Radio Ad Production Checklist: From Voice-Over Recording to Final Broadcast WAV/MP3

If a radio spot sounds fine in the studio but fails at delivery, the problem is usually not creative. It is process. The cleanest save is a checklist: capture the voice-over properly, edit with discipline, mix against a clear reference, and export the right file the first time.

Before you book the session, a few questions matter: Is the voice clean enough? Does the timing still match the spot length? Are the export settings correct? Will the broadcaster or client reject a clipped file for reasons nobody wants to explain twice? Those are the practical decisions that save time. As Ableton’s loudness primer notes, loudness is not the same thing as peak level; that distinction is where many delivery mistakes begin.

Radio production is a small chain of dependent steps. Recording, editing, mixing, and export each introduce their own risk. For a useful technical baseline, I also like to keep YouTube’s recommended audio specs and Audacity’s manual nearby when checking sample rates, mono/stereo choices, and basic cleanup workflows. The standards may vary by buyer and station, but the logic is stable: control the source, verify the mix, and document the deliverable.

In this article, I will walk through a studio-style checklist for finishing a radio ad correctly. You will see what to prepare before recording, how to clean and align the VO, how to manage loudness without flattening the spot, which export settings are the safest defaults, and how to name and version files so approvals stay readable instead of theatrical.

Terminology: the terms that actually matter

Term	What it means in practice
VO	Voice-over recorded for the spot.
Bed	The music track under the voice.
Gain staging	Setting levels so the recording is clean without clipping.
Clipping	Digital distortion caused when audio exceeds the available headroom.
Loudness	Perceived level, not just the tallest waveform peak.
WAV	Uncompressed delivery format, usually safest for master files.
MP3	Compressed format, usually useful for review copies or secondary delivery.

1. What to prepare before you record

Start with the documents that prevent avoidable confusion. I want the script, timing notes, brand references, pronunciation notes, and any required legal line in one place before the microphone is live. If the client has already approved a direction, keep that note visible. If they have not, separate the facts from the taste decisions. That is where approval chains get lost.

• Script version with the final read length target.
• Pronunciation guide for names, places, offers, and numbers.
• Reference spots if the client wants a similar pace or tone.
• Brand notes covering energy level, audience, and any forbidden wording.
• Delivery requirements such as WAV or MP3, mono or stereo, and preferred naming.

A sensible default is to review the spot against the project brief and the relevant service page before you hit record. If you need to align the production with the rest of the studio offering, the services page helps frame the work, while the portfolio is a better place to check how finished campaigns are usually presented.

2. Record a clean voice-over

Here the job is boring, which is good. Keep the mic distance consistent, reduce room noise, and set gain so the loudest takes never flirt with clipping. A dry studio voice usually beats a “characterful” room tone. Studios sometimes romanticize the ambience. The waveform does not share that enthusiasm.

Three practical rules hold up well:

1. Maintain a fixed mic distance so the tonal balance stays stable across takes.
2. Control the room by turning off fans, phones, and anything that drifts into the noise floor.
3. Capture backups-at least one extra pass, often two, because the take you love is sometimes the take that coughs at the last line.

When the session has more than one voice or read style, record a clean slate for each variation. That makes later comping straightforward. For background on audio workflow terminology and practical editing basics, the Audacity documentation is a reliable reference, especially for teams that move between tools.

3. Edit the spot before you mix it

The editing pass is where you remove the obvious distractions: breaths that jump out, clicks, mouth noises, chair squeaks, stray pops, and any pauses that stretch the spot past time. I prefer to think of this as reducing friction, not sterilizing performance. A radio ad should still sound human. It just should not sound under-rehearsed.

Watch the duration while you edit. If the approved script is 30 seconds, do not discover at the end that the final read is 32.4 seconds and “close enough” has become a proposal. Trim silence at the top and tail, smooth obvious cuts, and make sure the final read lands where it should.

Editing task	What to watch for	Reason it matters
Breath control	Remove only the distracting ones	Avoids a robotic read
Noise cleanup	Reduce hiss, hum, or room rumble	Keeps the VO clear under music
Timing alignment	Match the approved spot length	Prevents delivery mismatches
Comping	Choose the best phrase or sentence from backups	Improves delivery without overprocessing

If you need the broader studio workflow context, the earlier article on radio commercial production from script to spot covers the planning stage. This checklist picks up after the read is already written.

4. Mix the VO and music bed carefully

The mix is not the place to prove artistic courage. It is the place to make the message intelligible. Align the music bed so it supports the rhythm of the read, then check the VO/music balance at both low and normal monitoring volume. If the voice disappears when the speakers get smaller, the balance is wrong.

Basic EQ and compression can help, but they should serve clarity, not volume for its own sake. Use EQ to remove muddiness or harshness. Use compression lightly if it smooths the level, not if it makes the VO sound squeezed into a tax form.

• Music bed alignment: fade in and out cleanly, and avoid clashes with the voice’s important words.
• VO/music balance: keep the voice intelligible even when the music is energized.
• EQ: reduce clutter in the low mids and watch for brittle highs.
• Comping: choose one consistent tone and level across the final read.

For a practical discussion of loudness and perceived level, Ableton’s guide to loudness is useful because it separates peak concerns from the way listeners actually hear audio.

5. Check loudness, peaks, and silence

Peak level tells you one thing; loudness tells you another. That distinction is why an ad can look “safe” on a meter and still feel aggressive, or look healthy and still sound too quiet beside other spots. Keep enough headroom to avoid clipping, but also check the spot against other approved deliverables so no version suddenly behaves like a different species.

At minimum, verify the following:

• No clipped peaks in the VO or master.
• No long dead air at the front or end unless the brief calls for it.
• Consistent loudness across versions of the same campaign.
• One clean master file, not five nearly identical ones with mystery suffixes.

If your client shares a target loudness or delivery note, honor it. If they do not, use a reasonable internal house standard and document it. The key is consistency and traceability, not bravado.

6. Export the right delivery files

This is where many otherwise good spots fail. Export settings are not decorative. They decide whether the file is portable, editable, and usable by the next person in the chain. The safest default is usually a master WAV, with an MP3 only when requested for review or secondary delivery.

Format	Best use	Default note
WAV	Final master delivery	Preferred for broadcast handoff unless another spec is given
MP3	Client review or secondary copy	Use only when compression is acceptable
Mono	Many VO-led spots and compatibility-safe delivery	Often the simplest broadcast option
Stereo	Music-driven spots or creative mixes that depend on width	Use only if the production benefits from it

Match sample rate and bit depth to the project’s delivery needs, and keep them consistent with the production session whenever possible. For a quick reference on how compressed delivery behaves, Google’s audio specs guidance is a useful reminder that format choices and codec choices are not the same decision.

7. Name files and metadata clearly

Good file naming is not glamourous, but it saves approvals from becoming detective work. Include the client name, spot length, version number, and date. If the client asks for multiple edits, the naming convention should let anyone identify the latest approved file without opening six near-identical exports.

A clean pattern looks like this:

ClientName_30s_v03_2026-09-23.wav

That pattern is boring in the best way. It tells the truth without making anyone guess.

• Client name: short, stable, and spelled consistently.
• Spot length: 15s, 20s, 30s, 60s, or the approved runtime.
• Version: increment every time the approved content changes.
• Date: use one format everywhere.

If the team also needs a public-facing contact route for revisions or follow-up, use the contact page rather than burying requests in email threads that only one person can find later.

8. Keep draft, revision, and final versions separate

Version control is how you avoid fake certainty. Drafts are for experimentation. Revisions are for approved changes. Finals are for delivery. Mixing those states is how teams end up broadcasting a file that still contains yesterday’s music bed or a line the client already rejected.

Status	Purpose	What should be true
Draft	Early creative exploration	Unused options may still exist
Revise	Client feedback applied	Changes are tracked and visible
Final	Approved delivery	No unresolved comments remain

Keep the final file in a separate folder or at least a separate naming tier. Responsibility is not optional; file structure is one of the few places where that sentence can be made practical.

9. Common failure points and how to prevent them

Most delivery failures are ordinary. That is the bad news and the good news. If they are ordinary, they are preventable.

Failure point	What it looks like	How to prevent it
Wrong duration	Spot runs long or short	Check timing during edit, not after export
Clipping	Distorted peaks	Leave headroom and monitor loud passages
Noisy VO	Hiss, hum, room tone	Fix the recording environment before editing
Inconsistent levels	One spot feels louder than the next	Use a house reference and compare exports
Unclear naming	Multiple “final_final” files	Use version numbers and a fixed format

The safest rule is simple: if the session, edit, or export relies on memory instead of a checklist, expect avoidable mistakes. Chaos is easy. Order is the bill that arrives later.

Checklist summary

• Prepare the script, timing, references, and pronunciation notes before recording.
• Record a clean VO with stable mic distance and controlled room noise.
• Edit breaths, clicks, and silence before you mix.
• Balance VO and music so the message stays intelligible.
• Check peak level, loudness, and silence at the end of the chain.
• Export a master WAV and an MP3 only when it is actually needed.
• Name files clearly and separate draft, revision, and final versions.
• Use the safest reasonable default when the brief does not specify a detail.

When the production is finished, the best result is almost invisible: the client hears a spot that sounds intentional, readable, and ready. That is the whole game. Decide early, verify twice, and export once.