01.Evidence Collection
Locard's exchange principle
Every contact leaves a trace, in both directions. Foundation of trace evidence.
Documentation order
Wide → mid → close-up (with scale) → sketch → notes. All before anything is touched.
Wet biological packaging
Air-dry first, then paper. Never plastic — moisture kills DNA in days.
Bloodstain on porous surface
Excise the substrate + a clean control patch. Air-dry. Paper bag.
Search patterns
Grid (small enclosed) · Spiral (centred outdoor) · Strip (large open) · Zone (multi-area).
Panchnama (BNSS)
Witnessed inventory: each item, location, seal number, signed by two independent panch. For digital evidence, include SHA-256 hash on the memo.
Memory hook
Life first → document the disturbance → preserve what's left.
02.Analytical Instruments
Microscopy choice
Compound (thin, transmitted) · Stereo / Comparison (3-D opaque, raking light) · PLM (birefringence) · SEM-EDX (morphology + elements).
SEM modes
SE = topography (surface). BSE = atomic-number contrast (composition).
Beer-Lambert
A = ε × c × l. Linear range 0.1–1.0 absorbance. Above 1.0 → dilute and re-measure.
FTIR vs ATR
ATR-FTIR is non-destructive on solids (paint chips, fibres, films). KBr-pellet for ground powders. Diamond / ZnSe / Ge crystal, ~0.5–2 µm penetration.
Raman vs IR (water)
Water absorbs IR strongly, scatters Raman weakly → Raman wins for aqueous samples.
ICP-MS vs ICP-AES vs AAS
ICP-MS: ppt, multi-element, isotopic. ICP-AES: ppb-ppm, multi-element. AAS: ppb, single-element.
GSR (ASTM E1588)
Pb + Sb + Ba on a 1–5 µm spheroid = GSR-consistent. SEM-EDX is the workhorse.
GC-MS confirmation (SWGDRUG Cat-A)
Library mass-spectrum match + co-injected reference retention time. Both required.
LC-MS/MS
Polar / thermolabile / non-volatile compounds. Tandem MRM for high specificity. Use deuterated internal standards for matrix-effect correction.
XRD
Bragg's law: nλ = 2d sin θ. Match d-spacings vs ICDD-PDF library. Distinguishes polymorphs (cocaine HCl vs base).
LIBS
Pulsed laser → microplasma → atomic emission. Portable, near-non-destructive, depth profiling.
TLC R_f
R_f = spot distance / solvent distance. Reproducible identifier on the same plate. For basic drugs on silica: chloroform-methanol-ammonia (90:9:1).
Memory hook
Match the technique to the sample, not the other way around.
03.Pattern Evidence
Bloodstain angle of impact
sin θ = width / length. Tail points toward direction of travel.
Velocity classification
HVIS < 1 mm (gunshot, mist) · MVIS 1–4 mm (blunt force) · cast-off ≥ 4 mm (swung weapon).
Glass: 3R rule
Radial cracks form Right angles on the Reverse side from impact.
Glass: radial-termination rule
New cracks stop at old cracks. Earlier impact's radials dominate.
Bullet through glass
Cone of debris points in the direction the bullet was travelling.
Tool-mark hierarchy
Class (type) · Sub-class (manufacturing batch — the trap) · Individual (random acquired marks).
Tyre marks
Track width + wheelbase → vehicle size class. Cuts and embedded debris individualise.
Hair DNA
Tag with tissue → nuclear STR (individualising). Shaft only → mtDNA HV1/HV2 (maternal-line).
Fibre transfer hierarchy
Primary (direct) · Secondary (one intermediate) · Tertiary (two). Each step attenuates.
Footprint stature
Foot length ≈ 0.15 × stature. Class-level evidence; report a range, not a point.
Bite marks
Now considered weak (NAS 2009, PCAST 2016). Exclusion only — never identification.
Memory hook
Pattern evidence: class first, then look for individualising features.
04.Math & Statistics
Likelihood ratio (LR)
LR = P(E | H_p) / P(E | H_d). Strength of evidence. Court owns the prior.
Bayesian combination
Posterior odds = prior odds × LR. Multiply, never divide.
Prosecutor's fallacy
LR ≠ probability of guilt. Posterior depends on prior; high LR alone doesn't convict.
Empirical rule
68% within 1 SD · 95% within 2 SD · 99.7% within 3 SD.
SEM vs SD
SD = data spread. SEM = SD / √n = mean precision. Use SD for variability, SEM for confidence in mean.
Coefficient of variation
CV = SD / mean × 100%. Dimensionless. <5% for well-developed methods.
Random match probability (RMP)
Product of per-locus match probabilities (independence + HWE). 13 CODIS loci → ~10⁻¹³.
Theta (θ) correction
Population substructure correction. 0.01 (homogeneous) to 0.03 (subdivided). Conservative default 0.03.
Type I vs Type II
Type I = false positive (α). Type II = false negative (β). Power = 1 − β.
Memory hook
Forensic stats reports the LR; the court combines with prior to reach posterior.
05.Voice Authentication
Formants
F1, F2, F3, F4 = vocal-tract resonances. Adult male: F1 ~730, F2 ~1090, F3 ~2440 Hz on /a/. Adult female: F1 ~850, F2 ~1220, F3 ~2810 Hz.
F0 (pitch)
Adult male 85–180 Hz · Adult female 165–255 Hz · Children > 250 Hz.
Pitch-shift disguise
Raises F0 but not formants → acoustic inconsistency detectable on spectrogram.
GSM codec effects
Narrowband 300–3400 Hz. F4 lost, F3 at edge. Band-limit reference recording to match.
Cepstrum + MFCCs
Separates source (F0) from filter (formants). MFCCs are the standard ASR feature.
Voice quality
H1-H2 (breathiness), spectral tilt, jitter (F0 cycle-to-cycle), shimmer (amplitude). GRBAS / CAPE-V auditory scales.
Whispered speech
No F0. Formants + voice quality + prosody + lexis still discriminate.
ASR generations
GMM-UBM → i-vector → x-vector → ECAPA-TDNN. Modern EER < 1% on clean recordings.
ASVspoof
Anti-spoofing benchmark for TTS / voice-cloning / replay attack detection.
Memory hook
Speaker comparison = F0 + formants + prosody + idiolect + voice quality, reported via LR.
06.Video Analysis
Frame-rate verification
Count frames during a known-duration event. DVR nominal ≠ actual.
H.264/H.265 frame types
I (intra, full data, best) · P (predicted from past) · B (bidirectional). Extract I-frames for analysis.
DVR clock skew
Photograph DVR display + accurate reference at seizure. Document offset. Never reset clock.
De-interlacing
Comb-tooth artefact on moving objects in interlaced video. De-interlace before plate / face analysis.
Photogrammetric height
H_actual = h_image × distance / focal_length. Use scene reference markers + lens calibration.
CCTV speed
time = Δframes / fps. speed = distance / time. Verify frame rate first.
PRNU
Sensor noise fingerprint per camera instance. High cross-correlation → camera-source match.
Lens distortion
Wide-angle = barrel distortion. Calibrate with checkerboard, undistort before measurement.
Super-resolution caveat
Multi-frame SR = legitimate. AI hallucination = invents content; not defensible without disclosure.
Deepfake detection
Pixel artefacts + temporal flow + rPPG (heart rate) + double-compression + ML classifiers.
Memory hook
Preserve original, work on a copy, document every step, never invent detail.
07.Criminalistics & Engineering
IS 269 (cement)
Insoluble residue ≤ 4% · MgO ≤ 6% · 28-day strength per grade. Failures = adulteration / unsoundness.
IS 1786 (rebar)
Fe 500 = yield ≥ 500 N/mm². Below spec → structural risk + remediation.
IS 456 (concrete)
150 mm cubes, 28-day compressive test. M20 = 20 MPa minimum.
Fire (NFPA 921)
V-pattern apex points to origin. Floor-level apex + serpentine pour + ILR (GC-MS) = arson.
Arc bead vs fire-melt
Arc: sharp boundary, dendritic, oxide enrichment (rapid cool). Fire-melt: blobby, gradual transition, oxidised throughout.
Currency (RBI)
Intaglio relief is the touch test — counterfeits use offset / inkjet without relief.
Explosives
TNT M+ at m/z 227 (GC-MS). RDX/HMX/PETN by HPLC-MS/MS. Inorganic ions by IC. NH₄NO₃ phase IV at room temp by XRD.
ASTM E1618 IL classes
Gasoline · Light/Medium/Heavy petroleum distillate · Isoparaffinic · Naphthenic-paraffinic · Normal alkane · Oxygenated · Miscellaneous.
Diatom test
Diatoms in bone marrow (femur / sternum) with water species match → ante-mortem drowning.
Palynology
Pollen species + seasonal flowering windows → location / season / PMI.
Counterfeit gold (XRF)
Cu/Zn ~4:1 + small Au = brass with thin gold plating. Density (Au 19.3 vs brass 8.5) confirms.
Memory hook
Indian Standards anchor the analysis: IS 269 / 456 / 1786 / 4151 / 4031 series.
08.Collision Investigation
Skid speed
v = √(2gμd). Dry asphalt μ ≈ 0.7. 25 m skid → ~67 km/h. Lower bound on pre-skid speed.
ABS pattern
Intermittent dotted marks. Skid formula doesn't apply; use EDR data.
Yaw mark
v = √(g × r × μ). Curved mark with cross-tyre striations. Speed at grip-loss.
Pedestrian throw
Searle: v = √(2gd × μ / (1 + μ²)). Han: v ≈ 4.06 + 0.89√d. Report range, not point.
Conservation of momentum
m_A v_A + m_B v_B = (m_A + m_B) v_f for collinear inelastic. Vector decomposition for off-axis.
EDR (SAE J1698)
Last 5–15 s pre-impact: speed, throttle, brake, steering, ABS, seat-belt, delta-v.
AEB analysis
TTC = distance / closing speed. Required braking = v² / (2a). Compare against available distance.
Static Stability Factor
SSF = (track / 2) / CG height. Sedan 1.3–1.5 · SUV 1.0–1.2 · loaded tanker 0.6–0.8.
Helmet (IS 4151)
Shell + EPS liner crush + chinstrap + certification + drop-test attenuation.
Airbag + seatbelt
Complementary, not alternative. Airbag without seatbelt → facial / chest injury from inflation.
Memory hook
Always check: ABS or not, EDR available, μ measured (not generic).
Premium book
Read the full Forensic Physics book
These are the cheat-sheet notes. The full hand-built book — 80 pages, 15+ diagrams, worked examples, formula reference and an index of standards — is available inside your account.
Open the book →Next steps
- Take the FACT Forensic Physics mocks — Mocks #6–#11 cover this syllabus at three difficulty levels plus a 100-question full-length.
- Read related deep-dives on the blog
© ForensicSpot · Quick notes for educational use. Pair with primary references (Saferstein, Sharma, NFPA 921, ASTM, ENFSI, BIS standards) for full study.