Reviewed concrete example
Treat structure as evidence, not noise
Knowing how to handle intros breakdowns and outros prevents one average from flattening several different behaviors. An intro may be free-time, a breakdown may preserve pulse with fewer cues, and an outro may slow deliberately. Each section deserves either its own observation or an explicit exclusion from a steady-window claim.
Before counting, listen through the transitions and sketch a simple map: free intro, groove entry, verse, breakdown, return, ending. The map can stay in personal notes. The local calculator needs only the selected timestamps and intervals.
Intros: wait for the pulse to settle
A pickup, spoken count, ambient swell, or rubato opening can precede the stable tempo. Do not begin at 00:00 merely because it is convenient. Find the first repeatable pulse after the tempo settles and place the start on a musical boundary.
If the intro has its own stable ostinato, measure it separately and label the observation in non-sensitive terms such as “intro pulse.” Do not combine its intervals with the later groove unless the same pulse and rate genuinely continue through the boundary.
For example, an eight-second free opening followed by 32 intervals across 16 steady seconds should produce a 120 BPM groove observation from the steady span. Counting the same 32 intervals across all 24 seconds would yield 80 BPM, a number that describes neither the free intro nor the groove usefully.
Breakdowns: listen for the hidden grid
A breakdown may remove drums while harmony, bass, movement, or effects continue on the same underlying grid. If the pulse remains trackable, the section can be measured. State which cue was followed. If the pulse becomes ambiguous, stop the prior window before the breakdown and begin a new one when a reliable cue returns.
Half-time arrangement can change perceived pulse without changing source tempo. A backbeat moving from every two beats to a slower accent can tempt the observer to report half the prior BPM. Compare half and double candidates and listen to phrase length. Do not force both sections into one pulse level without documenting the choice.
Pauses, drops, and edits
A dramatic silence can occur inside an otherwise metered phrase. Whether it belongs in the tempo window depends on whether the underlying beat continues and the observer can place the next boundary confidently. If the silence suspends time or an edit removes elapsed material, separate the windows.
Video edits can also splice performances. A visual cut alone does not prove an audio edit, and an audio edit may be visually subtle. The worksheet cannot detect either. If a timestamp span produces an unexpected rate, listen around the cut and choose a passage that does not cross it.
Outros: describe ramps with several windows
A fade-out can retain stable tempo even as volume falls. Measure before the cue becomes too weak. A ritardando or accelerando should not be reduced to a single “song BPM” if the change itself matters. Select an early ending window and a later ending window, then report their numerical difference neutrally.
Worked example: an outro’s first twenty-second window contains 40 intervals, or 120 BPM. A later twenty-second window contains 36 intervals, or 108 BPM. Absolute difference is 12 BPM, and the later rate is 10 percent lower relative to 120. This describes the entered sections; it does not prove intent or rate of change between them.
Build section-specific observation cards
A useful set might contain “main groove A,” “breakdown pulse,” and “late outro,” each with timestamps, count convention, speed, and source-time rate. Labels should remain non-sensitive and local. The comparator can report range and median for comparable readings, while the drift checker preserves chronological direction for two sections.
Do not average section values merely to fill a catalog-style field. If a downstream application requires one representative rate, explain the selection rule—such as the longest stable main section—and retain the other cards.
Limits of the method
Manual observation cannot distinguish every edit, infer a hidden conductor beat, determine whether silence remains metered, or decide an arrangement’s intended pulse. Whole-second timestamps and weak cues constrain precision. Section choices are human judgments that should remain visible in notes.
The site has no access to the video, audio, URL, thumbnails, captions, or player controls. It does not call the YouTube API, listen through a microphone, upload media, or save a structural map. The viewer watches elsewhere and enters only the chosen numeric window.
Frequently asked questions
Should I include a count-in?
Only if it belongs to the pulse and question you are documenting. Otherwise begin at the stable musical boundary after it.
Is a drumless breakdown unmeasurable?
Not necessarily. Another reliable cue may preserve the grid. Stop if the pulse cannot be followed confidently.
Does a fade-out change BPM?
Not by itself. Falling volume can hide cues while tempo remains stable; measure before the boundary becomes uncertain.
How should I report a ritardando?
Use separate chronological windows and describe their rates and difference rather than one unexplained average.
Write a simple section map, choose a window that stays inside one behavior, and create a labeled card. When the transition is the subject, measure both sides independently instead of spanning the change.
