Log File Format

Bugs and Support
1

Hi,

I am having some artefacts in another service from having 3 points with the same power in a row.

When checking the zwift log I came across the following line after the activity, does anyone know what it means, in particular what is being measured by min, max, mean and stddev?

[20:27:06] DEBUG LEVEL: POWER SOURCE: Interval Stats: device = KICKR BIKE 7945 217, count = 38400, min = 0.000008, max = 0.300002, mean = 0.083143, stddev = 0.029104, success-ratio = 1.000000

2

Are you using a Kickr Bike v2 with race mode enabled?

The statistics measure time intervals between power readings (in seconds). Your mean = 0.083143 (83ms) indicates roughly 12Hz sampling.

Your KICKR BIKE has Race Mode enabled, which transmits power at 10Hz instead of standard 4Hz (250ms intervals). This explains why you’re getting far more frequent data points than the other service expects.

Why you see duplicate power values:

  • Standard trainers send power every 250ms (4Hz)
  • Race Mode sends every 100ms (10Hz)
  • Your log shows ~83ms (possibly Zwift’s internal sampling rate)
  • At steady efforts, power legitimately stays constant across multiple rapid readings

The fix: The other service likely expects 4Hz data. Either disable Race Mode on your KICKR BIKE, or ask the service provider if they support high-frequency (10Hz) power data from Race Mode-enabled trainers.

Your trainer is working correctly - it’s a compatibility issue with services designed for standard 4Hz broadcasts.

Statistics Breakdown

  • count = 38400: Total power samples received (~2h 40min ride)
  • min = 0.000008: Shortest gap between readings (8 microseconds) - likely buffered duplicates
  • max = 0.300002: Longest gap (300ms) - brief dropout or cadence change
  • mean = 0.083143: Average time between readings (83ms ≈ 12Hz update rate)
  • stddev = 0.029104: Timing variability (29ms jitter)
  • success-ratio = 1.000000: 100% of packets received successfully

Race Mode Details

  • KICKR BIKE v2 Race Mode enables 10Hz broadcast rate via WiFi or KICKR Direct Connect
  • Updates power upon each flywheel revolution
  • During high speed sprints can be up to 10 times per second
  • Captures small peaks in wattage that aren’t reflected with slower broadcast rates
  • Standard ANT+ power = 4Hz (250ms intervals)
  • Race Mode = 10Hz (100ms intervals)
  • This log shows ~12Hz (83ms intervals) - likely Zwift’s internal sampling rate

Why Duplicate Values Occur

  1. Power hasn’t changed - At steady state efforts, consecutive readings 83ms apart legitimately show identical watts
  2. Wahoo’s power smoothing - KICKR applies internal averaging that can hold power constant for 200-300ms
  3. Resolution limits - Power is reported in 1-watt increments; small changes get rounded to the same value

(I used claude code with zwiftalizer log parser source code in context to help me write this. It may be tripping, but I actually think it’s spot on :slight_smile:)

1 Like
3

Thanks,

Only had zwift open for around an hour which aligns closer with the number of samples at 12 hz. 12 hz could also just be the operating frequency of the flywheel, wahoo seems to update once per revolution. In that case it might just be a bit of aliasing as a result of zwift recoring power every second.

Assiming power output is a pure sine wave at 90 rpm 6 hz recording would be the minimum needed to perfectly reproduce the power input.

Look up nyquist frequency.