
challenge of determining authenticity in the age of digital inal and civil cases may hinge on a dispute about the cir-
data storage and the corresponding opportunities present- cumstances under which a recording was made and whether
ed by the increasingly wide deployment of audio recording audible material in the recording could have been deleted,
apparatus in memo recorders, digital cameras, and virtu- added, or otherwise edited after the fact (Koenig 1990; Au-
ally every contemporary smartphone. Finally, I sum up the dio Engineering Society, 2000).
field by mentioning areas in which interested members of
the Acoustical Society of America (ASA) can get involved to Enhancement involves signal-processing techniques that at-
provide new and innovative research that will help advance tempt to improve the intelligibility of speech, the clarity of
forensic acoustics. specific background sounds, or the overall signal-to-noise ra-
tio of the recording. Modern enhancement techniques use a
high-quality digital copy of the original recording so the origi-
What Is Forensic Acoustics?
nal recording medium is used only to obtain the work copy for
Forensic acoustics, or audio forensics, is the specialty field
enhancement (Musialik and Hatje, 2005; Koenig et al., 2007).
of acoustics and audio engineering that deals with the ac-
quisition, analysis, and evaluation of audio recordings that Finally, interpretation refers to the description of the acous-
are to be presented as evidence in an official inquiry or in a tical evidence in words, pictures, statistics, and graphs that
court of law (Maher, 2009). In addition to the analysis and help address investigative questions, explain the sequence
Summer 2015 | Acoustics Today | 23
Forensic Acoustics

of acoustical events, and educate the attorneys, defendants, 1958): Current advances in the technology of electronics
judges, and juries about the meaning, significance, and limi- and sound recordings make inevitable their increased use
tations of the recorded evidence (Audio Engineering Soci- to obtain and preserve evidence possessing genuine proba-
ety, 1996; Maher, 2009). Figure 2 depicts an interpretive an- tive value. Courts should deal with this class of evidence in
notation that has been added to a time waveform to assist a manner that will make available to litigants the benefits
with an investigation. of this scientific development. Safeguards against fraud or
other abuse are provided by judicial insistence that a proper
Forensic Acoustics and the US Courts foundation for such proof be laid.
Audio forensics traces its origins to the development of por-
[]A review of the authorities leads to the conclusion that,
table recording equipment, and examples of the use of audio
before a sound recording is admitted into evidence, a foun-
recordings in US courts date to the mid-1950s. Although the
dation must be established by showing the following facts:
courts generally began to accept the unique importance of
(1) that the recording device was capable of taking the con-
audio recordings, especially in cases involving speech ob-
versation now offered in evidence; (2) that the operator of
tained via clandestine surveillance or wiretaps, there were
the device was competent to operate the device; (3) that the
significant considerations regarding the Fourth Amend-
recording is authentic and correct; (4) that changes, addi-
ments protections against unreasonable searches and sei-
tions or deletions have not been made in the recording; (5)
zures and concern about the legal admissibility of a record-
that the recording has been preserved in a manner that is
ing as being a bona fide representation of the sonic events
shown to the court; (6) that the speakers are identified; and
actually present during the recording process.
(7) that the conversation elicited was made voluntarily and
The McKeever Case in good faith, without any kind of inducement.
Among the key cases in the US federal court system regard-
ing the admissibility of audio forensic evidence is United Since the early 1960s, US Federal Bureau of Investigation
States v. McKeever (1958). In the McKeever case, two de- (FBI) laboratories have developed techniques and proce-
fendants were indicted for extortion in an antiracketeering dures for assessing the authenticity and audible contents of
prosecution involving the International Longshoremens forensic audio recordings obtained from law enforcement
Association. After his indictment, the defendant McKeev- investigations, and similar capability has been instituted
er had arranged to make a surreptitious tape recording of in other public and private forensic acoustics labs around
a conversation he had with an individual who later was a the world (Koenig, 1990). As described in Modern Audio
witness in the trial. During the trial, McKeevers defense Forensics: Digital Good and Digital Bad, tenet 3, au-
team sought to challenge under cross-examination the wit- thenticity determination, has become more difficult in our
ness testimony by playing a portion of the clandestine tape era of read/write digital recording and computer processing
recording to refresh the witness recollection. The court al- compared with the historic reliance on physical analog mag-
lowed the tape to be played but only via headphones so that netic tapes.
the witness could hear it but not the jury. When the defense
Forensic Acoustics and Watergate
then sought to have the recording played in court so that
A significant turning point in the practice of forensic acous-
the jury also could hear it, the prosecution objected to the
tics in the United States occurred 15 years after McKeever
use of the tape because its admissibility as evidence had not
during the Watergate scandal. In 1971, late in his first term
been established. Specifically, the court had to address the
in office, President Richard Nixon directed the Secret Ser-
fact that the recording was obtained secretly out of court, the
vice to install audiotaping systems in the Oval Office and
participants were not sworn, the witnesses disputed whether
the Cabinet Room of the White House, in the presidents
or not they recognized even their own voices, and the legal
private office in the Executive Office Building (EOB) next
chain of custody had not been demonstrated to ensure ad-
to the White House, and at Camp David, the presidents re-
missibility of the audio evidence.
treat in rural Maryland. The existence of these recording
The court examined these questions and considered a num- systems was known only to a select group of individuals
ber of prior federal court rulings, then established what the and to the Secret Service (Nixon Presidential Library and
forensic acoustics community now refers to as the Seven Museum, 2015). President Nixon presumably assumed that
Tenets of Audio Authenticity (United States v. McKeever, the recording system's existence would be of interest to no
24 | Acoustics Today | Summer 2015
one other than presidential biographers and historians af- to make the original recording (Advisory Panel on White
ter he left office. However, it quickly turned into a political House Tapes, 1974).
and legal bombshell of historic proportions in 1973 when
The procedures employed by the Advisory Panel became the
White House aide Alexander Butterfield revealed during his
standard for audio forensic investigators: (1) examine the
congressional testimony that there were secret audiotape
physical tape, reels, and structural housing, documenting
recordings of conversations between the President and his
their characteristics, total length, and mechanical integrity;
advisors! After all, this was in the midst of the Watergate
(2) verify that the recording is complete and continuous and
investigation and the widespread public concern about the
does not exhibit any erasures, splices, or stop/start sequenc-
veracity of various White House officials who had testified
es; (3) listen carefully and critically to the entire tape; and
before Congress and in federal court. President Nixon even-
(4) use nondestructive signal processing as needed for intel-
tually agreed to release edited transcripts of the various con-
ligibility enhancement.
versations recorded by the secret taping system and later the
tapes themselves.
Modern Audio Forensics: Digital
In 1974, interest turned to a particular recording of a con- Good and Digital Bad
versation between President Nixon and his Chief of Staff H. For a few decades after Watergate, the discipline of acousti-
R. Haldeman recorded in 1972 in the EOB. The investigators cal forensics revolved around two common requests: estab-
were suspicious that the recorded conversation included re- lish the authenticity of a tape recording and identify the talk-
marks about the Watergate cover-up, but when the record- ers whose utterances are audible in the recording. The work
ing was examined, the investigators discovered that 18 generally focused on analog audio because up until recently,
minutes of the recording were obliterated by an unexplained the most frequently encountered recording medium in fo-
gap consisting of audible buzz sounds but no discernable rensic cases was analog tape.
speech. Investigators suspected that someone had deliber-
Analog magnetic tape had a variety of drawbacks including
ately erased or recorded over that section of the tape to de-
comparatively poor signal quality, stability, and storage ca-
stroy the originally recorded conversation, perhaps with the
pacity. However, as found with the 18-minute gap in the
intention of eliminating incriminating remarks.
Watergate tape, the physical medium itself could provide
John J. Sirica, Chief Judge of the US District Court for the useful forensic information about edits, splices, stop/start/
District of Columbia, determined that the potentially altered erase sequences, and other alterations affecting authentic-
tape required expert analysis beyond the routine capability ity. Our contemporary digital recorders can provide much
of the court (McKnight and Weiss, 1976). He requested that greater quality, stability, and storage capacity, but the ability
the Watergate Special Prosecutor and the counsel for the to read, alter, and resave the contents of a digital bitstream
president jointly nominate a group of six outside technical and even to manipulate the files date and other file system
experts (including several ASA members) to form a special data without leaving a physical trace in the flash memory
Advisory Panel on White House Tapes to study relevant or computer disk file has caused serious concern about our
aspects of the tape and the sounds recorded on it (Advisory ability to authenticate digital audio forensic recordings. If a
Panel on White House Tapes, 1974). forensic examiner cannot detect any evidence of tampering,
it is still conceivable that the digital data could have been
The Advisory Panel analyzed the physical tape itself and the
manipulated in some undetectable manner.
electrical signals observed on playback and, ultimately of
greatest importance, performed magnetic development us- One proposed approach to address digital audio authenticity
ing ferrofluid to reveal latent magnetic domain patterns on is based on detection of a tell-tale hum in the recording due
the tape and the magnetization signatures of the recording to interference from alternating current (AC) power leak-
and erase heads installed in the tape recorders known to be ing into the audio recording (Grigoras, 2005; Cooper, 2008).
present in the White House. The magnetic development of The electrical network frequency (ENF) of the AC power
the tape led the Advisory Panel to the conclusion that the grid actually varies slightly from its nominal 60-Hz (US) or
18-minute gap consisted of several overlapping start-stop 50-Hz (Europe) frequency by a time-varying deviation that
erasures performed with a specific tape recorder available depends on the instantaneous balance between power gen-
in the White House but not the same device that was used eration and consumption on the grid, which fluctuates from

Summer 2015 | Acoustics Today | 25

Forensic Acoustics

moment to moment. By maintaining a database of the

small, instantaneous aleatoric variations of the grid
frequency for every date and time, the precise hum
frequency present in an evidentiary recording can be
matched to the database. The degree to which the re-
corded hum matches the database pattern can either
confirm or refute the claimed date and time that the
recording was made. Nonetheless, even in a system in-
volving ENF detection and comparison, an examiner
may not be able to rule out the possibility that a clever
adversary filtered out the presence of any residual ENF.
Whats more, a skilled audio forger could conceivably
synthesize a fake ENF signal and additively mix it into
the evidentiary signal, saving the contrived composite Figure 3. Example of a time waveform (top) and spectrogram (bottom) de-
picting an urban surveillance recording containing gunshots, reverberation,
information back onto the digital storage medium as a
mechanical sounds, yelling, car horns, and other background sounds and
way to fool a subsequent forensic examiner into believ- noise. Several aspects of the recorded soundscape may be of interest in a par-
ing that the recording is authentic. ticular investigation, such as the sequence and timing of audible events and
any intelligible speech utterances.
Many digital audio recording and storage systems in-
corporate metadata in the digital file format. Metadata
may include information about the recording settings, date sounds and environmental acoustics at the recording scene
and time, manufacturer of the device, and its software ver- that can sometimes become an even more important part
sion. Although metadata can potentially be altered to con- of the forensic evidence. The reverberation and background
ceal tampering with the audio data contained in the file, an acoustic emissions can provide valuable information about
audio forensic examiner should always review the metadata the authenticity of the recording or help determine the geo-
as part of an authenticity investigation (Koenig and Lacey, metric orientation of the sound sources with respect to the
2014). microphone. Typical forensic recordings involve a compli-
cated superposition of many different sounds and the effects
Other evidence of digital audio tampering can be more sub-
of microphone and audio encoder peculiarities (Figure 3).
tle, requiring careful consideration of signal continuity and
the interpretation of background sounds. An attempt to edit In audio forensics, the term acoustical fingerprinting re-
out an embarrassing or possibly incriminating speech utter- fers to the analysis of background sounds, acoustic reflec-
ance, for example, could cause a momentary, but detectable, tions, reverberation, and peculiarities of the microphone
interruption of the distinctive background sounds present and recoding system that are detectable in the evidentiary
in the original recording (Maher, 2010). recording (Alexander et al., 2012; Moore et al., 2014). Be-
cause many common consumer recording devices include
Innovation in Acoustical Forensics automatic gain control and perceptual audio coding/data
Many key principles of audio forensics have been developed compression algorithms intended for speech signals, addi-
and evolved incrementally over the last 50 years, but cer- tional research is needed to understand the degree to which
tain aspects of the field gain increasing interest from time the acoustical surroundings and the characteristics of the
to time. Here are a few of the current hot topics in foren- recording microphone and digital audio-coding algorithms
sic acoustics that could entice ASA members to join the re- can be derived from the stored audio recording.
search effort.
Proliferation of Personal Audio/Video Recording Devices
Audio Scene Fingerprinting Technology for portable audio/video recorders has pro-
As noted previously, a forensic recording typically includes gressed to the point that audio forensic evidence from a
prominent foreground sounds that are of interest to the in- crime scene may come from one or more mobile smart-
vestigation, such as speech utterances, warning alarms, or phones or cameras carried by journalists or bystanders, se-
perhaps gunshots (National Academy of Sciences, 1982; curity surveillance systems, dashboard-mounted recorders
Maher, 2007; Beck et al., 2011). Yet it is the background in police cars, and, increasingly, pocket-sized personal video
26 | Acoustics Today | Summer 2015
recorders clipped to the vests of law enforcement officers Analysis of the four audio channels of CVRs primarily in-
to document their interaction with the public. As recorder volves transcribing the spoken words of the pilot and copi-
costs continue to decrease, it seems inevitable that more and lot, and any other utterances by members of the crew. In ad-
more investigations will include evidence obtained from dition to the flight crew's speech, the microphones can also
these portable devices and dashboard recorders (McGinty, 2015). pick up nonspeech sounds that can be very important to
accident investigators. Engine sounds, airframe vibrations,
The ubiquity of audio recording devices will require an
avionics audible warning alarms, and sounds of cockpit in-
emerging specialty in handling, searching, and analyzing
trusions or other commotion may all have significance to
the potentially huge number of recordings available from
the investigation. For example, the March 2015 demise of
a public incident that leads to a forensic investigation. The
Germanwings Flight 9525 in the French Alps has been at-
investigative challenge that occurred after the 2013 Boston
tributed to deliberate action by the copilot reportedly based
Marathon bombing involving dozens of surveillance record-
in large part on preliminary evaluation of the CVR evidence.
ings and hundreds of digital photos taken by smartphone
Whats more, even the sound of the pilots respirations can
and camera-toting bystanders near the finish line provides a
give information about the flight crews health, state of alert-
glimpse of what is likely to occur in future incidents involv-
ness, and level of anxiety or agitation (Stearman et al., 1997;
ing video, audio, and still photo evidence (NOVA, 2013).
Byrne, 2002; McDermott, personal communication).
Transportation Accident Investigations:
CVR systems are typically activated automatically whenever
The Cockpit Voice Recorder
the aircraft is powered up, whereas the FDR systems collect
Because commercial aircraft accidents are so remarkably
flight data only from the point at which the plane becomes
infrequent, aviation regulatory agencies can afford to spend
airborne. This means that the CVR may contain important
substantial resources to investigate the cause of accidents
information about flight crew checklist completion, preflight
when they do occur. Arguably the most important develop-
discussion, and similar audio information obtained before
ment in accident investigations today is the invention of the
takeoff that isn't covered by the FDR information (McDer-
flight data recorder (FDR) and cockpit voice recorder (CVR)
mott, personal communication).
equipment required on all civilian commercial passenger
flights, large private jets, and many military flights, the so- Assessing the Fidelity of Audio Forensic Findings
called black boxes (National Transportation Safety Board, 2015). Audio forensic findings that will end up in court are sub-
ject to the same validity considerations used for other types
Originally developed using special fireproof magnetic tape,
of scientific expertise before the information can be stated
contemporary FDRs and CVRs now record in digital form
in the courtroom. In 1993, the US Supreme Court affirmed
using nonvolatile solid-state memory. FDR systems on con-
in Daubert v. Merrell Dow Pharmaceuticals, Inc., that un-
temporary airliners can record hundreds of flight param-
der Rule 702 of the Federal Rules of Evidence (which cov-
eters, actuator positions, and sensor readouts every second,
ers both civil trials and criminal prosecutions in the federal
with memory capacity for up to 25 hours. Yet, even with
courts), The Rulesespecially Rule 702place appropriate
the plethora of flight data, the acoustical information from
limits on the admissibility of purportedly scientific evidence
the CVR is often indispensable for accident investigators to
by assigning to the trial judge the task of ensuring that an
piece together what happened leading up to the accident.
expert's testimony both rests on a reliable foundation and
CVRs capture four separate monophonic channels: the pi- is relevant to the task at hand (Daubert v. Merrell Dow,
lots headset microphone, the copilots headset microphone, 1993). Assessing the reliability of audio forensic opinions
a cockpit area microphone (CAM) mounted in the cockpits can be a challenge. Unlike DNA comparisons that can be
ceiling panel, and the fourth channel that often is used to expressed in a formal statistical sense, a forensic acoustics
record the intercom communications between the pilots and question such as Is the utterance present in the evidentiary
the flight attendants. Modern CVRs record up to 120 min- recording the voice of Suspect A? traditionally has not been
utes of audio in a memory buffer loop, sequentially over- amenable to having a strong statistical basis for the opinion.
writing the oldest data with new data. Thus, in the event of Due to growing judicial skepticism about the scientific ba-
an accident, the audio forensic examiner will have a record- sis for almost all forensic testimony in fields ranging from
ing containing the sounds from the 2 hours preceding the fingerprints and bite marks to handwriting and carpet fiber
crash (National Transportation Safety Board, 2007). comparisons, the 2009 report by the National Academy of
Summer 2015 | Acoustics Today | 27
Forensic Acoustics

Sciences pointed out many areas of concern (National Acad- was left by someone other than the suspect, which would
emy of Sciences, 2009). depend on the probability that someone else with the same
size and type of boots could have left the print at the crime
For example, voice comparison is a common and important
scene. If only 1 person in 100 owned that particular size and
forensic request. An audio recording contains an utterance
type of boot in the local population, the naive probability of
that the prosecution claims is the voice of the defendant
observing the bootprint if someone other than the suspect
while the defense denies that allegation and claims that the
was at the crime scene would be approximately 1/100, and
recorded speech was not the voice of the accused but was
so the LR in this example is 1/(1/100) or 100. Note that the
uttered by someone else. For many years, the common ap-
LR is not in and of itself an objective way to establish guilt.
proach was for an expert in speech analysis to employ the
The LR formulation needs to be interpreted as the statisti-
aural-spectrographic method, which entails recording the
cal increase in belief in a match to the suspect based on the
defendant reading a script version of the words spoken in
particular evidence. Thus, the LR must be multiplied by the
the evidentiary recording. The expert then compares the
odds of guilt that are established based on other evidence
sound and the visual spectrogram of the evidentiary record-
and testimony in the case (Perlin, 2010).
ing and the defendants exemplar recordings and renders
an opinion regarding the degree to which the defendants ex- A common difficulty for forensic examination is that the evi-
emplar recordings match the evidence. The problem is that dence is distorted, noisy, smeared, or otherwise incomplete.
the method of matching the recordings has traditionally Determining the degree to which a smudged fingerprint
been highly subjective and therefore subject to mistakes and matches a particular record in a fingerprint database or the
unconscious bias (Bolt et al., 1970; National Academy of Sci- degree to which a recorded utterance matches an exemplar
ences, 1979; Poza and Begault, 2005). recording ultimately requires discussion of the scientific
models and assumptions used.
Due in large part to the increasing use of statistically strong
DNA evidence in crime investigation, the acoustical foren-
Conclusions
sics field has motivation to move toward the use of likeli-
Forensic acoustics is an interesting specialty field that lies
hood ratio (LR) calculations as a way to deal systematically
between the scientific world and the legal world. Although
with the uncertainties present in acoustical comparisons
the world of science is accustomed to the process of develop-
(Morrison, 2011). The LR is a way to consider two compet-
ing new theories and practical techniques that are gradually
ing hypotheses pertaining to a particular event by assessing
tested, refined, and modified incrementally through new
the probability of an observation of the event given each of
experiments, publications, and peer review by fellow scien-
the hypotheses (Perlin, 2010; Lindley, 2014). Specifically, the
tists and engineers, the legal world revolves largely around
LR is the probability of an observation being made given
precedent and the need for judges and juries to make final
that the suspect is, in fact, the perpetrator (the prosecution
decisions in a reasonable and quick manner. Members of the
hypothesis) divided by the probability of the same observa-
ASA are encouraged to become involved in improving the
tion being made but that the perpetrator is, in fact, someone
reliability and flexibility of acoustic forensic science, thereby
other than the suspect (the defense hypothesis).
enabling law enforcement, judicial, and accident investiga-

tion professionals to work with increased scientific assur-
(1)
ance and confidence.
Consider the hypothetical situation in which a particular
bootprint is observed at a crime scene and a suspect is found Biosketch
who is wearing boots that "match" the observed print to a
reasonable degree of scientific certainty. The prosecutors hy- Rob Maher is Professor and Depart-
pothesis is that the suspect is the perpetrator, so the numera- ment Head of Electrical and Computer
tor of the LR has probability 1. In other words, if the prose- Engineering at Montana State Univer-
cution hypothesis is correct that the suspect was at the scene sity in Bozeman. He holds a BS degree
while wearing the boots, the probability is 1 that the particu- from Washington University in St.Lou-
lar bootprint would be observed. Meanwhile, the denomi- is, a MS degree from the University of
nator addresses the defense hypothesis that the bootprint Wisconsin-Madison, and a PhD from

28 | Acoustics Today | Summer 2015

the University of Illinois at Urbana-Champaign, all in Elec- Maher, R. C. (2007). Acoustical characterization of gunshots. Proceedings
trical Engineering. His professional interests are in digital of the IEEE SAFE Workshop on Signal Processing Applications for Public
Security and Forensics, Washington, DC, April 11-13, 2007, pp. 1-5.
audio signal processing, audio forensic analysis, music syn- Maher, R. C. (2009). Audio forensic examination: Authenticity, enhance-
thesis, and acoustics. He is a Fellow of the Audio Engineer- ment, and interpretation. IEEE Signal Processing Magazine 26, 84-94.
ing Society, a Senior Member of the IEEE, and for 31 years, a Maher, R. C. (2010). Overview of audio forensics. In: Sencar, H. T.,Velastin,
S.,Nikolaidis, N.,and Lian, S. (eds), Intelligent Multimedia Analysis for Se-
member of the Acoustical Society of America.
curity Applications. Springer-Verlag, Berlin, pp. 127-144.
McGinty, T. J. (2015). Why the Cuyahoga County Prosecutors Office Is
References Spending Half a Million Dollars on Dash Cams for Cleveland Police Cars.
Available at http://tinyurl.com/o737dw8. Accessed May 27, 2015.
Advisory Panel on White House Tapes. (1974). The EOB Tape of June 20, McKnight, J. G., and Weiss, M. R. (1976). Flutter analysis for identifying
1972. US District Court for the District of Columbia. tape recorders. Journal of the Audio Engineering Society 24, 728-734.
Alexander, A., Forth, O., and Tunstall, D. (2012). Music and noise finger- Moore, A. H., Brookes, M., and Naylor, P. A. (2014). Room identification
printing and reference cancellation applied to forensic audio enhance- using roomprints. Proceedings of the Audio Engineering Society 54th Inter-
ment. Proceedings of the Audio Engineering Society 46th International Con- national Conference: Audio Forensics, London, UK, June 12-14, 2014.
ference: Audio Forensics - Recording, Recovery Analysis, and Interpretation, Morrison, G. S. (2011). Measuring the validity and reliability of forensic
Denver, CO, June 14-16, 2012. likelihood-ratio systems. Science & Justice 51, 91-98.
Audio Engineering Society. (1996). AES27-1996: AES Recommended Prac- Musialik, C., and Hatje, U. (2005). Frequency-domain processors for effi-
tice for Forensic Purposes Managing Recorded Audio Materials Intended cient removal of noise and unwanted audio events. Proceedings of the Au-
for Examination. Audio Engineering Society, New York. dio Engineering Society 26th International Conference: Audio Forensics in
Audio Engineering Society. (2000). AES43-2000: AES Standard for Forensic the Digital Age, Denver, CO, July 7-9, 2005.
Purposes Criteria for the Authentication of Analog Audio Tape Record- National Academy of Sciences. (1979). On the Theory and Practice of Voice
ings. Audio Engineering Society, New York. Identification. National Academies Press, Washington, DC.
Beck, S. D., Nakasone, H., and Marr, K. W. (2011). Variations in recorded National Academy of Sciences. (1982). Report of the Committee on Ballistic
acoustic gunshot waveforms generated by small firearms. Journal of the Acoustics. National Academies Press, Washington, DC.
Acoustical Society of America 129, 1748- 1759. National Academy of Sciences. (2009). Strengthening Forensic Science in the
Begault, D. R., Heise, H. D., and Peltier, C. A. (2014). Forensic musicology: United States: A Path Forward. National Academies Press, Washington,
An overview. Proceedings of the Audio Engineering Society 54th Interna- DC.
tional Conference: Audio Forensics, London, UK, June 12-14, 2014. National Transportation Safety Board. (2007). Cockpit Voice Recorder
Bolt, R. H., Cooper, F. S., David, E. E., Denes, P. B., Pickett, J. M., and Ste- Handbook for Aviation Accident Investigations. Vehicle Recorder Division,
vens, K. N. (1970). Speaker identification by speech spectrograms: A sci- Office of Research and Engineering and Office of Aviation Safety, Wash-
entist's view of its reliability for legal purposes. Journal of the Acoustical ington, DC. Available at http://tinyurl.com/ntsb-cvr-pdf. Accessed May
Society of America 47, 597-612. 27, 2015.
Byrne, G. (2002). Flight 427: Anatomy of an Air Disaster. Springer National Transportation Safety Board. (2015). Cockpit Voice Record-
Science+Business Media, New York. ers and Flight Data Recorders. Available at http://tinyurl.com/ntsb-
Cooper, A. J. (2008). The electric network frequency (ENF) as an aid to au- cvr-fdr. Accessed May 27, 2015.
thenticating forensic digital audio recordings An automated approach. Nixon Presidential Library and Museum. (2015). History of the White
Proceedings of the Audio Engineering Society 33rd International Conference: House Tapes. Available at http://tinyurl.com/tapes-history. Accessed May
Audio Forensics Theory and Practice, Denver, CO, June 5-7, 2008. 27, 2015.
Daubert v. Merrell Dow Pharmaceuticals, Inc. (1993). Supreme Court of the NOVA. (2013). ManhuntBoston Bombers. Public Broadcasting Service,
United States, 509 U.S. 579 (1993), No. 92-102. Available at http://laws. broadcast aired May 29, 2013. Available at http://tinyurl.com/pbs-bomb-
findlaw.com/us/509/579.html. Accessed May 27, 2015. ers. Accessed May 27, 2015.
Grigoras, C. (2005). Digital audio recording analysis: The electric network Perlin, M. W. (2010). Explaining the likelihood ratio in DNA mixture inter-
frequency (ENF) criterion. International Journal of Speech, Language and pretation. Proceedings of Promegas Twenty-First International Symposium
the Law 12, 63-76. on Human Identification, San Antonio, TX, October 11-14, 2010.
Koenig, B. E. (1990). Authentication of forensic audio recordings. Journal of Poza, F., and Begault, D. R. (2005). Voice identification and elimination us-
the Audio Engineering Society 38, 3-33. ing aural-spectrographic protocols. Proceedings of the Audio Engineering
Koenig, B. E., and Lacey, D. S. (2014). Forensic authenticity analyses of the Society 26th International Conference: Audio Forensicsin the Digital Age,
metadata in re-encoded WAV files. Proceedings of the Audio Engineering Denver, CO, July 7-9, 2005.
Society 54th International Conference: Audio Forensics, London, UK, June Stearman, R. O., Schulze, G. H., and Rohre, S. M. (1997). Aircraft damage
12-14, 2014. detection from acoustic and noise impressed signals found by a cockpit
Koenig, B. E., Lacey, D. S., and Killion, S. A. (2007). Forensic enhancement voice recorder. Proceedings of the National Conference on Noise Control
of digital audio recordings. Journal of the Audio Engineering Society 55, Engineering 1, 513-518.
352-371. United States v. McKeever. (1958). United States District Court in the South-
Lindley, D. V. (2014). Understanding Uncertainty. John Wiley & Sons, ern District of New York, 169 F. Supp. 426 (S.D.N.Y. 1958). Available at
Hoboken, NJ. http://tinyurl.com/qhgtlqn. Accessed May 27, 2015.

Summer 2015 | Acoustics Today | 29