Latency / Buffer Size Calculator mechanism

Model scope

The arithmetic is valid only under the explicit 0 ms I/O-overhead assumption. The instrument cannot be verified as fully specified until the overhead estimate and round-trip toggle behavior are defined. The honest label correctly indicates that driver and DAC hardware latency add-ons are excluded, but it should also disclose that the I/O overhead term is either omitted or an explicitly defined estimate.

• The base formula bufferSize/sampleRate*1000 is correct for the duration of one buffer, assuming bufferSize is in samples and sampleRate is in samples per second (Hz).
• Multiplying by 2 is a reasonable simplified estimate for input-plus-output buffer latency, but it is not universally the complete round-trip latency formula because interface, driver, conversion, and scheduling delays may differ.
• The phrase '+ I/O overhead estimate' is incomplete: no overhead value, input, default, or calculation method is specified, so the stated formula cannot produce a unique output.
• The round-trip toggle semantics are not explicitly defined for whether the I/O overhead estimate is applied in one-way mode, round-trip mode, or both.

Units and conventions

• bufferSize is a non-negative number of samples per buffer.
• sampleRate is a positive number of samples per second, expressed in Hz.
• 1000 converts seconds to milliseconds.
• The test calculations assume the unspecified I/O overhead estimate is 0 ms; this assumption must be made explicit before implementation.
• A round-trip estimate is modeled as two buffer durations: 2*bufferSize/sampleRate*1000.

Inspectable vector

One of 4 source-bound review vectors:

{
  "inputs": {
    "bufferSize": 256,
    "sampleRate": 48000,
    "roundTrip": true
  },
  "expected_outputs": {
    "latencyMs": 10.666666667
  },
  "working": "One buffer: 256 / 48000 s = 0.005333333333 s. Convert: 0.005333333333 * 1000 = 5.333333333 ms. Round trip: 5.333333333 * 2 = 10.666666667 ms. I/O overhead assumed: 0 ms. Total = 10.666666667 ms."
}

Failure boundary

BREAK_CLAIMED · SELF_FLAGGED

A sample rate of zero causes an undefined division and must be rejected. Additionally, the specification provides no numeric I/O overhead estimate, so exact total latency is underdetermined unless the implementation documents or supplies H; the outputs above therefore include the computable base latency and express the total as base plus H.

Other records

Sources and evidence

These pointers bind the page to evidence. They do not grant authority or collapse verification into deployment.

• VERIFIED_LEDGER.json
• evidence-tranche4/tranche4-results.json
• BOUNDARY_HAZARD_CATALOG.json