The short answer

Before mastering a podcast, export a clean, uncompressed WAV or AIFF file with enough headroom for a later mastering pass. Keep the original multitrack recording untouched. Do not apply final loudness normalization or a restrictive peak limiter during the pre-master export.

When the mastered episode is ready, create a separate delivery copy. For most spoken-word workflows, a common practical baseline is −16 LUFS integrated loudness with true peak at or below −1 dBTP, but the written brief from your host, network or client takes priority. Apple Podcasts publishes guidance around −16 dB LKFS with a ±1 dB tolerance and a true peak that does not exceed −1 dB FS. Spotify’s −14 LUFS material concerns music playback and should not be treated as a universal podcast upload rule.

In short:

  • Edit and preserve: Keep the original multitrack session and source files.
  • Pre-master export: Use an uncompressed WAV or AIFF with headroom.
  • Master separately: Set integrated loudness and true-peak targets deliberately.
  • Delivery export: Create the format requested by the distribution destination.
  • Archive: Save the source, pre-master, mastered master and delivery copy separately.
  • Verify: Check measurements, metadata, duration and the final file before publishing.

Why the pre-master and delivery file are different jobs

A pre-master is not the file you should normally upload. Its job is to carry the finished edit into mastering without throwing away quality or committing the episode to one final level.

A delivery file has a different purpose. It must work inside the RSS feed, hosting platform or client’s processing chain. That is why producing a “podcast-ready” file directly from the editing timeline can work against you: it may skip a mastering stage, double-process loudness, or use a lossy format that is no longer needed at that point.

Use two distinct outputs:

Output Main purpose Practical choice
Edit or pre-master Give the mastering editor a clean, lossless starting point WAV or AIFF, usually 24-bit PCM
Archival master Preserve the approved finished audio Lossless WAV or AIFF, with the measured loudness and peak recorded
Distribution copy Upload to the podcast host The host’s specified format and settings, commonly MP3 or another supported delivery format

The exact bit depth can depend on your editing software, the original recording and the receiving workflow, but the important decision is whether the intermediate file is lossless and has useful headroom.

1. Preserve the sources before changing anything

The safest pre-master workflow begins before export. Preserve the original recordings separately from the edited project and keep multitrack sources intact whenever possible.

A practical folder structure looks like this:

Episode-024/
├── 01-sources/
├── 02-edit-session/
├── 03-premaster/
├── 04-master/
├── 05-delivery/
└── 06-documentation/

Use descriptive names rather than a pile of files called final, final2 and really-final. Include the episode slug, version and date where useful. If you collaborate with an editor, agree on a naming convention before the first export.

Do not overwrite a source when replacing or cleaning up the audio. If your workflow includes archival storage and later migration, the podcast audio archive workflow covers naming, source preservation, metadata and checksums in more detail.

2. Choose the pre-master export settings

For a typical edited voice track, choose these settings unless your mastering engineer provides a different specification:

  • Format: WAV or AIFF
  • Encoding: Linear PCM, uncompressed
  • Bit depth: 24-bit is a practical studio choice; retain the source depth if it is already higher
  • Sample rate: Match the source material, commonly 44.1 or 48 kHz, unless the production brief says otherwise
  • Channels: Preserve the project’s native mono or stereo layout rather than duplicating channels for convenience
  • Processing: No final mastering limiter, codec encoding or irreversible cleanup at this stage
  • Headroom: Leave enough level headroom for compression, EQ, leveling and final peak control

Why not export MP3 first?

An MP3 delivery copy may be necessary for the podcast host, but it is a poor hand-off between editing and mastering unless no other option exists. Lossy encoding discards information and can introduce artifacts that are harder to correct later.

The reverse order is more defensible:

Lossless sources → edited session → lossless pre-master → mastered master → required delivery copy

If your editing tool exports only compressed audio, check whether it also supports a separate lossless mixdown. If not, document that limitation and avoid creating multiple lossy generations.

Should you normalize loudness before mastering?

Usually, no—not as the final operation. Early normalization can make the mix sound louder without making the underlying balance better, and it reduces the mastering editor’s room to make level-related decisions.

You may still need a rough level adjustment for monitoring, but treat that as a working control. Save a clean pre-master without final normalization, limiting or encoding.

If you want help choosing between automatic enhancement and a manual editing workflow, see automatic podcast enhancement vs manual editing.

3. Set loudness as a deliberate workflow, not a universal constant

LUFS measures perceived loudness across a program. True peak measures the reconstructed peak level, including peaks that can occur between digital samples. You need both checks:

  • Integrated loudness: How loud the episode is overall.
  • True peak: How close the signal comes to clipping after conversion or encoding.

The practical target depends on the destination and the content. PRX, for example, publishes a working range of −16 to −19 LUFS and notes that there is no single industry standard. Mono and stereo files can also behave differently during playback because a mono file may be routed to both channels.

A sensible decision sequence is:

  1. Find the written delivery brief. Check with the network, sponsor, client or podcast host.
  2. Choose a target within that brief. If none exists, around −16 LUFS is a defensible starting point for stereo spoken-word production.
  3. Set the peak ceiling separately. A ceiling at or below −1 dBTP is a common practical choice for a delivery master.
  4. Measure the actual exported file. Do not rely only on the editor’s project display.
  5. Check episode-to-episode consistency. Keep the same measurement method and targets across the show.
  6. Listen on more than one playback path. Use headphones, a phone speaker and a car or Bluetooth setup.

This is a normalization workflow, not a guarantee that every listener will hear identical volume. Platforms may transcode, normalize or process files, while devices, apps and listener settings differ. Check the published episode after release rather than assuming the source file’s measurements tell the entire story.

Avoid these common loudness mistakes

  • Confusing peak normalization with loudness normalization. Matching the loudest peak does not create a consistent perceived level across episodes.
  • Treating −14 LUFS as a podcast rule. That number appears in music-oriented playback guidance; it is not proof that every podcast file should be mastered to it.
  • Using one number for mono and stereo without checking playback. A −19 LUFS mono convention is not interchangeable with a −16 LUFS stereo convention.
  • Crushing dynamics to solve a mix problem. If speech is inconsistent because the performance or recording level changed, loudness normalization cannot repair the underlying balance.
  • Checking only the waveform. A waveform may look healthy while still containing clipping, a missing tail or an incorrect channel configuration.

4. Create the delivery copy from the approved master

Do not overwrite the archival master with a host-specific upload file. Export a new copy from the mastered master using the destination’s requirements.

Many podcast workflows use MP3 for distribution, but the correct choice must come from the hosting platform or client brief. Check these items before export:

  • Supported container and codec
  • Mono or stereo channel requirement
  • Sample rate and bitrate
  • Maximum file size
  • Whether the host transcodes the file
  • Required file naming convention

Do not assume that a larger bitrate or a WAV upload improves the final listening experience. If the host expects MP3 and transcodes uploaded files, follow that brief rather than creating an unsupported experimental delivery format.

The main rule is simple: the mastered archival file and the distribution file should both exist, but they need not be the same format.

5. Add metadata without confusing file metadata and show metadata

Podcast episode metadata includes more than what is embedded in the audio file.

At the file level, confirm:

  • Episode title
  • Artist or show name
  • Version or “master” label
  • Track number if your workflow uses one
  • Creation or modification date
  • Any notes required by the mastering engineer

At the episode or RSS level, verify separately:

  • Episode title and description
  • Publish date and episode number
  • Artwork
  • Episode artwork where supported
  • Transcript
  • Chapter markers
  • Explicit-content flag
  • Sponsor or ad disclosures

Do not assume that tags embedded in an MP3 or WAV will populate every field in your podcast host. Enter and verify the episode-level metadata in the publishing system itself.

6. Verify the final file before distribution

A delivery file can meet the correct settings and still be wrong. Before uploading, run a short verification pass.

File-level checks

  • The file opens and plays from beginning to end.
  • The duration matches the intended episode length.
  • The first and last words are present, with no accidental fade or truncation.
  • The channel count and sample rate match the brief.
  • The codec, bitrate and file size are accepted by the destination.
  • Integrated loudness is measured on the exported delivery file.
  • True peak is measured on the exported delivery file.
  • There is no clipping or unexpected silence.
  • Sponsor tags, music, IDs and required credits are present where needed.

Listening checks

  • Dialogue remains intelligible on headphones.
  • The mix remains comfortable on a phone speaker.
  • The episode is not obviously louder or quieter than recent episodes.
  • Embedded and side-chain tracks do not clip when mixed together.
  • Music beds and transitions do not hide the conversation.

Publishing checks

  • The uploaded file plays in the host preview.
  • The RSS enclosure points to the intended file.
  • The episode title, number, date and artwork are correct.
  • A private or unlisted feed has been tested before public release.
  • The published episode has been checked on at least one mobile and one desktop path.

For a repeatable final pass, use this compact export card:

EPISODE:
DELIVERY FORMAT / CODEC:
CHANNELS / SAMPLE RATE / BITRATE:
INTEGRATED LOUDNESS:
TRUE PEAK:
DURATION:
MASTER FILE LOCATION:
DELIVERY FILE LOCATION:
VERIFIED BY / DATE:
NOTES:

A decision guide by production situation

If you are editing and mastering yourself

Use a lossless pre-master, then make a separately measured master and a host-specific delivery copy. This adds a file and a short verification stage, but it preserves control and makes revision easier.

If you are sending the episode to an editor

Agree on the pre-master format, target loudness, peak ceiling, naming convention and required metadata before they begin. Do not send a previously compressed copy unless the workflow requires it.

If a client provides a strict delivery brief

Treat that brief as the source of truth. Keep an archival master separately, but do not let generic podcast conventions override the client’s written requirements.

If your show has a history of inconsistent playback

Do not immediately change the target. Compare recent episodes, inspect their measurements and consider whether the difference comes from mono/stereo handling, gain staging, platform processing or inconsistent editing.

If you publish video as well as audio

Create the podcast master and delivery audio from the shared edit, but do not assume the video platform’s loudness guidance is interchangeable with the podcast host’s requirements. Each destination can have its own brief and processing chain.

Frequently asked questions

What audio file should I upload to a podcast host?

Use the format required by that host or client. Many accept MP3, but the supported codec, channel layout, sample rate and bitrate can vary. Keep a lossless archival master and create the upload copy from it.

Should I upload WAV or MP3?

Upload whichever the destination explicitly supports. For distribution, an MP3 is common; for editing and mastering hand-offs, WAV or AIFF is generally safer because it avoids an extra lossy generation.

Is −16 LUFS the only correct podcast loudness target?\n

No. It is a useful reference point, and Apple publishes guidance around it, but a written client or network brief takes priority. Mono and stereo playback differences also matter. The important goal is a clean, consistent master that meets the destination’s requirements.

Should I normalize every track separately?

Usually not. Normalizing individual tracks can disrupt the intended balance between speech, music and effects. Build and level the complete episode first, then measure and process the finished mix.

Can a loudness-normalized master still clip?\n

Yes. Raising perceived loudness and controlling peaks are separate tasks. Measure true peak after processing and export, and use a suitable limiter or gain adjustment as part of the mastering chain.

Final pre-upload checklist

Before approving the episode, confirm that you have:

  1. Preserved the original source recordings and edit session.
  2. Exported a clean, lossless pre-master with headroom.
  3. Created and approved a separate mastered master.
  4. Measured integrated loudness and true peak on the exported file.
  5. Compared the result with recent episodes.
  6. Created a delivery copy using the destination’s required settings.
  7. Verified file metadata and episode-level publishing metadata.
  8. Archived the source, pre-master, master, delivery copy and verification record.
  9. Listened to the uploaded episode after publishing.

The best export setting is not the one with the most impressive numbers. It is the setting that preserves your options, follows the destination’s brief, produces a consistent listening experience and leaves you with files you can still use when the show, team or platform changes.

Sources