Splitting Long Audio: Cut Units, Precision and Naming

Cut by content: drop the knife at topic changes and agenda boundaries so each segment stands alone — best for archiving, search and on-demand listening. Cut by duration: fixed 30- or 60-minute chunks for platform or email size limits. Prefer content cuts when possible — uniform cuts routinely split sentences in half, which is the worst listening experience.

Before content cuts, skim the recording once and note each cut point in minutes:seconds. The ten extra minutes buy you segments that each make sense on their own; the trade is always worth it.

Cut segment by segment with a trim tool against your noted time points; second-level precision is enough. Leave half a second of margin between segments, erring on the generous side — a clipped word is nearly impossible to fix, while half a second of breath is nothing. Time-based trimming does not re-encode, so the split itself is lossless.

For size-limited scenarios (email attachments, per-file platform caps), check each segment’s size after cutting: re-export any oversized segment alone at one bitrate step lower. Bitrates across segments need not be identical, but keep them in the same tier — audible jumps between segments are noticeable in continuous listening.

Name segments with numeric prefixes: 01-Opening, 02-Topic-one. Numbers keep sorting stable on any device; descriptions keep the content recognizable three months later. Never archive under names like “New recording (1)” — that just re-shuffles your mess.

Consistency check: segment-processed material often ends up segment-inconsistent — volume jumps, uneven noise floors. Normalize loudness across all segments to one level so continuous listening does not sound like a different person mid-stream. For published material, finish with one compression pass at the target platform’s spec.