R↗RACESOUNDSENGINE AUDIO / FIELD LAB
PROTOTYPE 0.1.4 Audio offline

FEEL THE RESPONSE. HEAR THE DIFFERENCE.

The engine room.

A test bench for the sound your drive is missing.

Local processingNo coordinates logged or uploaded
Choose an engine, then start audio. Use the bench while parked.
VIRTUAL ENGINEBENCH / MANUAL
ROAD SPEED—mph
VIRTUAL LOAD15%
STANDBY

Virtual RPM and gears are simulated. Recorded engines use a single loop each.

Set up your run

01 / INPUT
—Loading local audio library
850 rpm
IDLE7,000 RPM
30%
INPUT INTERVAL ⓘ—msNo GPS observations
READING AGE—msInput timestamp → now
CONTROL ACK · P95 ⓘ—msMessage round trip, not audible delay
REPORTED OUTPUT—msBrowser estimate, not speaker measurement

Speed response

Last 30 seconds · mph

WAITING
Simulated truthObservedEngine input

Bench mode auditions RPM directly. Select a simulated drive or live GPS for speed traces.

Audio monitor

Waveform · after volume and limiter

— dBFS
PEAK —RMS —LIMITER —

Digital signal monitor. It cannot detect distortion or delay added by the car’s speakers.

Response tuning

02 / COMPARE
1200 ms
50 ms
8 ms
Tuning assumption, not measured RPM
Playback ratio 1.00×Range limited to 0.55–3.00×

Capture the evidence

03 / EXPORT

Log a run, compare settings, and keep an audio take. Data stays here unless you download it or choose an encrypted QR link.

Capture idle · 0 data pointsAudio takes: up to 30 sec

QR to phone uploads an encrypted snapshot for 30 minutes. Anyone with the QR link can open it.

Sync pulse gives a visual flash and a beep for an external video test. Exported timings and internal audio recordings do not measure physical speaker latency.

SYNC
Detailed diagnostics CAPABILITIES / TIMING / SESSION

Control ACK includes return delivery and main-thread scheduling. Base and output latency are browser reports, may be zero or unavailable, and do not establish accelerator-to-speaker latency. UI frame gaps and long tasks are main-thread signals, not measured audio dropouts.

Audio provenance & testing guide LICENSES / LIMITATIONS

01 / Listen while parked

Start at low volume. Sweep RPM, compare sources at the same load, and listen with crossfade at 0 and 8 ms. Pitch clamps and single-loop stretching will be audible.

02 / Compare response

Use the same simulated drive with prediction on and off. Export separate captures. The truth trace exists only in simulation; GPS mode has no independent reference.

03 / Test in the Tesla

Use a trusted HTTPS address. Set up while parked; have a passenger handle diagnostics. Record the screen and speakers externally for a sync-pulse test and physical pedal-to-sound comparison.

Read full audio credits & download details ↗

Take this capture to your phone

Preparing the encrypted link…

Only an encrypted copy is stored on race.shaa.one. The decryption key is in this QR link. The link expires after 30 minutes; it can be removed sooner.