Audio Forensic Examination - Montana State University
" 2009 IEEE. Personal use of this material is permitted.However, permission to reprint/republish this material foradvertising or promotional purposes or for creating newcollective works for resale or redistribution to servers or lists,or to reuse any copyrighted component of this work in otherworks must be obtained from the IEEE."[Robert C. Maher]Audio ForensicExamination[Authenticity, enhancement, and interpretation] BRAND X PICTURESThe field of audio forensics involves many topicsfamiliar to the general audio digital signal processing (DSP) community, such as speech recognition,talker identification, and signal quality enhancement. There is potentially much to be gained byapplying modern DSP theory to problems of interest to theforensics community, and this article is written to give the DSPaudience some insight into the types of problems and challengesthat face practitioners in audio forensic laboratories. However,this article must also present several of the frustrations and pitfalls encountered by signal processing experts when dealingwith typical forensic material due to the standards and practicesof the legal system.Digital Object Identifier 10.1109/MSP.2008.931080INTRODUCTIONAudio forensics refers to the acquisition, analysis, and evaluation of audio recordings that may ultimately be presented asadmissible evidence in a court of law or some other officialvenue. Audio forensic evidence is typically obtained as part of acivil or criminal law enforcement investigation or as part of theofficial inquiry into an accident or other civil incident.The principal concerns of audio forensics are i) establishingthe authenticity of audio evidence [1], [2], ii) performingenhancement of audio recordings to improve speech intelligibility and the audibility of low-level sounds [3], [4], and iii) interpreting and documenting sonic evidence, such as identifyingtalkers, transcribing dialog, and reconstructing crime or accident scenes and timelines [5]. The requirements of admissibilityinto a court of law mean that the techniques employed mustIEEE SIGNAL PROCESSING MAGAZINE [84] MARCH 20091053-5888/09/ 25.00 2009IEEEAuthorized licensed use limited to: MONTANA STATE UNIV BOZEMAN. Downloaded on April 2, 2009 at 14:16 from IEEE Xplore. Restrictions apply.
speech transcription andpreviously have been proven toFORENSIC AUDIO RECORDINGSspeaker identification wouldthe court to be unbiased, toTYPICALLY SUFFER FROM NOISE,play the original tape (or a refhave known reliability statistics,DISTORTION, INTERFERING SOUNDS,erence copy) repeatedly whileto be nondestructive, and to beAND OTHER SIGNAL PROCESSINGadjusting the filter and gainwidely accepted by experts inCHALLENGES THAT CAN IMPEDEsettings to get the best subjecthe field.tive result.For example, a modern digiPROPER ANALYSIS.tal speech enhancement techAUDIO FORENSICS AND THE LAWnique may seem like an obvious choice for cleaning up aIn the United States, the controlling legal case establishingrecorded surveillance recording prior to preparing a transcript,the admissibility of audio forensic evidence in the form ofbut the court may need to be convinced that the “enhancerecorded conversations is generally regarded to be the 1958ment” could not have resulted in a change to the meaning orruling in United States versus McKeever (169 F.Supp. 426,interpretation of the recorded dialog. If noise is removed by the430, S.D.N.Y. 1958). The judge in the McKeever case had tospeech enhancement procedure, the defense and prosecutingrule on the admissibility of a tape-recorded conversationattorneys and the court may be concerned that an underlyinginvolving the defendant. Although in the McKeever case aphoneme was also inadvertently (or deliberately!) altered. Forwritten transcript was ultimately presented to the jury ratherinstance, the court could reasonably worry that a signal prothan playback of the tape itself, the judge’s ruling establishedcessing technique might change the interpretation of a noisya set of requirements that are now used, with some variation,utterance “I didn’t do it!” to the phrase “I did too do it!”, within most state and federal courts in the United States whenthe obviously different implications for the transcript.considering the authenticity of audio recordings. The sevenSimilarly, establishing the proper chain of custody and authentenets of audio authenticity from the McKeever case are listedticity when dealing with digital audio files typically goes wellin Table 1.beyond the standard operating procedure of academic DSPTenets 1 and 2 are now considered less significant than in theresearch labs. Thus, due to these special legal considerations,late 1950s when tape recorders were not the ubiquitous andthis article must include some of the arcane history and pracfamiliar devices judges and juries know of today. Tenets 3 and 4tices of audio forensics even though the topics and techniquestend to imply the involvement of audio experts who can examineseem out-of-date to those of us in the signal processingthe physical tape and the subtle magnetic characteristics of theresearch field.recording device to determine if any splices, stop/start sequences, overrecordings, or other issues exist that would indicate inadHISTORYvertent damage or deliberate tampering. Tenet 5 demands aOver the preceding 40 years, forensic audio examination hasproper chain of custody of the recording, while tenet 6 requiresgradually become a recognized profession. Prior to the availabilthat the participants in the recording be identified either byity of DSP tools, most forensic audio analysts worked exclusivelyvoice or by corroborating witnesses to the recording. Finally,with analog magnetic tape and basic recording studio electrontenet 7 requires that the recorded conversation be spontaneousics such as analog filters, variable-speed playback equipment,and not coerced.gain compressors, and test equipment such as oscilloscopes,In a legal sense, both the authenticity of the physical evimicroscopes, and voice spectrographs [3].dence, consisting of the tape and recording system, and the legalSince the 1960s, the U.S. Federal Bureau of Investigationimplications of the transcript—which is often subject to inter(FBI) has conducted examination of audio recordings for speechpretation and dispute—are issues for the court to sort out. Forintelligibility enhancement and authentication [3]. Law enforceexample, the fact that a recorded conversation is generallyment organizations around the world also have developed proobtained out of court, the participants are not sworn, and witcedures and standards for handling audio forensic material innesses may or may not be available for cross-examination, makesthe context of local legal rules and customs.it necessary for the court to determine whether or not theThe authenticity of analog tape recordings was (and still is)recording is admissible as evidence in a trial.assessed using magnetic development, which uses a colloidalsuspension of fine magnetic particles in a fluid thatevaporates after being spread on the tape [1]. When[TABLE 1] MCKEEVER CASE REQUIREMENTS FOR AUDIO AUTHENTICITY.dry, the magnetic particles stick to the magneticdomains recorded on the tape, thereby revealing vis- 1) THAT THE RECORDING DEVICE WAS CAPABLE OF TAKING THE CONVERSATION NOWOFFERED IN EVIDENCE.ibly under a microscope the underlying magnetic 2) THAT THE OPERATOR OF THE DEVICE WAS COMPETENT TO OPERATE THE DEVICE.patterns. Determining authenticity from modern 3) THAT THE RECORDING IS AUTHENTIC AND CORRECT.4) THAT CHANGES, ADDITIONS, OR DELETIONS HAVE NOT BEEN MADE IN THE RECORDING.digital files can be problematic, as described later in 5) THAT THE RECORDING HAS BEEN PRESERVED IN A MANNER THAT IS SHOWN TO THE COURT.this article.6) THAT THE SPEAKERS ARE IDENTIFIED.THAT THE CONVERSATION ELICITED WAS MADE VOLUNTARILY AND IN GOOD FAITH,In predigital days, the forensic audio examiner 7)WITHOUT ANY KIND OF INDUCEMENT.tasked with the problem of signal enhancement forIEEE SIGNAL PROCESSING MAGAZINE [85] MARCH 2009Authorized licensed use limited to: MONTANA STATE UNIV BOZEMAN. Downloaded on April 2, 2009 at 14:16 from IEEE Xplore. Restrictions apply.
EXPERT WITNESSESIn the United States, the various state and federal jurisdictionsapply a variety of standards for admitting the testimony of topical experts. The standards are commonly based on the 1923Frye case (Frye versus United States, 54 App. D.C. 46,293F.1013, DC Ct App 1923), the Daubert case (Daubert versusMerrell Dow Pharmaceuticals, 509 U.S. 579 1993) or somesimilar interpretation. The Frye standard requires that themethods and techniques of the expert be generally accepted bythe scientific community. Daubert uses the Federal Rules ofEvidence to support an acceptability test of relevance and scientific reliability of the expert‘s testimony. Subsequent cases,such as Kumho Tire Co. versus Carmichael (526 U.S. 1371999), have extended the Daubert standards beyond scientifictestimony to technical and other specialized knowledge (suchas audio engineering).original recording. The panel‘s conclusion was based primarilyon the characteristic start/stop magnetic signatures present onthe subject tape.The form of the panel’s examination quickly became thestandard approach for assessing the authenticity of forensicaudio recordings:1) Physically observe the entire length of the tape.2) Document the total length and mechanical integrity of thetape, reels, and housing.3) Verify that the recording is continuous with no unexplained stop/start sequences or erasures.4) Perform critical listening of the entire tape.5) Use nondestructive signal processing as needed for intelligibility enhancement.Other well-known forensic audio cases include crime reconstruction attempts using acoustic evidence of a Dallas PoliceDepartment Dictaphone recording purportedly from the assassination of President Kennedy [7], interpretation of backgroundsounds from cockpit voice recorder (black box) data [8], and theuse of voice identification techniques for authenticating recordings of Osama bin Laden and other terrorists [9].THE 18½-MINUTE GAPThe watershed event for audio forensics was arguably the 1974investigation of a White House conversation between U.S.President Richard M. Nixon and Chief of Staff H.R. Haldemanrecorded in the Executive OfficeAUTHENTICITYBuilding in 1972. InvestigatorsTHERE IS POTENTIALLY MUCH TOLike other types of forensic evidiscovered that the audioBE GAINED BY APPLYING MODERNdence, forensic audio may berecording contained an unexsubject to accidental or deliberplained section lasting 18½DSP THEORY TO PROBLEMS OFate tampering. The court mustminutes during which buzzINTEREST TO THE FORENSICS COMMUNITY,be convinced of the authenticisounds could be heard but noBUT EXPERTS MUST UNDERSTANDty and integrity of the audiodiscernable speech sounds wereTHE STANDARDS AND PRACTICESevidence. What is authenticity?present. Due to the highly speOF THE LEGAL SYSTEM.How can it be demonstrated?cialized nature of the technicalHow might DSP help?evidence, Chief Judge John J.Ideally, an audio recording made for forensic purposes willSirica of the U.S. District Court for the District of Columbiabe produced with authenticity verification in mind. For examappointed a special Advisory Panel on White House Tapes to giveple, the recording will include an audio slate, consisting of aexpert advice to the court. The advisory panel consisted of sixspoken announcement of the relevant information governingtechnical experts, jointly nominated by the counsel for the presthe recording, such as location, date, time, participants, model,ident and the special prosecutor, with the court‘s direction “. . .and serial number of the recorder and the recording medium,to study relevant aspects of the tape and the sounds recorded onetc. The recording should be made in one continuous sessionit” [6]. The panel members included many familiar names fromwithout pauses or stop/start sequences. Authenticity is alsothe engineering and acoustics communities: Richard H. Boltmore verifiable if one deliberately allows uniquely identifiable(Bolt, Beranek, and Newman), Franklin S. Cooper (Haskinsbackground sounds such as clock tower chimes or radio broadLaboratories), James L. Flanagan (AT&T Bell Laboratories),casts to be included in the audio recording.John G. McKnight (Magnetic Reference Laboratory), Thomas G.Stockham Jr. (University of Utah), and Mark R. Weiss (FederalFORENSIC AUDIO RECORDINGSScientific Corp.).The authenticity of audio forensic evidence has traditionallyThe advisory panel performed a series of objective analysesfocused on analog magnetic tape recordings. The forensic examof the tape itself, the magnetic signals on it, the electrical andiner performs a series of observations and tests to evaluate theacoustical signals generated by playback of the tape, and theintegrity of the recording [2], [5], [10]. Despite the fact that digproperties of the recording equipment used to produce theital recorders with solid-state flash memory are increasinglymagnetic signals on the tape. Analysis included observation ofavailable, many law enforcement agencies in the United Statesthe audio signals as well as magnetic development of thestill rely almost exclusively on analog cassette and microcassettedomain patterns and head signatures on the tape. Ultimatelyrecorders due to the agencies’ inventory and familiarity withthe panel determined that the 18½-minute gap was due to sevthese devices—as well as authenticity concerns with digital data,eral overlapping erasures performed with a specific model ofas noted later in this section. The forensic audio authenticitytape recorder that differed from the device that produced theIEEE SIGNAL PROCESSING MAGAZINE [86] MARCH 2009Authorized licensed use limited to: MONTANA STATE UNIV BOZEMAN. Downloaded on April 2, 2009 at 14:16 from IEEE Xplore. Restrictions apply.
procedure for media such as analog tapetypically follows the strategy employed bythe Advisory Panel on White House Tapes.The methodology requires the examinerto observe the physical integrity of therecording medium, the quality of therecorded audio, and the consistency ofthe magnetic signatures present on thetape. The details of the process aredescribed next.PHYSICAL HANDLINGAND INSPECTIONThe examiner documents the conditionand properties of the evidentiary recording, including the length and condition [FIG1] Magnetic development image of a record-over signature from a segment ofof the tape, the condition of reels and analog magnetic tape. The two striped bars on the left side of the figure are thehousing, any manufacturing serial num- magnetic traces of a two-channel recording. The vertical line near the middle and thesmudged image to the right are due to the magnetic erase head being energized andbers or batch numbers, and the magnetic destroying the underlying magnetic patterns previously recorded. A continuous,configuration on the tape (number of unaltered segment of magnetic tape would not exhibit this erasure signature, so in thistracks, mono or stereo, etc.). The tape case the forensic examiner would suspect the tape had been deliberately altered afterthe original recording was made (from [11]).itself is inspected to look for any splicesor other changes, and the recorder usedis difficult to exclude the possibility that the audio content itselfto produce the tape is also inspected and tested.was edited or manipulated prior to a surreptitious reencodingstep. In general, ancillary information and meticulous chain-ofCRITICAL LISTENINGcustody practices are essential.The examiner carefully listens to the entire recording and notesOne recent effort in audio authentication that is applicableany apparent alterations or irregularities. Any audible evidenceto digital recordings is analysis of residual signals due to couof edits, splices, or audible discontinuities in backgroundpling of the electrical power line frequency into the audiosounds, buzzes, tones, and so on are noted.recording system [12–16]. The electric network frequency(ENF), nominally 60 Hz in the United States and 50 Hz inMAGNETIC SIGNATURE AND WAVEFORM OBSERVATIONSmany other parts of the world, is not precisely constant butThe condition of the magnetic signals on the tape is examinedvaries up to 1/20.5 Hz from time to time in an unpredictableusing magnetic development techniques and compared to referfashion due to small mismatches between the electrical systemence signatures of recordings obtained from the same recordingload and system generation. The alternating magnetic fielddevice. An example magnetic development image is shown inemanating from AC power lines can cause audible hum in theFigure 1. The magnetic signatures associated with the recordrecording, which is usually considered undesirable. However,and erase heads, as well as the transitions from stop to record,if hum is present, the characteristic frequency fluctuations canrecord to pause, overrecording, and so forth, are examined forconceivably be traced to a particular date and time if sufficientconsistency. The examiner also observes and measures the elecdata is available, since the ENF is consistent over the entiretrical waveforms obtained during playback of the tape.geographic region served by a synchronous AC network “grid.”Thus, the ENF information extracted from a forensic recordREPORT PREPARATIONing could be compared to a database of known ENF data forFinally, the examiner analyzes the observations and writes athe electrical power network to verify the date and time of thereport explaining whether the tape is believed to be authentic,recording [15].a copy, or altered in any manner after the original recordingOne system proposed for extracting the ENF from audiowas made.recordings is shown in Figure 2. A sample comparison of ENFdata obtained from an audio recording and the reference ENFDIGITAL MEDIAdata from the electrical power system is shown in Figure 3.The question of authenticity becomes more complicated withThe ENF procedure may not always be applicable in practicedigital recordings because the evidence of tampering or alterabecause magnetic field coupling from the power system may betions is more difficult to discover than mechanical splices orminimal in well-designed audio equipment, when condenser oroverdubbing signatures in a physical master analog tape. A digipiezoelectric microphones are used, or when battery-operatedtal recording can be encoded with a checksum, processed withequipment is used in an area away from the power grid [16].an embedded digital watermark, or otherwise encrypted, but itIEEE SIGNAL PROCESSING MAGAZINE [87] MARCH 2009Authorized licensed use limited to: MONTANA STATE UNIV BOZEMAN. Downloaded on April 2, 2009 at 14:16 from IEEE Xplore. Restrictions apply.
Input DataDecimationto300 HzBandpass Filter49.5–50.5 HzSplit Data into JOverlapping FramesWeightingWindowFrame 1WeightingWindowFrame 2WeightingWindowFrame JSegments3 to J –1Zero PadData by aFactor 4Zero PadData by aFactor 4Zero PadData by aFactor 4FFT Frame 1FFT Frame 2FFT Frame JQuadraticInterpolationof Peak ValueFrame 1QuadraticInterpolationof Peak ValueFrame 2QuadraticInterpolationof Peak ValueFrame JVector of ENF Estimates: f1, f2,.fJ[FIG2] Proposed ENF processing procedure (from [15]).50.2Archive (London)Frequency50.1550.1ENHANCEMENTForensic audio recordings typically suffer from noise, distortion,interfering sounds, and other signal processing challenges thatcan impede proper analysis.Perhaps the most common enhancement issue for forensicaudio examiners is the clandestine surveillance recording of arelevant conversation via a hidden microphone. The surreptitious nature of the recording system often leads to poormicrophone placement with respect to the participants, interference from wind and other environmental sounds, and rubbing or muffling due to clothing that comes in contact withthe microphone
with typical forensic material due to the standards and practices of the legal system. INTRODUCTION Audio forensics refers to the acquisition, analysis, and evalua-tion of audio recordings that may ultimately be presented as admissible evidence in a court of law or some other official venue. Audio forensic evidence is typically obtained as part .
Forensic Science is the integration of core scientific disciplines. Forensic science involves a variety of careers. 1. Students will recognize the major contributors to the development of Forensic . Worksheets Lab; Activity Project assessments Research activities such as “famous forensic scientists and their contributions” or “careers inFile Size: 444KBPage Count: 21People also search forforensic science for high school textbook pdfdo forensic criminologist investigate the cri forensic criminology bookswhat is a dental hygienisthow to check fingerprint forensic criminologyare dental hygienists and dentist same thing
765 S MEDIA TECHNOLOGY Designation Properties Page Audio Audio cables with braided shielding 766 Audio Audio cables, multicore with braided shielding 767 Audio Audio cables with foil shielding, single pair 768 Audio Audio cables, multipaired with foil shielding 769 Audio Audio cables, multipaired, spirally screened pairs and overall braided shielding 770 Audio Digital audio cables AES/EBU .
Sample Forensic Examination Worksheet Examination of Firearms Lab File No.:_ Examiner: _ . 08ws_firearm.pdf Page 1 of 6: Sample Forensic Examination Worksheet Examination of Firearms . National Forensic Science Technology Center (NFSTC)
Forensic science is the application of science to law. Any science can be applied into a legal situation, but some of the commonest forensic sciences include forensic biology, forensic chemistry, and forensic toxicology. The word forensic in today’s world simply
forensic science discipline (or equivalent). Experience It is essential that the post holder is an experienced forensic scientist in forensic drug analysis, forensic toxicology and preferably in forensic criminalistics, with a minimum of 10 years performing multi-disciplined forensic
Forensic Psychology Chapter ObjeCtives ·orensic Define f psychology. · Review career areas in the forensic sciences. · Distinguish forensic psychology from forensic psychiatry. · Identify and describe the major subareas of forensic psychology. · Review the educational, training, and certification requirements to become a forensic psychologist.
Delivering forensic services (Report 21: 2018-19) 4 . Summary of audit findings . Delivering forensic services . We audited four types of forensic services: fingerprints, deoxyribonucleic acid(DNA), forensic medical examinations and illicit drugs. Three of these services accounted for approximately 92 per cent of all forensic services .
accordance with asset management guidelines and procedures established by the Director of Finance & Administration or a designee. 8 Asset Inventory Departments will conduct a full inventory of all property under their stewardship, in accordance with the inventory schedule developed by F&A, and will provide the results of that inventory to the Director of Finance & Administration or a designee .
