Back to all guides

Sound Diagnostics

Engine Sound Diagnostic Checklist

A step-by-step checklist for recording engine noises so the AI sound analyzer can give you the most accurate diagnosis: vehicle info, complaint, recording equipment, environment, required clips, locations, supporting OBD-II data, and final verification.

Vehicle information

  • Vehicle year
  • Make
  • Model
  • Engine size
  • Engine code (if known)
  • Mileage
  • Fuel type (gas / diesel / hybrid)
  • Transmission type

Customer complaint

  • No start
  • Hard start
  • Rough idle
  • Misfire
  • Knocking
  • Ticking
  • Squealing
  • Rattle
  • Belt noise
  • Exhaust noise
  • Loss of power
  • Stalling
  • Other

Engine condition

  • Cold engine
  • Fully warmed up
  • Outside temperature
  • Check Engine Light on?
  • Stored DTCs (if available)

Recording equipment

  • External microphone (preferred)
  • Phone microphone
  • Wind protection installed
  • Phone in Airplane Mode
  • Remove phone case if it blocks microphones

Recording environment

  • Quiet area
  • No radio
  • No talking
  • Hood open
  • Vehicle in Park / Neutral
  • Parking brake set

Required audio recordings

Capture each of these separately — short, focused clips beat one long ramble.

  • Cold start
    • 20–30 seconds
  • Warm idle
    • 30 seconds
  • Raise RPM slowly
    • Idle → 2500 RPM
    • Hold 5 seconds
    • Return to idle
  • Snap throttle
    • 3 quick throttle snaps
  • Hold RPM
    • 1500 RPM
    • 2000 RPM
    • 3000 RPM
    • Hold each for 10 seconds

Recording locations

Capture each test from every viable position — location narrows the source faster than any other input.

  • Front of engine
  • Driver side
  • Passenger side
  • Top of engine
  • Near accessory belt
  • Tailpipe
  • Under vehicle (if safe)

Optional advanced tests

  • A/C ON
  • A/C OFF
  • Steering turned fully left / right
  • Transmission in Drive (foot on brake)
  • Electrical load (headlights, blower, rear defrost)
  • Cranking only (no start)
  • Power-braking (where safe)

Helpful supporting data

If you have any of these, upload or paste them alongside the audio — they meaningfully raise diagnostic confidence.

  • OBD-II live data
  • Freeze-frame data
  • Fuel trims
  • Misfire counters
  • Engine RPM
  • Coolant temperature
  • Intake air temperature
  • MAF / MAP readings
  • Battery voltage
  • Ignition timing
  • Relative compression results
  • Vacuum readings
  • PicoScope waveforms

Maintenance history

  • Last oil change
  • Oil viscosity used
  • Spark plugs replaced?
  • Ignition coils replaced?
  • Timing belt / chain serviced?
  • Recent repairs?
  • Accident history?

AI confidence boost

Give the AI multiple reference clips so it can compare relative changes across conditions and locations.

  • Cold start
  • Hot idle
  • Light acceleration
  • Heavy acceleration
  • Deceleration
  • Belt area
  • Valve cover
  • Intake
  • Exhaust
  • Alternator
  • Water pump
  • A/C compressor
  • Power steering pump

Final verification

Before you submit, walk this list one more time.

  • Recording is clear
  • No wind noise
  • No voices
  • No music
  • Entire engine audible
  • At least 30-second idle recording
  • Multiple engine locations recorded
  • OBD-II data attached (if available)

Features that can significantly improve diagnostic accuracy

A wish-list of capabilities that raise confidence when combined with a good recording.

  • AI noise reduction to isolate engine sounds.
  • Automatic detection of recording location (front, side, tailpipe) using prompts or AI.
  • Spectrogram analysis in addition to raw audio.
  • Frequency analysis (FFT) to identify knocks, ticks, bearing noise, and belt squeal.
  • Automatic engine RPM estimation from audio when OBD-II data isn't available.
  • Comparison against a library of known-good recordings for the same engine.
  • Combining audio analysis with OBD-II live data, DTCs, and vibration data (from the phone's accelerometer) for a more reliable diagnosis.

Related guides

Ready to run the checklist?

Record your clips, gather the supporting data, and upload it all to the sound analyzer.

Open the sound analyzer