A timegrapher log can look beautifully scientific while quietly telling you a half-truth. The dots march, the rate number flickers, amplitude winks at you, and suddenly your watch bench feels like a tiny courtroom with no judge. If your readings bounce around, split into ghost traces, or change when you touch the caseback, you need a way to score the signal-to-noise ratio before trusting the result. This guide gives you a practical DIY rubric you can use today, in about 15 minutes, to separate useful timing data from bench-top static, microphone drama, and wishful thinking.
Why Timegrapher Noise Matters
A timegrapher is not a magic window into the soul of a watch. It is a microphone, a timing algorithm, a clamp, a setting menu, and your bench habits all wearing one small lab coat. When those parts behave, the log is useful. When they do not, the numbers become persuasive confetti.
Signal-to-noise ratio scoring is a way to ask a simple question: how much of this log is the watch, and how much is everything around the watch? That distinction matters because hobbyists often make expensive decisions from cheap noise. A watch that looks erratic may only be sitting poorly on the mic. A watch that looks healthy may be hiding a position problem because the test was too short.
I once watched a perfectly serviceable Seiko read like it had swallowed a drawer of paperclips. The crime scene was not magnetism or a dying movement. It was a rubber desk mat touching the timegrapher stand and feeding vibration into the pickup. Tiny thing, large embarrassment. The bench has a sense of theater.
For hobbyists, the goal is not laboratory perfection. It is decision-grade repeatability. You want to know whether a log is strong enough to support one of three actions: leave the watch alone, adjust cautiously, or stop and inspect more deeply.
- Score the log before you judge the watch.
- Look for repeated patterns across positions, not one heroic reading.
- Separate measurement noise from mechanical behavior.
Apply in 60 seconds: Before changing anything, save a screenshot or photo of the current trace and label the position, lift angle, and test length.
What Signal-to-Noise Means in Watch Timing
In plain English, the signal is the watch’s actual beat behavior. Noise is anything that makes the timegrapher hear, calculate, or display that behavior incorrectly. The problem is that noise often dresses up as knowledge. It wears a tidy number on the screen and acts innocent.
What counts as signal?
Useful signal includes rate trend, beat error trend, amplitude trend, positional differences, and stability over time. A strong signal repeats under similar conditions. If dial up reads +4 seconds per day three times in a row with similar amplitude and beat error, that is a useful clue.
Signal is not always pretty. A real mechanical issue can produce ugly traces. The point is that the ugly trace should be repeatable, position-aware, and explainable. If the pattern follows the watch across careful retests, it becomes evidence.
What counts as noise?
Noise can come from microphone pressure, loose caseback contact, wrong lift angle, low mainspring state, background vibration, poor beat detection, magnetization, or user handling. Noise can also come from logging too quickly. A 20-second sample can flatter or frighten you, depending on the watch’s mood and the algorithm’s coffee intake.
If you have seen sudden amplitude changes from small clamp adjustments, compare that with your own notes or this related guide on timegrapher mic pressure versus amplitude. Contact pressure is one of the sneakiest little gremlins in hobby timing.
The practical definition
For home watch work, a high signal-to-noise timegrapher log has five traits: stable trace, adequate sample length, controlled setup, repeated position data, and consistent notes. A low signal-to-noise log has jumpy data, missing context, short samples, unexplained setting changes, and readings that cannot be reproduced.
Measurement professionals, including NIST, talk often about traceability, uncertainty, and repeatability because numbers need context to be meaningful. You do not need a metrology lab in your spare room. You do need a habit that keeps your timing logs from becoming decorative fiction.
The DIY Rubric: 100-Point Timegrapher Log Score
This rubric scores the quality of the log, not the health of the watch. That distinction is the hinge. A watch can be unhealthy and still produce a high-quality log. A watch can be healthy and produce a low-quality log because your setup was doing interpretive dance on the bench.
Use this as a repeatable hobbyist rubric. It is not a certification tool. It is a decision filter for whether a reading is strong enough to guide adjustment, comparison, or further inspection.
Category 1: Trace clarity, 25 points
Trace clarity measures whether the timegrapher is locking onto the beat cleanly. A clean single trace earns more confidence than a fuzzy spray of dots. Occasional speckling is normal. A double trace, wandering trace, or random scatter means you should pause before believing the headline number.
If your log shows two parallel tracks, read this companion article on the double trace mystery. Double traces can come from real beat issues, poor pickup, case resonance, or detection confusion. The trick is not to panic. Panic is just bad data wearing a cape.
Category 2: Setup control, 20 points
Setup control means the test conditions are stable. Same timegrapher, same lift angle setting, same microphone pressure, same position, similar wind state, and no weird table vibration. If you moved the crown, bumped the clamp, changed the angle, and then compared the two readings, the log should receive fewer points.
One evening, I logged a vintage hand-wind watch beside a laptop fan. The trace looked nervous, almost shy. I moved the timegrapher to a heavier table and the dots calmed down. The movement had not healed itself. The room had stopped gossiping into the microphone.
Category 3: Sample length, 15 points
A short log is a snapshot. A longer log is a sentence. For hobby scoring, 30 seconds may be enough for a rough check, but 2 to 5 minutes gives you better confidence. If you are testing drift, warm-up, power reserve behavior, or position change, longer windows matter.
For slow changes over a few minutes, compare your notes with timegrapher reading drift over 10 minutes. Some watches need time to settle after movement, winding, or position change.
Category 4: Position coverage, 20 points
A single position can be useful, but it rarely tells the full story. At minimum, hobbyists should log dial up, dial down, crown up, crown down, crown left, and crown right when diagnosing. If you only want a quick “is this watch wildly off?” check, one position is fine. If you plan to regulate, position coverage becomes much more important.
Beat error can behave very differently by position. That is why a single excellent dial-up reading can still miss trouble hiding on the vertical positions. For a deeper related example, see beat error versus position.
Category 5: Context notes, 20 points
Context notes are the bread crumbs that let future-you understand present-you. Record movement type, lift angle setting, wind state, position, test length, case condition, microphone contact, temperature if relevant, and whether the watch was recently moved, demagnetized, opened, or adjusted.
I have found old timing screenshots with no labels and the emotional usefulness of a napkin map found in a rainstorm. Was it dial up? Fully wound? Before regulation? After coffee? No one knows. Label your logs. Your future self deserves a small lantern.
Show me the nerdy details
A mechanical watch timegrapher estimates rate by listening for acoustic events tied to escapement action. Noise enters when the pickup hears extra vibration, misses impulses, misclassifies beat timing, or applies settings that do not match the movement. Lift angle mainly affects displayed amplitude, not rate in the same direct way, but a wrong value can make amplitude comparisons misleading. For scoring, treat rate stability, trace coherence, and repeatability as separate signals. A log with a stable rate but wrong lift angle may still be useful for rate trend, while a log with unstable trace detection should be treated cautiously for all metrics.
How to Run a Clean Timegrapher Log
The best scoring system in the world cannot rescue a chaotic test. Good logs start before the first dot appears. Treat your setup like a kitchen scale: not sacred, not fussy, but useless if someone keeps leaning on the counter.
Step 1: Stabilize the bench
Put the timegrapher on a firm surface. Avoid laptop fans, speakers, air purifiers, desk motors, typing, and the heroic thump of closing drawers. If you use a phone app rather than a dedicated timegrapher, isolate the microphone and keep the room quiet.
For vibration contamination, this related test on speaker vibration contamination is worth reading. Background vibration can make an innocent watch look guilty.
Step 2: Wind consistently
Wind state affects amplitude. For manual watches, use a consistent full-wind protocol unless you are testing power reserve behavior. For automatic watches, decide whether you are testing after wrist wear, after rotor winding, or after manual top-up, then write it down.
Automatic winding habits can create surprisingly different results. If you are comparing wrist behavior against bench results, see manual top-up versus pure rotor winding.
Step 3: Confirm lift angle, then keep it fixed
Lift angle affects amplitude display. If you do not know the lift angle, do not pretend certainty. Use the manufacturer’s documentation when available. If not, label it as assumed. For trend tracking, consistency is often more useful than false precision.
Two good internal follow-ups are whether a wrong lift angle still allows trend tracking and how to infer unknown lift angle. The short answer: rate trend may remain useful, but amplitude confidence depends on the setting.
Step 4: Let the watch settle
After moving a watch into a new position, give it 30 to 60 seconds before logging. Some movements settle quickly. Some vintage pieces need a little dignity and a quiet chair. Do not judge the first few seconds too harshly unless the goal is to test transition behavior itself.
Step 5: Log each position the same way
Use a fixed routine. For example: 60 seconds settling, 120 seconds logging, screenshot, note, next position. If you change the timing window halfway through, your comparisons become softer than a worn leather strap.
- Control vibration before blaming the movement.
- Use the same wind state for comparable logs.
- Let each position settle before recording.
Apply in 60 seconds: Write one standard logging routine on a sticky note and tape it near your timegrapher.
Scorecard Table: Grade Your Log Without Guessing
Here is the core money block: a scoring table you can use every time. Print it, copy it, or keep it in a notes app. The point is not to become a stern clipboard person. The point is to stop one wobbly log from bossing you around.
| Category | Max Points | Full-Credit Standard | Deduct When |
|---|---|---|---|
| Trace clarity | 25 | Single, stable trace with few random dots. | Double trace, scatter, unstable lock, or sudden unexplained jumps. |
| Setup control | 20 | Same mic pressure, lift angle, position method, and quiet bench. | Clamp moved, case shifted, room vibration, inconsistent contact. |
| Sample length | 15 | At least 2 minutes after a short settling period. | Quick snapshots under 30 seconds or no settling time. |
| Position coverage | 20 | Relevant positions tested for the question being asked. | One-position conclusions for multi-position problems. |
| Context notes | 20 | Wind state, lift angle, position, duration, date, and changes recorded. | Screenshots with no labels or missing test conditions. |
How to interpret the score
| Score | Confidence Level | Best Use | What Not to Do |
|---|---|---|---|
| 85-100 | Strong | Regulation decisions, before-and-after comparisons, service notes. | Do not assume wrist performance will match perfectly. |
| 70-84 | Useful | Trend checks, rough diagnosis, deciding whether to retest. | Do not make aggressive adjustments from one position. |
| 50-69 | Weak | Screening only; repeat with tighter controls. | Do not diagnose a movement from this alone. |
| 0-49 | Noise-heavy | Find setup errors, microphone issues, or logging mistakes. | Do not regulate, demagnetize repeatedly, or open the watch based only on it. |
Think of 85 or above as “decision-grade.” Think of 70 to 84 as “useful but not smug.” Below 70, the log may still contain clues, but it needs backup. Below 50, you are mostly measuring the bench, the clamp, the room, or a tiny electrical goblin in the pickup.
Mini Calculator: Your Timegrapher SNR Score
This simple calculator uses three inputs to estimate the quality of your log. It is intentionally modest. It will not know whether your movement is dry, magnetized, or damaged. It only helps you decide whether the log itself deserves confidence.
DIY Log Quality Calculator
Enter honest scores. The calculator rewards clarity, repeatability, and context.
Total score: Click the button to calculate.
Buyer checklist for timegrapher gear
If you are buying a dedicated timegrapher or upgrading from a phone app, use a practical checklist. The nicest screen in the world does not help if the pickup is fussy and the stand wobbles like a folding chair at a county fair.
- Stable microphone stand: The watch should sit securely in common positions without odd pressure.
- Clear lift angle setting: You should be able to set and record lift angle easily.
- Readable trace display: Dots should be visible enough for trace-quality judgment.
- Consistent beat detection: Reviews should mention stable readings across common movements.
- Quiet power behavior: Avoid units known for electrical noise or poor grounding.
- Export or photo-friendly display: You need records, not just a momentary number.
Cost-wise, many hobbyists start with an entry-level dedicated unit or a careful phone-app setup. A pro bench tool may be justified if you service watches often, but for occasional regulation checks, your method matters more than your equipment flex. The watch does not care how expensive your tool is. It cares whether you can hear it properly.
Visual Guide: The Four Noise Traps
Visual Guide: Four Noise Traps in Timegrapher Logs
Too much or too little contact can change pickup quality and displayed amplitude.
A quick snapshot may catch settling behavior instead of stable timing.
No wind state, lift angle, or position note means the log cannot be compared later.
Fans, speakers, typing, and hollow tables can feed false rhythm into the pickup.
Decision card: What the pattern is trying to tell you
If rate is stable but amplitude seems odd: Check lift angle, wind state, and mic pressure before worrying about the movement.
If beat error changes by position: Repeat the same positions and compare with positional beat-error behavior.
If the trace splits or ghosts: Retest with better contact and quieter surroundings before assuming escapement trouble.
If amplitude collapses suddenly: Check setup first, then consider actual mechanical causes if the pattern repeats.
For sudden amplitude changes that repeat after setup checks, compare with this guide on amplitude collapse thresholds. A real collapse is worth attention, but a fake collapse is just a bench prank with decimal places.
Common Mistakes That Make Good Watches Look Sick
Most timegrapher mistakes are not dramatic. They are small, repeatable, and annoyingly believable. They do not announce themselves with trumpets. They quietly tilt the result until your adjustment plan starts walking into the weeds.
Mistake 1: Trusting the first 10 seconds
The first few seconds after placing a watch on the microphone are often settling time. The watch, case, clamp, and pickup are finding their conversation. Logging immediately can confuse transient behavior with stable behavior.
Mistake 2: Comparing different wind states
A watch tested fully wound on Monday and barely wound on Thursday may not produce comparable amplitude. That does not automatically mean something changed mechanically. It may only mean your test condition changed.
Mistake 3: Ignoring temperature and wear context
Cold rooms, warm wrists, and resting positions can all affect real-world behavior. If your watch behaves differently after winter walks, this internal article on cold wrist syndrome gives a practical angle on wrist conditions and timing reality.
Mistake 4: Treating amplitude as a moral score
Amplitude is important, but it is not a single-number certificate of virtue. A lower reading may be caused by wind state, lift angle setting, position, or mic contact. A healthy movement can also vary by design, age, service history, and position.
Mistake 5: Forgetting crown position
Crown orientation affects how the movement sits, how gravity loads the balance staff, and how the case contacts the pickup. If your readings change oddly, this article on crown position during measurement is a practical companion.
Mistake 6: Chasing magnetization without a pattern
Magnetization can cause sudden timing changes, but not every weird log is magnetic. If you demagnetize repeatedly without a repeatable symptom, you may be treating the bench instead of the watch. For pattern recognition, see magnetization signatures in logs.
- Do not compare full-wind amplitude to low-wind amplitude.
- Do not judge positional behavior from one position.
- Do not treat an unlabeled screenshot as evidence.
Apply in 60 seconds: Add “wind state” and “position” to every timing note before saving the log.
Who This Is For and Not For
This rubric is for hobbyists who want better confidence before making small regulation decisions, comparing before-and-after logs, or deciding whether a watch deserves professional attention. It is also for collectors who want to stop spiraling over one dramatic screenshot from a watch sale listing.
This is for you if:
- You own a timegrapher or use a phone-based timing app.
- You compare rate, amplitude, and beat error across positions.
- You want a neutral way to score log quality.
- You are deciding whether to regulate, retest, demagnetize, or seek service.
- You want better notes for future maintenance decisions.
This is not for you if:
- You need a professional service diagnosis for a valuable or fragile watch.
- You are working on a sealed, water-resistant watch without proper tools.
- You expect one score to replace inspection, cleaning, lubrication, or parts assessment.
- You are handling radium-era dials, damaged crystals, broken mainsprings, or unknown contamination.
One collector sent me six screenshots from a newly bought watch and asked whether the escapement was doomed. The screenshots had no positions, no wind state, and three different lift angles. The answer was not “doomed.” The answer was “start again, kindly.” Measurement mercy is real.
When to Seek Help
Timegrapher scoring is safe as a logging habit, but watch work can become physically risky or financially painful once you open cases, remove hands, handle old dials, or adjust delicate parts. A screwdriver in the wrong place can turn a simple timing question into a parts hunt.
Safety and responsibility note
This guide is for hobbyist measurement and decision support. It is not a professional repair manual. If a watch is valuable, rare, sealed for water resistance, under warranty, or personally irreplaceable, get help from a qualified watchmaker before opening it or adjusting the regulator. Be especially careful with vintage watches that may have fragile luminous material, brittle gaskets, sharp case edges, or damaged crystals.
OSHA’s general hand-tool safety guidance is meant for workplaces, but the common-sense ideas still fit a home bench: use the right tool, control sharp edges, and avoid forcing parts. Watchmaking tools are small, but they can still bite.
Get professional help when:
- The watch has sentimental or market value you cannot comfortably risk.
- Amplitude is low across positions after a full wind and careful retest.
- Beat error is high and you do not have the skill to adjust the stud carrier or beat setting safely.
- The watch stops in certain positions.
- The trace is erratic even after a clean setup and repeated tests.
- You suspect water damage, impact damage, missing lubrication, or broken parts.
Consumer protection groups and the FTC have long warned that repair claims and warranty language should be understood carefully. For watches, that means reading warranty terms before opening a caseback, especially on modern watches where one curious twist can change your service options.
Build a Repeatable Log Template
A good log template turns scattered screenshots into a useful maintenance record. It also lowers emotional temperature. Instead of “my watch is cursed,” you get “dial up, fully wound, 2-minute log, clean trace, +6 seconds per day.” Much calmer. Fewer candles required.
Quote-prep list for a watchmaker
If you decide to contact a watchmaker, send useful information without burying them in 47 screenshots and a novella about your wrist habits. A clean message gets better answers.
- Brand, model, movement if known, and approximate age.
- Whether the watch is manual, automatic, quartz, or hybrid.
- What problem you observe on wrist, not only on the timegrapher.
- Recent events: drop, water exposure, magnet exposure, shipping, service, or regulation attempt.
- Your best high-score logs in several positions.
- Your question: regulate, diagnose, service estimate, or inspect only.
Simple log template
| Field | Example | Why It Matters |
|---|---|---|
| Date and time | 2026-06-24, 8:15 PM | Helps compare logs across days. |
| Wind state | Fully wound, 30 turns | Amplitude depends heavily on power state. |
| Position | Dial up | Gravity changes timing behavior. |
| Lift angle | 52 degrees, assumed | Needed for amplitude interpretation. |
| Duration | 60 sec settle, 120 sec log | Separates settling from stable data. |
| Observed trace | Single trace, slight scatter | Adds quality context beyond numbers. |
Short Story: The Watch That Failed Only on Paper
A hobbyist brought a mid-size automatic to a meet-up and said it had become “unregulatable.” He had photos: +2, then -18, then +11, then a strange double trace, all from the same afternoon. The watch looked like it had developed opinions. We repeated the test on a heavy table, fully wound it, set the same lift angle, waited one minute between positions, and logged two minutes per position. The score jumped from noise-heavy to useful. The watch was not perfect, but it was not chaotic. The earlier readings came from shifting caseback contact and a mic stand sitting half on a soft pad. The practical lesson was simple: before you adjust the watch, adjust the experiment. The bench is part of the instrument, even when it pretends to be furniture.
- Use the same fields every time.
- Send only your cleanest logs when asking for help.
- Record setup changes immediately, not later from memory.
Apply in 60 seconds: Create a note titled “Watch Timing Log Template” and paste the fields from the table above.
FAQ
What is a good signal-to-noise ratio score for timegrapher logs?
For hobby use, 85 to 100 is strong enough for careful decisions if your test matches the question you are asking. A score from 70 to 84 is useful for trends but should be confirmed before regulation. Anything below 70 deserves a cleaner retest.
Can I regulate a watch from one timegrapher reading?
You can, but it is usually not wise. One reading may be enough for a rough check, but regulation decisions are stronger when you have stable logs across relevant positions. A single dial-up number can miss vertical-position behavior.
Does wrong lift angle ruin every timegrapher log?
No. A wrong lift angle mainly affects displayed amplitude. Rate trends may still be useful if the setting stays consistent, but amplitude comparisons become weaker. Label assumed lift angles clearly so you do not overstate the result.
Why does my timegrapher show a double trace?
A double trace can come from poor pickup, case resonance, escapement behavior, beat detection confusion, or real mechanical trouble. First retest with a quiet bench, stable contact, and consistent position. If it repeats under clean conditions, it becomes more meaningful.
How long should I run a timegrapher test?
For quick screening, 30 to 60 seconds can show obvious problems. For decision-grade hobby logs, use about 60 seconds of settling and 2 minutes of recorded data per position. Longer tests help when you are studying drift or power behavior.
Why does amplitude change when I move the watch on the microphone?
Microphone pressure and contact angle can affect pickup quality and displayed amplitude. That is why setup control matters. If amplitude changes dramatically after a tiny clamp change, retest before assuming the movement has a mechanical problem.
Can a timegrapher predict real wrist accuracy?
It can suggest likely behavior, but it cannot perfectly predict your wrist. Real use includes motion, temperature, resting position, wind state, and wearing habits. Timegrapher logs are best used with actual wear tracking.
Should I demagnetize a watch every time the log looks weird?
No. Demagnetizing is useful when symptoms fit, such as a sudden large rate gain after magnetic exposure, but weird logs can also come from bench vibration, bad contact, short sampling, or wrong settings. Find a repeatable pattern first.
Is a phone timegrapher app good enough for hobby scoring?
It can be useful for screening if the microphone setup is controlled and the app detects beats reliably. A dedicated timegrapher is usually easier to repeat across positions, but disciplined setup still matters more than gadget pride.
What score should I require before buying a used mechanical watch?
If a seller provides timegrapher photos, treat them as supporting evidence, not a guarantee. Look for labeled positions, reasonable sample duration, visible trace, and consistent conditions. Unlabeled screenshots are better than nothing, but not by much.
Conclusion: Trust the Pattern, Not the Panic
The hook at the beginning was simple: a timegrapher log can look official while quietly mixing watch behavior with bench noise. Now you have a way to challenge it. Score the log first. Judge the watch second. That order protects your movement, your wallet, and your weekend mood.
In the next 15 minutes, run one clean two-minute log in dial-up position after a short settling period. Score it using the 100-point rubric. Then repeat one vertical position using the same routine. If both logs score high, you have useful evidence. If they score low, you have not found a sick watch yet. You have found a noisy experiment that needs tidying.
Good timing work is not about staring harder at one number. It is about building a trail of repeatable clues. The quieter the test, the louder the watch can speak.
Last reviewed: 2026-06
