Updated October 2026 · Editing audio well
A reliable editing workflow runs in a fixed order: back up the raw file, listen through once, fix the faults, cut the structure, clean and level the voices, add music and effects, then check and export. Following the same order every time prevents rework and keeps problems from spreading through the edit.
Beginners often open an editor and start clicking, cutting and processing without a plan. The result is usually a file that takes many hours, sounds uneven and is hard to finish. An editing workflow is a fixed sequence of steps that solves the problems in a sensible order. Each stage prepares the next, so the work moves forward rather than looping back. Once the order is familiar, it becomes automatic, and the edit becomes faster and calmer.
This post sets out a workflow for a typical spoken episode, with notes on how it changes for music and video. It is a starting point, and experienced editors adapt it, but the principle of working in order holds.
Before you open the editor, copy the raw recording to a safe place and do not change it. Work on a project file that points to a copy. The raw file is the one thing you cannot replace, and a mistake in an edit can always be undone if the original is safe. This single habit prevents more disasters than any other part of the workflow.
Name the project with the episode and the date, so you can find it later and tell it apart from earlier versions.
Play the whole recording straight through, on headphones, without making changes. Note the problems and the strongest moments. Write the times down. The first listen gives you the shape of the episode and a list of tasks, which is far more useful than starting at the beginning and fixing as you go.
Clicks, dropouts, hum and clipping are faults, and they should be repaired before the structure changes. A fault repaired after a cut can end up on the wrong side of a join. Fix each one at its position in the raw file, then check that the repair sounds natural in context.
Next, shape the episode. Remove the repeats, the tangents that do not serve the story, the long dead air and the false starts. Work on the story rather than the sound at this stage, and keep the natural pauses that help the listener follow. The edit by words guide shows a fast way to make these cuts from a transcript.
Once the structure is set, clean the voices and balance them. Reduce steady noise gently, then level each speaker so they sit together. Clean after cutting, not before, because processing applied to a long file may behave differently at each join. The levelling guide explains the method for two or more speakers.
Add music, intros and effects last, once the voices are settled. Lower the music under speech, and check that every element is licensed for your use. Music can hide problems in the voices, so listen to the speech on its own after adding it, to be sure it is still clear.
Play the finished episode from start to end, on headphones and on a phone speaker. Check the joins, the levels, the music balance and the ending. Listen for anything that makes you want to turn the volume up or down, since those moments point to the problems a listener will feel.
Export the file in the format your platform requires, and check its loudness and true peak. Keep the master project and the export together, so you can re-export later for a different platform or a clip.
Each stage depends on the one before. Cleaning before cutting spreads effort over material you will remove. Adding music before levelling the voices hides problems in the speech. Exporting before checking the whole episode means a fault may slip out. Working in order keeps each decision meaningful, and it prevents the endless loop of fixing and re-fixing.
For music, the order changes: the structure and arrangement come first, then the balance of parts, then the processing. For video, the cuts follow the picture, so the audio is edited first and the picture is cut to match. The principle of working in a deliberate order still applies, and the best workflow is the one you can repeat.
Write the steps down as a checklist, and tick them off for each episode. A checklist protects the workflow on busy weeks, when the temptation to skip a step is strongest. Review the list every few months, and add any step that would have caught a problem you missed.
A workflow is a guide, not a rule. Some episodes need a different order, and some editors prefer to work in different ways. The aim is a repeatable process that produces a clean, finished episode. If your process is working, keep it, and change it only when a specific problem shows you why.
A final check before publishing catches the problems that editing hides. The Vocal Checker shows clipping, noise and level problems in a recording you own, and the Tunedzy Go page covers cleaning a phone recording. For the cutting technique itself, read the long recording guide in this series.
Spot clipping, noise, rumble and level problems. Runs in your browser, nothing is uploaded.
Backing up the raw file and never changing it. Everything else can be undone if the original is safe.
After cutting, so processing is applied to the material you keep, not the material you remove.
Last, once the voices are settled, so it does not hide problems in the speech.
The order changes for music, with structure and balance first, but working deliberately still applies.
