STUDY MATERIAL

[Link]. Forensic Science

Course Code- NYFSE06
Course Name- Speaker Identification

Unit-V “Casework, Legal Aspects and Forensic Reporting”

Forensic report writing is the process of documenting findings, observations, and conclusions
from a forensic investigation in a clear, objective, and professional manner. These reports are
used in legal, criminal, or civil cases and must be understandable to non-experts like judges,
lawyers, and juries.

Definition:
Forensic report writing is the formal documentation of evidence analysis, methods used, and
expert conclusions prepared by a forensic professional to support legal proceedings.

Forensic evidence evaluation refers to the process of analysing and interpreting physical or
digital evidence collected from a crime scene or related to a legal investigation, in order to
draw meaningful and scientifically supported conclusions.
Admissibility Standards-
Daubert and Fryes Criteria- Notes Attached for the same.
Work Flow for Forensic Speaker Comparison-

1. Case Intake & Definition

 Receive case materials (recordings, transcripts, case notes).

2. Evidence Handling & Authentication

 Verify integrity of audio (check for edits/tampering).

 Maintain chain of custody.
 Convert files to consistent formats (e.g., WAV, 16-bit PCM).

3. Audio Pre-processing

 Noise reduction (carefully, to avoid altering speaker traits).

 Segmentation:
o Remove silence, isolate speech segments.
 Channel normalization (telephone vs. studio differences).
4. Feature Extraction

This is the core technical step, often based on Acoustic Phonetics.

A. Acoustic Features

 Fundamental frequency (pitch)

 Formants (vocal tract resonances)
 Spectral features (e.g., MFCCs)
 Voice quality (breathiness, creakiness)

B. Linguistic / Phonetic Features

 Accent/dialect
 Pronunciation patterns
 Speech rate, rhythm, intonation

5. Reference vs. Questioned Sample Comparison

 Questioned sample (Q): unknown speaker (e.g., crime recording)

 Known sample (K): suspect’s voice

Comparison methods:

 Auditory (expert listening)

 Acoustic-phonetic analysis
 Automatic systems (e.g., likelihood-based models)

6. Statistical Evaluation

Most modern systems rely on likelihood ratios grounded in Bayesian Inference.

7. Validation & Calibration

 Use background population data (same language/accent group).

 Validate system performance:
o False acceptance/rejection rates
 Calibration ensures reliable LR outputs.

8. Interpretation of Results

9. Reporting

A forensic report includes:

 Case description
 Materials analyzed
  Methods used
 Results (LR + explanation)
 Limitations (noise, short samples, disguise)

10. Court Presentation

 Expert explains:
o Methodology
o Scientific basis
o Strength of evidence
 Must be understandable to non-experts (judge/jury)

Components of Forensic Report Writing-

1. Report title
2. Date
3. Time
4. Type of Crime Scene
5. Type of Crime
6. Case No
7. FIR No
8. Charge Sheet No
9. Name of Investigation Officer
10. Name of Forensic Expert
11. Purpose of Report
12. Brief summary of case
13. Investigation Details
14. Evidences Description
15. Documentation Details
16. Interpretation of Findings
17. FSL/RFSL results findings and interpretation
18. Summary of Findings
19. Expert Opinion
20. Conclusion
21. Authorises Signature and Government seals

Report Format of RFSL/FSL-

1. Header Information
o Forensic laboratory name and logo
o Case number or lab number
o Report number
o Date of report
2. Requestor Information
o Name of the investigating officer or agency
o Department or division
o Contact details
3. Subject/Case Information
o Case title or description
o Type of case (e.g., homicide, arson, sexual assault)
o Date of offense (if known)
4. Evidence Submitted
o List and description of evidence items (with item numbers)
o Date of evidence submission
o Name of person who submitted the evidence
o Chain of custody details
5. Purpose of Examination
o Why the evidence was submitted
o What analysis was requested
6. Methods and Techniques Used
o Scientific methods or instruments used for testing
o Standard operating procedures or protocols followed
o Any reference standards used
7. Observations / Results
o Findings from tests or analysis
o Measurements, reactions, or identification details
o Photos or diagrams (if needed)
8. Interpretation / Conclusion
o Expert interpretation of results
o Relevance to the case
o Any comparison results (e.g., match/no match with known sample)
9. Remarks (if any)
o Limitations or uncertainties
o Notes on sample condition or anomalies
10. Reporting Officer Details
Name and signature of forensic analyst or examiner
Designation and section (e.g., toxicology, ballistics)
Date of completion
Reviewed/approved by (if applicable)
11. Attachments/Appendices
Photographs of evidence
Lab data sheets
Instrument printouts or chromatograms

Constitutional Validity and Admissibility of Evidences in Court of Law-

Indian Evidences Act 1872

Section 136 of the Indian Evidence Act, 1872, deals with the judge's role in determining the
admissibility of evidence. It empowers the judge to decide whether evidence presented by a
party is relevant and should be admitted into the court record. The judge can ask the party to
explain how the fact they are trying to prove is relevant to the case, and will only admit the
evidence if it is deemed relevant. If the fact's relevancy depends on another fact being proven
first, the judge may decide the order in which those facts are presented.
1. Section 3: Defines evidence as including oral and documentary evidence.

2. Section 65: Deals with the admissibility of secondary evidence in cases where the original
document is unavailable.
3. Section 65B: Addresses the admissibility of electronic records.

4. Section 17: Defines an admission as a statement, oral, documentary, or electronic, which

suggests an inference as to a fact in issue or a relevant fact.
5. Section 157: Allows the proof of a previous statement made by a witness to corroborate
their testimony.

6. Section 27: Allows the admission of information received from an accused in police custody,
if it leads to the discovery of a fact.

Bharatiya Sakshya Adhiniyam 2023

1. Section 57: Primary Evidences- Original Form of Evidences

2. Section 58: Secondary Evidences- Photocopies/Xerox Copies of original Evidences.
[Link] 63: Admissibility of Electronic Evidences before court law. E.g. Digital
Photographs, audio, video and any related to digital media.

Information Technology (IT) Act 2000

Section 79 A: Any electronic form of evidences is admissible in court law.

Interpretation and Presentation of Reports in the Court:

In legal proceedings, reports (such as medical, forensic, financial, or technical reports) are
essential pieces of evidence. Proper interpretation and presentation of these reports ensure
that the court understands the findings clearly and fairly.

Interpretation of Reports

This involves explaining the meaning, significance, and implications of the report content.

Steps in Interpretation:

Understand the Report Thoroughly:

o Know who prepared it, why it was prepared, and the methodology used.
Summarize Key Findings:
o Highlight important data or conclusions relevant to the case.
Clarify Technical Terminology:
o Use plain language to explain scientific or technical terms.
Identify Limitations:
o Be honest about any assumptions, limitations, or uncertainties in the report.

Presentation of Reports in Court

Presenting the report involves explaining it clearly in front of the judge, jury, or tribunal—
either as a lawyer, witness, or expert.

A. Preparing for Presentation

Organize findings clearly: Introduction → Method → Results → Conclusion.
Use visual aids if permitted: graphs, tables, diagrams.
Rehearse explaining it in layman's terms.

B. Oral Testimony in Court

When presenting verbally:
Speak slowly and clearly.
Maintain professionalism and objectivity.
Refer to the report directly but paraphrase rather than read word-for-word.
Be ready to explain how and why the conclusions were reached.

C. Responding Under Cross-Examination

Stay calm and focused.
Do not guess or overstate findings.
Acknowledge any uncertainties or limitations if questioned.

Expert Witness and Expert Testimony-

Indian Evidences Act 1872

Section 45-55 of the Act defines an expert as someone with specialized knowledge, skills,
or experience
The court can consider expert opinions on topics like foreign law, science, art, handwriting,
and finger impressions
The court can also consider facts that support or contradict the expert's opinion

How are expert opinions used-

Expert opinions can help judges and juries understand technical details
They can help establish facts that are important to the case
They can help authenticate evidence
They can help clarify complex issues
They can help judges and juries make informed decisions

Bharatiya Sakshya Adhiniyam 2023

In the Bharatiya Sakshya Adhiniyam (BSA), 2023, expert opinions are relevant facts when the
court needs to form an opinion on specific subjects like foreign law, science, art, handwriting,
or finger impressions. Persons especially skilled in these fields are considered experts, and their
opinions are admissible as evidence. These opinions are not conclusive and should be
considered with other evidence.
Section 39:
This section deals with the relevance of opinions of experts in various fields, including foreign
law, science, art, and handwriting.

Sections 45-55:
These sections in the Indian Evidence Act, 1872 (IEA), which is now superseded by the BSA,
also dealt with expert evidence and expert opinion. Similar principles apply under the BSA.

Ethical Consideration and Challenges in Forensic Phonetics-

Ethical Considerations in Forensic Phonetics

1. Objectivity and Impartiality

 Experts must remain neutral—serving the court, not prosecution or defense.

 Avoid confirmation bias (e.g., being influenced by case details or suspect identity).

2. Transparency of Methods

 All methods used (acoustic, auditory, statistical) should be clearly documented.

 Courts must understand how conclusions were reached.
 Black-box systems (e.g., some AI models) raise ethical concerns if not explainable.

3. Validity and Scientific Reliability

 Techniques must be grounded in established principles like Acoustic Phonetics.

 Methods should be empirically tested and peer-reviewed.
 Avoid overstating the strength of evidence.

4. Use of Likelihood Ratios

 Interpretation often relies on Bayesian Inference.

 Ethical duty:
o Present results as probabilities, not certainties.
o Avoid misleading statements like “this is the same speaker.”

5. Informed Use of Technology

 Automatic speaker recognition systems must be:

o Validated
o Calibrated
 Ethical concern: blind reliance on software without understanding limitations.

6. Privacy and Consent

 Voice recordings are biometric data.

 Issues include:
o Unauthorized recording
 o Use of personal speech samples without consent
 Must comply with legal standards on data protection.

7. Reporting Limitations

 Experts must clearly state:

o Audio quality issues
o Duration limitations
o Environmental noise
 Ethical obligation: disclose uncertainty, not hide it.

8. Avoiding Misleading Testimony

 Use clear, non-technical language in court.

 Avoid exaggeration or selective reporting.
 Ensure the judge/jury correctly interprets findings.

Key Challenges in Forensic Phonetics

1. Poor Audio Quality

 Background noise, distortion, compression.

 Common in real-world recordings (calls, surveillance).

2. Short or Incomplete Samples

 Limited speech reduces reliability of comparisons.

 Hard to extract stable features.

3. Speaker Variability

 Same speaker can sound different due to:

o Emotion
o Health
o Age
o Intentional disguise

4. Channel Mismatch

 Differences between:
o Telephone vs studio recordings
o Different microphones/devices
 Affects acoustic measurements.

5. Accent and Language Diversity

 Regional accents and multilingual speakers complicate analysis.

 Requires appropriate population reference data.
6. Lack of Standardization

 No universally adopted protocols across labs.

 Variation in:
o Methods
o Reporting scales
o Interpretation

7. Statistical and Database Limitations

 Reliable evaluation depends on large, representative datasets.

 Many languages/dialects lack sufficient databases.

8. Risk of Cognitive Bias

 Contextual information (e.g., suspect background) may influence analysis.

 Blind testing procedures are not always implemented.

9. Courtroom Misinterpretation

 Judges/juries may misunderstand:

o Probabilistic conclusions
o Likelihood ratios
 Risk of overvaluing or undervaluing evidence.

10. Emerging AI Challenges

 Deepfake voices and synthetic speech.

 Raises questions about:
o Authenticity verification
o New forms of voice spoofing
 Ethical concern: keeping pace with technology.

Real Life Case Study in Forensic Speaker Identification-

Nira Radia Tapes Controversy (2010) – Detailed Case Study-India

1. Who was involved?

 Niira Radia – Corporate lobbyist representing major business groups

 Politicians (especially linked to telecom ministry appointments)
 Senior journalists
 Corporate leaders

She represented companies like:

 Tata Group
 Reliance Industries
2. How it started

The Income Tax Department suspected:

 Tax evasion
 Possible illegal financial activities

They got permission to tap Niira Radia’s phone calls (2008–2009).

 Around 5,800+ calls recorded

 Conversations included discussions on:
o Cabinet formation
o Business deals
o Media influence

3. Leak to the Public (2010)

In 2010:

 Selected recordings were leaked to media outlets like magazines and online platforms
 The tapes created a national political storm

4. Link to the 2G Spectrum Scam

The tapes suggested:

 Lobbying efforts to influence who becomes Telecom Minister

 Alleged role in favoring certain corporate interests

Focus centered on:

 Appointment of A. Raja as Telecom Minister

This linked the tapes to the 2G spectrum allocation controversy, one of India’s biggest
corruption scandals.

5. Role of Media & Journalists

The tapes revealed:

 Conversations between Radia and prominent journalists

 Allegations that some journalists acted as intermediaries between:
o Politicians
o Corporates

This raised serious questions about:

 Media ethics
 Conflict of interest
6. Forensic Speaker Identification Role

This is where forensic science became crucial.

Why it was needed:

Many individuals:

 Denied the voices were theirs

 Questioned authenticity of recordings

Agencies involved:

 Central Bureau of Investigation (CBI)

 Central Forensic Science Laboratory (CFSL)

What experts did:

1. Voice Sample Collection

 Suspects were asked to give voice samples under controlled conditions

2. Audio Analysis

 Removal of noise from phone recordings

 Segmentation of speech

3. Spectrographic Comparison

 Matching voiceprints (visual patterns of speech)

4. Feature Analysis

 Pitch
 Tone
 Formants
 MFCC (Mel-Frequency Cepstral Coefficients)

Result:

 Many voices were matched with high probability

 Reports indicated recordings were authentic and not tampered significantly

The Lindbergh Kidnapping Case (1932) USA

Key People

 Charles Lindbergh – Famous pilot, father of the victim

 Anne Morrow Lindbergh – Mother
 Charles Augustus Lindbergh Jr. – 20-month-old child
 Bruno Hauptmann – Accused kidnapper

Timeline of Events
March 1, 1932 – The Kidnapping

 The Lindbergh baby was abducted from his home in Hopewell.

 A handmade wooden ladder was found outside the window.
 A ransom note demanding $50,000 was left behind.

Ransom Communication

 Over the next weeks, multiple ransom notes were sent.

 A retired school principal, John F. Condon, acted as intermediary.

The Cemetery Meeting

 Condon met the kidnapper in a Bronx cemetery.

 Charles Lindbergh listened from a distance and briefly heard the kidnapper’s voice.
 The kidnapper spoke with a German accent and called himself “John.”

Tragic Discovery

 Despite paying the ransom, the child was not returned.

 On May 12, 1932, the baby’s body was found in woods near the Lindbergh home.

Investigation Breakthrough
Marked Money

 Police tracked the ransom money using serial numbers.

 In 1934, a gas station attendant noticed suspicious bills.

Arrest

 The money led to Bruno Hauptmann, a German immigrant carpenter.

 Police found:
o $14,000 of ransom money in his garage
o Wood matching the ladder in his home
Role of Voice Identification
What Happened

 Charles Lindbergh testified in court that:

o He recognized Hauptmann’s voice
o It sounded like the man from the cemetery

Nature of the Evidence

 This was not scientific forensic analysis (no spectrograms at the time).
 It was based on:
o Memory of voice
o Accent and tone

Problems

 The identification occurred over 2 years later

 Human memory of voices is unreliable
 The encounter was brief and under stress

Today, this would be considered weak evidence on its own.

Trial and Verdict (1935)

 The trial took place in Flemington.

 Hauptmann was convicted and sentenced to death.
 He was executed in 1936.

Other Forensic Evidence Used

This case is famous because it used multiple early forensic methods:

1. Wood Analysis

 The ladder used in the kidnapping matched wood from Hauptmann’s attic.

2. Handwriting Analysis

 Ransom notes were compared to Hauptmann’s writing.

 Experts claimed strong similarities.

3. Currency Tracking

 Serial numbers of ransom bills were recorded and traced.