The EDL: how non-destructive podcast editing becomes real audio

In Podcast Studio, deleting a word does not erase samples from the source recording. The editor records the decision in an edit decision list, or EDL. The EDL says which time ranges to keep, which sections to mute, how tracks should mix, and whether Studio Sound should run when the final master is built.

That gives you a useful safety net: the Web Studio capture or imported source stays intact, while each export creates a new audio result from the current decisions.

The transcript is the editing surface

Clicking a transcript word seeks to that moment. Marking filler words or shortening silence turns the transcript state into delete regions. The remaining continuous words become ordered keep segments with an in-point and out-point on the original timeline. A mute region silences audio without collapsing time.

Track settings travel with the edit: volume, pan, solo and mute state, start offsets, and music ducking flags. Visual overlays can also be part of a video-capable export. Autosave stores the EDL state, not a rewritten copy of the source object.

Export compiles the decisions

When you export, the client turns that state into a validated JSON payload. At least one keep segment is required. Standard exports can be MP3, WAV, or AAC. The Pro Audio surface also has a DAW package path for stems. Studio Sound is carried as an explicit flag so the worker sees the setting the editor shows.

Validation runs before the job is queued. The gateway resolves the media asset IDs to storage keys on the server and checks export and tier entitlements there. A browser cannot point a worker at an arbitrary URL, and a client cannot grant itself Creator-level processing by changing a field.

What the mastering worker produces

The worker downloads the referenced tracks, extracts audio, applies per-track delay, volume, pan, and music ducking, then mixes to a working WAV. It applies the EDL keep list and concatenates the selected sections. Where possible, the cut list uses stream copy so unchanged material is not needlessly re-encoded.

If Studio Sound is enabled and the account is Creator or above, FFT denoising runs after the EDL assembly. Loudness normalization follows, and the delivery step produces the master MP3 in the 256 kbps class. That master becomes the MASTER asset used by downloads and RSS enclosures; the raw Studio source does not become the public enclosure.

Why keep the source untouched?

• You can change the keep list and export again without re-recording.
• The original capture remains available for archive or legal review.
• A failed export cannot corrupt the only copy of the session.
• A Web Studio draft keeps the recording as its source of truth.

The tradeoff is that an edit is not a finished listener file until export completes. The draft is a set of decisions; the master is those decisions rendered into audio.

Filler removal, silence shortening, Studio Sound, chapters, and show notes follow the same separation. Editing tools change EDL state or episode metadata. The mastering worker creates the new audio object only when you export.

How to work safely in a long episode

Start by making the obvious transcript edits, then play the joins around each keep segment. A cut that looks clean in text can remove the breath or setup that makes the next sentence understandable. Because the source remains intact, you can widen a segment and export again instead of trying to recover a deleted waveform.

Use mute regions when timing matters. A mute leaves the timeline in place, which is useful when a pause, music cue, or visual moment needs to stay aligned. Use keep-segment cuts when you want the finished episode to move past material entirely.

The export payload is the handoff between those choices and the worker. If validation rejects it, the source and the saved EDL are still available; the failed request does not create a half-master that looks complete.

A simple model

Source recording
        ↓
EDL: keep, mute, mix, processing flags
        ↓
Validated export payload
        ↓
Mastering worker → working WAV → loudness → master MP3
        ↓
RSS enclosure and download

If you know non-linear video editors, the EDL is the sequence. If you know a cue sheet, it is the list of moments to use. The benefit is the same in both cases: the source remains available when your editorial opinion changes.

The edit becomes real audio at export. Until then, the source is safe and the decisions are reversible.

Change the keep list, adjust the mix, turn Studio Sound on or off, and compile another master. The source remains the recording you actually captured.