Noise Nuisance Log

Overview

Recurring noise is one of the most common residential conflicts, and the side with the record almost always wins. "It's loud every night" carries no weight with a landlord, an HOA, or a city noise office; "14 incidents in 30 days, 11 after 22:00, median 47 minutes, mostly bass music, including weeknights before work days" gets a letter sent, a lease clause cited, or an inspector dispatched. The gap between suffering and resolution is almost never the noise itself — it's documentation discipline.

This skill is that discipline, distilled:

1. 10-second incident logging — time, duration, type, estimated loudness, impact (woken up / couldn't work / kid disturbed), and a note.
2. Pattern analysis — time-of-day histogram, weekday split, night-hours rate, per-type totals, trend vs the previous period.
3. Ordinance context — typical municipal quiet hours (22:00–07:00) and common decibel limits by zone; every incident inside quiet hours is flagged. Adjust to your local code (the reference shows how to find it in 10 minutes).
4. Impact score — a 0–100 severity index from frequency, night share, duration, and sleep disruption, so "how bad is it really" has a defensible number.
5. Complaint letter generator — a factual, non-hostile letter to landlord/HOA/city citing your documented pattern, the relevant clause concepts (quiet enjoyment, nuisance), and a specific remedy request.

When to Use

• Recurring neighbor noise (music, TV, footsteps, arguments, instruments)
• Barking dogs (yours-to-complain-about, or defending against a complaint about yours)
• Construction/renovation outside permitted hours
• Businesses/venues/HVAC affecting your home
• Preparing for a landlord conversation, HOA complaint, mediation, or a city noise office filing — anywhere a pattern beats an anecdote
• Don't use for: one-off events (call non-emergency police line if it's happening now), domestic-violence-sounding situations (that's emergency services), or workplace noise (different framework entirely).

Commands

# Log an incident (start time defaults to now; duration in minutes)
python3 scripts/noise_log.py log --type music --duration 45 --loudness loud \
    --impact "woke us up" --note "bass, clearly audible in bedroom"
python3 scripts/noise_log.py log --type barking --start "2026-09-05 06:15" --duration 20 \
    --loudness moderate --impact "woke baby"
python3 scripts/noise_log.py log --type footsteps --duration 30 --loudness moderate

# Types: music, tv, talking, party, barking, construction, renovation, hvac,
#        traffic, venue, footsteps, other

# The patterns — this is what goes in the letter
python3 scripts/noise_log.py stats            # all-time summary
python3 scripts/noise_log.py stats --days 30
python3 scripts/noise_log.py stats --neighbor "upstairs 4B"   # if you log --neighbor

# Severity score with interpretation
python3 scripts/noise_log.py score

# Time-of-day and weekday distribution
python3 scripts/noise_log.py histogram

# Generate the complaint letter (landlord | hoa | city)
python3 scripts/noise_log.py letter --to landlord \
    --recipient "Jane Smith, Smith Property Mgmt" \
    --your-name "Alex Tenant" --address "Apt 4A, 12 Oak St"
python3 scripts/noise_log.py letter --to city --your-name "Alex Tenant" \
    --address "12 Oak St" --source "venue across the street"

# Manage
python3 scripts/noise_log.py list --limit 10
python3 scripts/noise_log.py export --csv noise-log.csv

Log lives at ~/.noise-log.json (--file to override).

What the Output Looks Like

Noise Log — 30 days ending 2026-09-08
  incidents: 14   (3.4/week)   total 9.2 hours
  night-hours (22:00–07:00): 11 of 14  (79%)  ⚠ 11 potential quiet-hours violations
  by type:   music 8 · party 3 · footsteps 2 · barking 1
  weekday vs weekend: 9 weekday · 5 weekend
  peak window: 22:00–01:00 (9 incidents)
  impact: sleep disruption reported in 10 incidents (71%)
  severity score: 68/100 — HIGH (frequent + mostly night + sleep impact)

Common Pitfalls

1. Logging only the worst nights. The pattern is the evidence; a log with 3 entries in a month reads as "occasional," however brutal those nights were. Log everything you'd mention in conversation.
2. Guessing decibels. The tool asks for a band (faint/moderate/loud/very loud), not a number — because uncalibrated phone dB apps are often inadmissible-quality evidence, and a defensible band beats a pseudo-precise number. If you want real numbers, the reference explains what makes a measurement credible (calibrated meter, documented position) — and that your timestamped pattern usually matters more.
3. Hostile letters. The generator is deliberately factual and calm. Insults, threats, and speculation ("they're doing it on purpose") sink complaints; dates, durations, and impact get results.
4. Wrong quiet hours. 22:00–07:00 is the common default, but codes vary (some cities use 23:00; some have daytime limits too). Set your local hours with --night-start/--night-end after checking the municipal code — the reference shows where to look.
5. Skipping contemporaneous notes. Add a one-line --note (what it sounded like, what you did). "Called the front desk at 23:40, no answer" is a fact that later becomes leverage.
6. Anonymous complaints as step one. The letter generator includes your name because identified, documented complainants get processed; anonymous ones get filed. (If you fear retaliation, the reference covers escalation paths that protect you.)

Verification Checklist

• Every incident logged same-day (contemporaneous = credible)
• --night-start/--night-end match your local ordinance
• stats --days 30 shows the real pattern (aim for ≥2 weeks before escalating)
• Letter reviewed before sending: facts only, specific remedy requested
• Kept a copy of everything sent, with dates — the paper trail is the asset

One-Shot Recipes

Upstairs neighbor, week 3 of nightly music

python3 scripts/noise_log.py log --type music --duration 45 --loudness loud --impact "woke us up"
# ...two weeks of this...
python3 scripts/noise_log.py stats --days 14
python3 scripts/noise_log.py letter --to landlord --recipient "Jane Smith, Acme Mgmt" \
    --your-name "Alex Tenant" --address "Apt 4A, 12 Oak St"

Construction starting at 6:00 on Saturdays

python3 scripts/noise_log.py log --type construction --start "2026-09-06 06:05" \
    --duration 150 --loudness "very loud" --note "jackhammer; permit says 08:00 start"
python3 scripts/noise_log.py letter --to city --your-name "Alex Tenant" \
    --address "12 Oak St" --source "site at 14 Oak St"