01
Understand the three headline readings
A Timegrapher listens to the beat of a mechanical movement and estimates several values. Rate is expressed as estimated seconds gained or lost per day. Amplitude, expressed in degrees, describes the balance’s arc under the configured measurement assumptions. Beat error, expressed in milliseconds, describes asymmetry in the timing between alternating beats.
Read the values together rather than searching for one universal pass mark. For acoustic measurement, verify the movement-specific lift angle where possible. The setting is an input to the amplitude calculation, so an incorrect default or assumed angle directly biases the reported amplitude. Movement design, state of wind, test position, and measurement method also influence interpretation.
02
Create repeatable capture conditions
Wind the watch consistently, place it securely, choose a quiet environment, and keep the microphone or phone in a stable relationship to the case. Record the position and allow the trace to settle before saving a result. Changing several conditions at once makes it impossible to know whether a different number reflects the movement or the measurement setup.
Example: capture one reading dial up after a consistent wind, then repeat that method on another day. If you also want crown-down context, store it as a separate position instead of averaging it invisibly with dial up. A named sequence turns isolated screenshots into comparable observations and makes a surprising value easier to reproduce.
03
Look for patterns across position and time
Mechanical watches can behave differently by position and as available power changes. Compare like with like first, then inspect the spread between positions and the direction of change over time. One noisy capture should prompt a repeat under controlled conditions; a persistent shift across comparable sessions is more useful context for a service conversation.
Do not infer a specific internal fault from a reading alone. Similar symptoms can have different causes, and an acoustic capture cannot inspect lubrication, wear, magnetism, water resistance, or damaged components directly. Store notes about wind state, position, unusual noise, impacts, or recent service beside the Patina AI reading so the history does not lose its conditions.
04
Know when the reading reaches its limit
A phone-based Timegrapher does not replace professional service diagnosis. It cannot determine overall movement health, authenticate a watch, or tell you which repair is required. An incorrect configuration, unusual escapement, unstable sound, or environmental interference can also make a result unreliable. Treat the output as measurement context rather than a remote inspection.
Repeat a concerning reading under consistent conditions and consult a qualified watchmaker when the concern persists, changes suddenly, follows an impact or moisture event, or matters to a purchase. Share the recorded positions and history rather than a single cropped number. A professional can combine appropriate instruments with direct examination and the movement-specific knowledge the app does not have.