Ordering clips and unifying sample rates or channels locally
Joining voice intros, sponsor reads, and recorded chapters requires matching sample clocks and channel counts before concatenating samples. Combining MP3, WAV, M4A, and OGG tracks in custom order inside browser memory protects private podcasts and unreleased demos from third-party upload queues.
Reordering tracks with Move up and Move down controls
Dropping files onto Drop audio files here in any order or clicking Load sample clips builds an indexed playlist under Clip order. Each card displays exact duration, native sample rate, and channel count alongside Move up, Move down, and Remove buttons.
If an individual clip still carries sloppy head noise or a cough in the middle, open Audio Cutter to trim unwanted intros before building your final sequence. Moving any entry automatically triggers a fresh worker join and updates your output waveform.
Resampling mixed inputs to 48,000 Hz or 44,100 Hz stereo
Compressed browser audio decodes locally into linear floating-point channels matching the MDN AudioBuffer PCM storage reference. Directly concatenating a 22,050 Hz voice memo onto a 44,100 Hz music stinger without resampling would play the voice at double speed and chipmunk pitch.
Inside our worker, Catmull-Rom cubic audio resampling converts every track cleanly to 48,000 Hz or 44,100 Hz, inserting a low-pass anti-aliasing filter at 0.44 times the target rate whenever downsampling exceeds 10%. Selecting Stereo duplicates mono channels evenly across left and right buses, whereas Mono averages multi-channel stems.
Why does an equal-power crossfade prevent volume dips?
Linear fades drop combined acoustic energy by roughly 3 dB at the splice midpoint because uncorrelated audio waveforms sum by power rather than straight voltage. Sine and cosine weighting keeps the squared sum of incoming and outgoing gains equal to one, preserving steady perceived loudness across every transition.
Balancing sine-cosine overlap within the 45% duration limit
Crossfade blending, resampling, and peak scanning run inside everyday.worker.ts using the MDN Web Workers API specification so long multi-chapter episodes stay responsive. During Equal-power crossfade, the outgoing track follows cos(0.5 * pi * t) while the incoming track rises along sin(0.5 * pi * t).
Although Crossfade (ms) spans 100 ms to 4,000 ms, the worker caps overlap at 3,000 ms and at most 45% of either neighboring clip’s sample length. That 45% boundary stops a short transition chime from bleeding across three consecutive tracks at once.
Preventing boundary pops in Silent gap mode with 5 ms fades
Selecting Silent gap turns Gap (ms) active from 0 ms to 3,000 ms (250 ms by default) to insert clean digital silence between spoken segments. Rather than splicing abruptly onto raw samples, the worker applies 5 ms anti-click cosine micro-fades (applyCosineFadesInPlace) at every clip start and end.
Comparing 5 ms anti-click gap fades versus 45 percent overlap crossfades comes down to content type. If stitched tracks have wildly different vocal loudness before joining, pass quiet stems through integrated LUFS loudness normalization first so chapters match consistently.
Merge audio controls for silent gaps, crossfades, and levels
Reviewing Transition mode, Channel layout, Sample rate, and Normalize peak to -1 dBFS clarifies how each option shapes the final mix before clicking Merge audio. Output WAV files follow standard PCM headers documented in the Library of Congress RIFF WAVE format guide for universal DAW compatibility.
| Control | Range or options | Default setting | Worker DSP behavior | Best workflow fit |
|---|---|---|---|---|
| Transition mode | Silent gap / Equal-power crossfade | Silent gap | Inserts silence or sin/cos overlap |
Spoken episodes vs continuous music mixes |
| Gap (ms) / Crossfade (ms) | 0–3,000 ms gap / 100–4,000 ms fade |
250 ms / 600 ms |
Caps overlap at 3,000 ms and 45% length |
Podcast pauses vs musical segues |
| Channel layout | Stereo / Mono | Stereo | Duplicates mono or averages stereo stems | Music-rich podcasts vs lean speech files |
| Sample rate | 48000 Hz / 44100 Hz |
48000 Hz |
Catmull-Rom cubic resample + anti-alias LPF | Video/broadcast vs CD/audio distribution |
| Normalize peak to -1 dBFS | Checked / Unchecked | Checked | Scales global peak to -1 dBFS or soft-limits |
Maximizing volume safely across joined clips |
If one recorded segment contains distracting background fan static or electrical buzz between words, strip distracting fan hiss or electrical buzz beforehand so quiet pauses match cleanly across joined segments.
Tip: Leaving Normalize peak to -1 dBFS unchecked preserves intentional relative volume differences between quiet spoken intros and punchy stingers, while
softLimitSampleautomatically catches any crossfade sum exceeding0.99amplitude.
Exporting merged WAV files and chapter cue sheets offline
Once rendered, the result panel reports Total duration, Peak level in dBFS, Chapters count, WAV size in bytes, and a 160-bar SVG Merged audio waveform. Every temporary audio buffer stays in client memory until you save your track or close the browser tab.
Copying MM:SS chapter markers and saving .cue sheets
Because crossfades pull later tracks earlier on the timeline while gaps push them back, manual timestamp math breaks quickly across five or ten clips. Under Chapter markers & CUE sheet, the worker strips file extensions cleanly and computes exact startSec and endSec boundaries after all transitions.
Tracks under one hour format cue lines as MM:SS Title, switching automatically to HH:MM:SS Title at 3,600 seconds or longer. Clicking Copy chapters places the formatted timestamp list onto your clipboard for YouTube or podcast descriptions, while Download .cue sheet saves merged-audio.cue.
Choosing 16-bit PCM WAV, AAC M4A, or Opus WebM output
Clicking Download merged WAV exports uncompressed 16 bits per sample RIFF PCM (merged-audio.wav) with zero compression artifacts. On browsers supporting WebCodecs AudioEncoder, Download .m4a encodes 128 kbps AAC aligned with the Library of Congress MPEG-4 AAC format description alongside 128 kbps Opus via Download .webm.
Podcasters or musicians who want to turn a merged highlight reel into a custom phone alert afterward can cut merged melodies into 30-second mobile ringtones directly in browser memory. Uncompressed WAV output guarantees full archival fidelity when handing unified stems to video editors.