The six steps
Almost every podcast, whatever its size, goes through the same six steps.
- Plan: the format, the length, the guest and the questions.
- Record: capture each voice cleanly, ideally on its own track.
- Edit: remove noise, false starts, repeats and long pauses.
- Sound and mix: add the intro, outro and music, then bring every voice to the same level.
- Master and check: set the final loudness and measure the file against a spec.
- Publish: upload to a host, write the title and notes, and release.
Plan and record
Decide the format, the target length and how often you will publish before the first recording. Those three choices set the effort of every step that follows.
Recording is where most of the quality is decided. A clean recording is worth more than a rescued one, and the recording checklist covers what to set up before you press record.
Edit, mix and master
Editing is listening. The editor hears the whole take, marks noise and level jumps, then removes false starts, repeats and long pauses at points where the cut does not show. Cleaning reduces room echo, hum and clicks.
The mix puts every voice at the same level and sits the music underneath. Mastering sets the loudness of the finished file so it plays at a comfortable level next to other shows; the loudness and delivery spec explains the numbers.
Publish
Publishing is separate from production. A podcast host stores your files and creates the feed that apps such as Apple Podcasts and Spotify read. You or your team write the title, the description and any show notes, upload the file and choose the release date.
Who does each step
A small team can do all six steps in-house. For many teams the step that stalls first is editing, because it needs focused hours and a trained ear.
At radio.app the work starts at the edit: you plan and record, we edit, mix, master and check, and you publish. What we do and what stays with you is set out on one page.
How long it takes
The time an episode takes depends on its length, the number of voices, the state of the recording and the rounds of changes you need. A clean single-voice episode is the quickest case; a long panel with overlapping voices takes the longest.